<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>GraphKit Blog</title>
    <link>https://graphkit.com/blog/</link>
    <atom:link href="https://graphkit.com/blog/rss.xml" rel="self" type="application/rss+xml"/>
    <description>Practical notes on outbound across LinkedIn, Instagram, WhatsApp, Telegram, Facebook and email: limits, consent, sequencing, inboxes and agency operations.</description>
    <language>en-gb</language>
    <lastBuildDate>Sun, 15 Nov 2026 16:00:00 GMT</lastBuildDate>
    <image><url>https://graphkit.com/web-app-manifest-192x192.png</url><title>GraphKit Blog</title><link>https://graphkit.com/blog/</link></image>
    <item>
      <title>LinkedIn invitation limits in 2026 and how to pace them</title>
      <link>https://graphkit.com/blog/linkedin-invitation-limits/</link>
      <guid isPermaLink="true">https://graphkit.com/blog/linkedin-invitation-limits/</guid>
      <description>What LinkedIn documents about invitation limits, what it leaves out, and how to pace invites per account: 15 a day by default, 80 a week, ceilings 25 and 100.</description>
      <content:encoded><![CDATA[<p>LinkedIn limits how many connection invitations you can send, and it does not tell you the number. That makes “how many invites per day is safe” one of the most-asked questions in outbound and one of the least well answered. This post separates three things that usually get blended together: what LinkedIn actually documents, what nobody outside LinkedIn knows, and the pacing defaults we chose for GraphKit and why.</p>
<p>One caveat first, because it frames everything below. LinkedIn’s <a href="https://www.linkedin.com/legal/user-agreement">User Agreement</a> (section 8.2) prohibits using bots or other unauthorised automated methods to add contacts or send messages, and its <a href="https://www.linkedin.com/help/linkedin/answer/a1341387">page on prohibited software</a> says accounts that use such tools risk being restricted or shut down. No tool, including ours, can promise that an account will not be restricted. Pacing reduces risk; it does not remove it.</p>
<h2 id="what-linkedin-documents-about-invitation-limits">What LinkedIn documents about invitation limits</h2>
<p>LinkedIn’s help article on <a href="https://www.linkedin.com/help/linkedin/answer/a550555">invitation limits</a> is short. These are the points it makes, in its own words where it matters:</p>
<ul>
<li><strong>Everyone is limited.</strong> “All LinkedIn members, including those with Basic and Premium accounts, are subject to these limits.” Paying for Premium or Sales Navigator does not buy a higher invitation allowance.</li>
<li><strong>Hitting the limit restricts invitations, usually for a week.</strong> Your account “may be temporarily restricted from sending invitations”, the restriction “typically lasts one week”, and LinkedIn “cannot remove or shorten the wait period”.</li>
<li><strong>Withdrawing invitations does not help after the fact.</strong> “Withdrawing pending invitations will not remove the restriction.”</li>
<li><strong>Personalised notes are capped for free members.</strong> Basic members can include a personalised message on five connection requests per month. Premium members can add one to every request.</li>
<li><strong>LinkedIn’s advice is to send fewer, better invitations</strong>, ideally to people you know.</li>
</ul>
<p>LinkedIn’s <a href="https://www.linkedin.com/legal/professional-community-policies">Professional Community Policies</a> add one more rule that is directly relevant: do not use the invitation feature to send promotional messages to people you don’t know. That shapes what belongs in a connection note, which we cover in <a href="https://graphkit.com/blog/linkedin-connection-notes-that-get-replies/">writing connection notes that get replies</a>.</p>
<h2 id="what-linkedin-does-not-publish">What LinkedIn does not publish</h2>
<p>The number. LinkedIn gives no daily or weekly invitation figure anywhere in its help centre. The figures that circulate (“about 100 a week”, “around 200 a week”) come from vendors and practitioners comparing notes. Some of them are reasonable observations, but they are estimates, they change, and they vary by account. Treat any specific number as a starting assumption, not a rule.</p>
<p>What does seem consistent across those reports, and matches LinkedIn’s own advice, is that limits behave per account and react to how invitations are received. An account whose invitations are mostly ignored or rejected is in a worse position than one whose invitations are mostly accepted, at the same volume.</p>
<h2 id="graphkits-defaults-and-ceilings">GraphKit’s defaults and ceilings</h2>
<p>Because the platform number is unknown, we set conservative defaults and hard ceilings that no workspace can raise. They apply per connected LinkedIn account, across all campaigns that account runs in, so two campaigns cannot each spend a full allowance.</p>
<table>
<thead>
<tr>
<th>Action (per account)</th>
<th>Default</th>
<th>Ceiling you cannot exceed</th>
</tr>
</thead>
<tbody>
<tr>
<td>Invitations per day</td>
<td>15</td>
<td>25</td>
</tr>
<tr>
<td>Invitations per rolling week</td>
<td>80</td>
<td>100</td>
</tr>
<tr>
<td>Messages per day</td>
<td>50</td>
<td>100</td>
</tr>
<tr>
<td>Profile visits per day</td>
<td>40</td>
<td>80</td>
</tr>
<tr>
<td>InMails per day</td>
<td>20</td>
<td>Your InMail credits</td>
</tr>
</tbody>
</table>
<aside class="callout callout-note" aria-label="Recommendations, not platform rules"><p class="callout-title">Recommendations, not platform rules</p><div class="callout-body"><p>These numbers are our engine’s defaults. They are not limits LinkedIn has published, and staying under them does not guarantee anything. They are deliberately conservative starting points that you can lower, and raise only up to the ceiling.</p></div></aside>
<p>If you set a value above a ceiling, it is clamped. The weekly ceiling applies even if you never set a weekly limit. Setting a daily limit to zero turns that action off for the account. You can see and change these per account on the <a href="https://graphkit.com/senders/">connected accounts</a> screen.</p>
<h3 id="why-there-is-a-weekly-cap-as-well-as-a-daily-one">Why there is a weekly cap as well as a daily one</h3>
<p>A daily cap alone lets an account send its maximum every day, including weekends, which adds up quickly. The weekly cap is a rolling seven days: today plus the six days before it, in the account’s time zone. When the week is full, the next invitation waits (the step shows <code>weekly_quota_exhausted</code>) until enough old days roll out of the window.</p>
<p>At the defaults the two caps rarely fight. Fifteen a day across a five-day working week is 75, just under the weekly 80. They start to interact when you raise the daily number: at 25 a day, the weekly ceiling of 100 is reached on the fourth working day, and the account simply sends nothing more until the window moves on. That is intended. The weekly ceiling is what keeps a high daily setting from turning into a high weekly volume.</p>
<h2 id="pacing-inside-the-day">Pacing inside the day</h2>
<p>How invitations are spread across a day matters as much as the daily count. A hundred actions fired in ten minutes looks nothing like a person working through a list.</p>
<p>GraphKit only acts inside the overlap of two windows: the account’s working hours (by default Monday to Friday, 09:00 to 18:00 in the account’s time zone) and the campaign’s schedule (by default Monday to Friday, 09:00 to 17:00). Both keep their wall-clock times across daylight-saving changes.</p>
<p>Inside that window, actions are spread rather than queued back to back. Each next action waits roughly the remaining open time divided by the remaining actions plus one, so a campaign that starts at 15:00 spreads its work over the two hours left instead of bursting. There is a minimum gap of two minutes between any two actions from the same account, and a small, bounded amount of jitter on top of the nominal gap. To be clear about the purpose: that jitter keeps many accounts from acting in lockstep. It is not designed to disguise automation, and we don’t describe it that way.</p>
<h2 id="acceptance-rate-is-the-number-to-watch">Acceptance rate is the number to watch</h2>
<p>The daily count is the number people argue about. The acceptance rate is the one that tells you whether an account is in trouble.</p>
<p>GraphKit pauses a LinkedIn account automatically when its invitation acceptance falls below <strong>15% over the last seven days, once it has sent at least 30 invitations</strong> in that period, and notifies the workspace. The threshold is ours, not LinkedIn’s. It exists because a low acceptance rate usually means the targeting or the note is wrong, and sending more of the same rarely fixes that.</p>
<p>When an account is paused for low acceptance, look at three things before resuming:</p>
<ol>
<li><strong>Who you are inviting.</strong> Are these people who have a reason to know your name, or a scraped list that loosely matches a job title?</li>
<li><strong>Your profile.</strong> The person deciding whether to accept looks at your headline and photo, not your sequence.</li>
<li><strong>The note, or the absence of one.</strong> A specific one-line reason usually beats a pitch. More on that in <a href="https://graphkit.com/blog/linkedin-connection-notes-that-get-replies/">connection notes and follow-ups</a>.</li>
</ol>
<h3 id="pending-invitations">Pending invitations</h3>
<p>Old pending invitations accumulate. GraphKit does not withdraw them automatically today (it is on our roadmap as a step option), and LinkedIn says withdrawing them does not lift a restriction once you have one. Withdrawing invitations that have sat unanswered for weeks is still sensible housekeeping, done manually and before you run into a limit. A campaign can also branch on “still pending after N days” so a lead who never accepts gets a different path rather than a second invitation. GraphKit never sends a second invitation to someone who already has one pending, or who is already connected.</p>
<h2 id="new-accounts-and-recently-restricted-accounts">New accounts and recently restricted accounts</h2>
<p>GraphKit has no automatic warm-up or ramp schedule for new accounts. That is a deliberate gap we would rather state than paper over. Our recommendation, which again is not a platform rule, is to start a new or newly connected account well below the default (five to ten invitations a day for the first two weeks) and raise it only while acceptance stays healthy. The same applies to an account coming back from a restriction.</p>
<h2 id="what-happens-when-linkedin-says-no">What happens when LinkedIn says no</h2>
<p>If LinkedIn refuses an invitation because the account has hit a limit, GraphKit treats the step as out of quota for that account and retries it on the account’s next quota day. Rate-limit responses are retried with back-off, and an action whose outcome is unknown is reconciled against the platform rather than resent blind. If LinkedIn restricts the account, its status changes to restricted and campaigns stop using it until you resolve it.</p>
<h2 id="planning-volume-with-these-numbers">Planning volume with these numbers</h2>
<p>Because limits are per account, capacity is a function of how many accounts you connect, not of a setting you turn up. At the defaults, one LinkedIn account sends up to 75 invitations in a normal working week (the weekly cap of 80 leaves a little room for a sixth day if you schedule one).</p>
<table>
<thead>
<tr>
<th>Connected LinkedIn accounts</th>
<th>Invitations per week at defaults</th>
<th>Per four-week month</th>
</tr>
</thead>
<tbody>
<tr>
<td>1</td>
<td>about 75</td>
<td>about 300</td>
</tr>
<tr>
<td>5</td>
<td>about 375</td>
<td>about 1,500</td>
</tr>
<tr>
<td>20</td>
<td>about 1,500</td>
<td>about 6,000</td>
</tr>
</tbody>
</table>
<p>If the plan needs more than that, the honest options are more accounts (each owned by a real person who agreed to it), better targeting so each invitation counts, or other channels for people who have already shown interest. <a href="https://graphkit.com/pricing/">Pricing</a> is per connected account, so the cost of adding capacity is predictable.</p>
<h2 id="a-short-checklist">A short checklist</h2>
<ul>
<li>Keep each account at or below 15 invitations a day until you have a few weeks of acceptance data.</li>
<li>Leave the weekly cap on. It is what makes a higher daily setting safe to try.</li>
<li>Start new and recovering accounts lower, and raise slowly.</li>
<li>Watch acceptance, not just volume. Below 15% for a week, stop and fix the list or the note.</li>
<li>Withdraw stale pending invitations by hand, before you need to.</li>
<li>Never put a pitch in an invitation. Save it for after the connection, if at all.</li>
</ul>
<p>GraphKit is not affiliated with or endorsed by LinkedIn. For how we handle account connections and platform risk in general, see the <a href="https://graphkit.com/security/#platforms">platforms section of our security page</a>.</p>]]></content:encoded>
      <pubDate>Sun, 15 Nov 2026 16:00:00 GMT</pubDate>
      <dc:creator>GraphKit team</dc:creator>
      <category>LinkedIn</category>
      <category>Limits</category>
      <category>Pacing</category>
    </item>
    <item>
      <title>LinkedIn connection notes and follow-ups that get replies</title>
      <link>https://graphkit.com/blog/linkedin-connection-notes-that-get-replies/</link>
      <guid isPermaLink="true">https://graphkit.com/blog/linkedin-connection-notes-that-get-replies/</guid>
      <description>How to write LinkedIn connection notes and follow-up messages people answer, with templates that use variables and fallbacks so no lead gets a broken message.</description>
      <content:encoded><![CDATA[<p>Most connection notes fail for the same reason: they are written for the sender. “I’d love to add you to my network” tells the reader nothing. “We help companies like yours 10x pipeline” tells them you want something. The notes that get accepted, and the follow-ups that get answered, give a specific reason to talk that the reader recognises as true.</p>
<p>This post covers what a note is for, when to skip it, how to write follow-ups after someone accepts, and how to build templates with variables and fallbacks so a missing field never produces “Hi ,” in someone’s inbox. The templates use the variable syntax GraphKit’s <a href="https://graphkit.com/campaigns/">campaign builder</a> understands, but the writing advice applies whatever you use.</p>
<h2 id="the-rules-you-are-writing-inside">The rules you are writing inside</h2>
<p>A few constraints come from LinkedIn and are not negotiable:</p>
<ul>
<li><strong>Notes are short.</strong> GraphKit checks invitation notes against 300 characters, and 200 for free accounts.</li>
<li><strong>Free accounts barely get notes at all.</strong> LinkedIn’s <a href="https://www.linkedin.com/help/linkedin/answer/a550555">invitation limits page</a> says Basic members can personalise five connection requests a month. Premium members can personalise all of them.</li>
<li><strong>No pitching in invitations.</strong> LinkedIn’s <a href="https://www.linkedin.com/legal/professional-community-policies">Professional Community Policies</a> say not to use the invitation feature to send promotional messages to people you don’t know.</li>
<li><strong>Messages only reach connections.</strong> A regular LinkedIn message goes to first-degree connections. Before someone accepts, the note is all you get (InMail aside, which needs Sales Navigator or Recruiter and spends credits).</li>
</ul>
<p>Put together: the note’s only job is to get the invitation accepted. Selling, if it happens at all, happens later.</p>
<h2 id="when-to-send-no-note-at-all">When to send no note at all</h2>
<p>A blank invitation is a reasonable default. It costs nothing, it does not use one of a free account’s five monthly notes, and nothing in it can read as a pitch. People decide on a blank invitation by looking at your name, photo and headline, which is a good reason to make those clear before you start.</p>
<p>Add a note when you have a reason the reader would recognise without explanation: you met, you were both at an event, they wrote something you have an actual response to, you share a specific community, or someone they know suggested it. If the best you can write is “I came across your profile”, leave it blank.</p>
<h2 id="what-a-good-note-looks-like">What a good note looks like</h2>
<p>A note that works has three parts, and often fits in two sentences:</p>
<ol>
<li><strong>The real reason you are reaching out</strong>, specific to them.</li>
<li><strong>A light bridge</strong> to why connecting makes sense.</li>
<li><strong>No ask.</strong> No meeting request, no link, no deck.</li>
</ol>
<p>Compare:</p>
<blockquote>
<p>Hi Dana, I’d love to connect and share how we help teams like yours grow pipeline.</p>
</blockquote>
<blockquote>
<p>Hi Dana, your post on routing inbound leads by territory matched a problem we’re working through. Would be good to follow your writing.</p>
</blockquote>
<p>The first is about the sender. The second is about something Dana did, and it is true. It is also only possible if someone read the post, which is the point: personalisation that a merge field can produce on its own is rarely personal.</p>
<h2 id="templates-with-variables-and-fallbacks">Templates with variables and fallbacks</h2>
<p>Templates make the structure repeatable. Variables fill in the parts that differ per lead. Fallbacks decide what happens when a field is empty, which on real lead lists is often.</p>
<p>GraphKit’s syntax is double braces with an optional fallback after a pipe:</p>
<pre class="astro-code github-dark" style="background-color:#24292e;color:#e1e4e8;overflow-x:auto" tabindex="0" data-language="plaintext"><code><span class="line"><span>{{first_name}}          the lead&#39;s first name</span></span>
<span class="line"><span>{{first_name|there}}    &quot;there&quot; when the lead has no first name</span></span>
<span class="line"><span>{{first_name|}}         nothing at all when it is missing</span></span>
<span class="line"><span>{{company|your team}}   company, or &quot;your team&quot;</span></span>
<span class="line"><span>{{trigger}}             a custom field from your import</span></span></code></pre>
<p>Built-in fields include <code>first_name</code>, <code>last_name</code>, <code>full_name</code>, <code>company</code>, <code>job_title</code>, <code>headline</code>, <code>location</code> and the sender’s own name fields. Any other column you import becomes a custom field you can reference by name. Names are case-insensitive, spaces inside the braces are ignored, and a value that is only whitespace counts as missing.</p>
<h3 id="note-templates">Note templates</h3>
<p>Each of these stays under 200 characters with typical values, so they work on free accounts within the monthly allowance.</p>
<p><strong>Shared event or community.</strong> Needs a custom <code>event</code> column you filled in yourself.</p>
<pre class="astro-code github-dark" style="background-color:#24292e;color:#e1e4e8;overflow-x:auto" tabindex="0" data-language="plaintext"><code><span class="line"><span>Hi {{first_name|there}}, we were both at {{event}}. I&#39;m working on</span></span>
<span class="line"><span>similar problems in {{topic|outbound}}, would be good to stay</span></span>
<span class="line"><span>in touch.</span></span></code></pre>
<p>Note that <code>{{event}}</code> has no fallback. That is deliberate: if you do not know where you met, this note should not be sent. In GraphKit, a step can carry <strong>fallback copy</strong> (a whole alternative note) that is used when a variable without an inline default is missing. Set the fallback note to something that is still true without the field, or leave the invitation blank.</p>
<p><strong>Something they wrote.</strong> Needs a <code>trigger</code> column, written by a person.</p>
<pre class="astro-code github-dark" style="background-color:#24292e;color:#e1e4e8;overflow-x:auto" tabindex="0" data-language="plaintext"><code><span class="line"><span>Hi {{first_name|there}}, {{trigger}}. Would be good to follow</span></span>
<span class="line"><span>along with what {{company|your team}} is doing there.</span></span></code></pre>
<p>Here <code>trigger</code> is a short phrase like “your post on routing inbound by territory was useful”. It is the only part of the note that matters, and it is the only part a template cannot write for you.</p>
<p><strong>Mutual introduction.</strong> For when someone suggested you talk.</p>
<pre class="astro-code github-dark" style="background-color:#24292e;color:#e1e4e8;overflow-x:auto" tabindex="0" data-language="plaintext"><code><span class="line"><span>Hi {{first_name|there}}, {{referrer}} suggested we connect.</span></span>
<span class="line"><span>Happy to compare notes on {{topic|outbound}} whenever useful.</span></span></code></pre>
<aside class="callout callout-note" aria-label="How GraphKit handles a missing field"><p class="callout-title">How GraphKit handles a missing field</p><div class="callout-body"><p>If a variable is missing and has an inline default, the default is used. If it has no default, the step’s fallback copy is used instead. If that cannot render either, or the result is too long for the channel, the lead stops with a clear reason. A message with an empty gap is never sent. Previews show unresolved variables as <code>{{name}}</code> so you see the problem before launch.</p></div></aside>
<h2 id="follow-ups-after-someone-accepts">Follow-ups after someone accepts</h2>
<p>Acceptance is permission to talk, not interest in buying. The first message after acceptance should continue the reason you gave, and every later message should be useful on its own if it is the only one they read.</p>
<p>A sequence that respects that, written as GraphKit steps:</p>
<ol>
<li><strong>Invite</strong> with a note, or blank.</li>
<li><strong>Condition: connected?</strong> Check after up to a week. Leads who don’t accept take the “no” path, which here simply ends. GraphKit will not send a second invitation to someone with one pending.</li>
<li><strong>Wait one business day</strong> after acceptance.</li>
<li><strong>Message 1.</strong> Continue the original reason. Ask one easy question.</li>
<li><strong>Wait four business days.</strong></li>
<li><strong>Message 2.</strong> Add something useful: a short answer to a common problem, a relevant example, a resource. No new ask beyond the first.</li>
<li><strong>Wait a week.</strong></li>
<li><strong>Message 3.</strong> Close the loop politely and make it easy to say no.</li>
</ol>
<p>Stop-on-reply is on by default, so anyone who answers at any point, on any connected channel, leaves the sequence immediately. The check runs again right before each send, so a reply that arrives seconds before a scheduled message still stops it.</p>
<h3 id="follow-up-templates">Follow-up templates</h3>
<p><strong>Message 1: continue the thread.</strong></p>
<pre class="astro-code github-dark" style="background-color:#24292e;color:#e1e4e8;overflow-x:auto" tabindex="0" data-language="plaintext"><code><span class="line"><span>Thanks for connecting, {{first_name|}}. Since you&#39;re looking at</span></span>
<span class="line"><span>{{trigger_topic|outbound}} at {{company|your company}}: how are you</span></span>
<span class="line"><span>handling {{pain|replies across channels}} today?</span></span></code></pre>
<p>The empty fallback on <code>first_name</code> means a missing name reads “Thanks for connecting.” rather than “Thanks for connecting, there.” Small, but it is the difference between a template and a person.</p>
<p><strong>Message 2: give something.</strong></p>
<pre class="astro-code github-dark" style="background-color:#24292e;color:#e1e4e8;overflow-x:auto" tabindex="0" data-language="plaintext"><code><span class="line"><span>One thing that helped teams we work with: {{tip|keeping every</span></span>
<span class="line"><span>reply in one inbox so nothing waits on someone&#39;s phone}}. Happy</span></span>
<span class="line"><span>to share how they set it up if it&#39;s relevant.</span></span></code></pre>
<p><strong>Message 3: close the loop.</strong></p>
<pre class="astro-code github-dark" style="background-color:#24292e;color:#e1e4e8;overflow-x:auto" tabindex="0" data-language="plaintext"><code><span class="line"><span>I&#39;ll stop here so I&#39;m not cluttering your inbox. If {{topic|this}}</span></span>
<span class="line"><span>becomes a priority later, reply anytime. Either way, good to be</span></span>
<span class="line"><span>connected.</span></span></code></pre>
<h2 id="mistakes-worth-avoiding">Mistakes worth avoiding</h2>
<ul>
<li><strong>Fake familiarity.</strong> “Hope you had a great weekend!” to a stranger reads as automated because it is.</li>
<li><strong>Personalisation that only proves you have a list.</strong> “I see you’re the <code>{{job_title}}</code> at <code>{{company}}</code>” restates their profile back to them.</li>
<li><strong>Links in the first message.</strong> They look like a pitch and give the reader homework.</li>
<li><strong>A meeting request before there is a conversation.</strong> Ask a question they can answer in one line.</li>
<li><strong>Following up after a reply.</strong> If your tool does not stop on reply, across every channel, fix that before you tune copy. We cover this in <a href="https://graphkit.com/blog/multichannel-outbound-without-spamming/">multichannel outbound without spamming</a>.</li>
<li><strong>Testing on everyone at once.</strong> Run a campaign as a dry run first. In GraphKit, a dry run makes the same decisions as a live run without sending anything, so you can check how templates render for real leads.</li>
</ul>
<h2 id="volume-still-matters">Volume still matters</h2>
<p>Good copy does not rescue bad pacing. A well-written note sent 200 times a day from one account is still 200 invitations a day. Keep invitation volume conservative per account (our defaults and reasoning are in <a href="https://graphkit.com/blog/linkedin-invitation-limits/">LinkedIn invitation limits</a>) and spend the effort on who you invite and why. When replies come in, handle them in one place; <a href="https://graphkit.com/blog/unified-inbox-for-outbound-replies/">why a unified inbox matters</a> explains how we think about triage.</p>]]></content:encoded>
      <pubDate>Sun, 15 Nov 2026 15:00:00 GMT</pubDate>
      <dc:creator>GraphKit team</dc:creator>
      <category>LinkedIn</category>
      <category>Copywriting</category>
      <category>Templates</category>
    </item>
    <item>
      <title>Multichannel outbound without spamming people</title>
      <link>https://graphkit.com/blog/multichannel-outbound-without-spamming/</link>
      <guid isPermaLink="true">https://graphkit.com/blog/multichannel-outbound-without-spamming/</guid>
      <description>How to combine LinkedIn, Instagram, WhatsApp, Telegram, Facebook and email in one sequence without spamming: consent per channel, sequencing and stop on reply.</description>
      <content:encoded><![CDATA[<p>Adding channels to outbound is easy. Adding them without becoming the person who messages someone on LinkedIn, then WhatsApp, then Instagram, then email, all in the same week, takes some discipline. The difference is not the tool. It is three decisions: which channels you have permission to use for a given person, in what order you use them, and what stops everything.</p>
<p>This post goes through each, using the rules GraphKit’s <a href="https://graphkit.com/campaigns/">campaign engine</a> enforces as concrete examples. None of it is legal advice; where a platform has published rules, we link to them.</p>
<h2 id="each-channel-comes-with-its-own-permission-model">Each channel comes with its own permission model</h2>
<p>The most common multichannel mistake is treating every channel like email, or every channel like LinkedIn. They have very different expectations about who may message whom.</p>
<table>
<thead>
<tr>
<th>Channel</th>
<th>Who it’s for in outbound</th>
<th>GraphKit default per account</th>
</tr>
</thead>
<tbody>
<tr>
<td>LinkedIn</td>
<td>Professional contacts; invitations and messages to connections</td>
<td>15 invites and 50 messages a day</td>
</tr>
<tr>
<td>Instagram</td>
<td>People who follow you, engaged with you, or were imported with consent</td>
<td>20 messages a day, 3 min apart</td>
</tr>
<tr>
<td>WhatsApp</td>
<td>Existing contacts and people who opted in</td>
<td>20 messages a day, 5 min apart</td>
</tr>
<tr>
<td>Telegram</td>
<td>Existing contacts and people who opted in</td>
<td>20 messages a day, 5 min apart</td>
</tr>
<tr>
<td>Facebook</td>
<td>Friends, existing conversations and opt-ins, from a personal profile</td>
<td>15 messages a day, 4 min apart</td>
</tr>
<tr>
<td>Email</td>
<td>Your connected email platform’s rules and your own lawful basis</td>
<td>Set in the email platform</td>
</tr>
</tbody>
</table>
<aside class="callout callout-note" aria-label="Defaults are ours, not the platforms'"><p class="callout-title">Defaults are ours, not the platforms&#39;</p><div class="callout-body"><p>The per-day numbers are GraphKit’s conservative defaults, not limits any platform has published. You can lower them, and raise some up to a fixed ceiling.</p></div></aside>
<h3 id="linkedin">LinkedIn</h3>
<p>LinkedIn is the one channel where reaching out to people you don’t know is part of normal use, within limits. Its <a href="https://www.linkedin.com/legal/professional-community-policies">Professional Community Policies</a> still say not to use invitations to send promotional messages to strangers, and its <a href="https://www.linkedin.com/legal/user-agreement">User Agreement</a> restricts automated tools. We cover pacing in detail in <a href="https://graphkit.com/blog/linkedin-invitation-limits/">LinkedIn invitation limits</a>.</p>
<h3 id="whatsapp-telegram-and-facebook">WhatsApp, Telegram and Facebook</h3>
<p>These are not cold channels. WhatsApp’s <a href="https://whatsappbusiness.com/policy/">Business Messaging Policy</a> says you may contact people only if they have given you their number and you have received opt-in permission confirming they want to hear from you, and that you must respect opt-out requests made “either on or off WhatsApp”. Telegram’s <a href="https://telegram.org/tos">Terms of Service</a> prohibit using the service to send spam, with temporary or permanent bans as the consequence.</p>
<p>GraphKit enforces this structurally. A WhatsApp, Telegram or Facebook step only runs for a lead with a recorded consent basis on that channel: an opt-in, an existing conversation, or an import you marked as collected with consent (and, on Facebook, an existing friendship). When someone you don’t know writes to you first, the lead is created with an existing-conversation basis automatically. Only admins can change a consent basis, and every change is audited. Without one, the step is skipped, or the lead stops if you configured the step that way.</p>
<p>Facebook campaigns run from personal profiles. Facebook Pages work in the <a href="https://graphkit.com/inbox/">unified inbox</a> but cannot start campaign conversations, because Pages can only answer people who messaged them recently.</p>
<h3 id="instagram">Instagram</h3>
<p>Instagram sits in between. Instagram’s <a href="https://www.facebook.com/help/instagram/581066165581870">Terms of Use</a> restrict automated access, and its official messaging API only lets businesses reply to people who messaged first. In GraphKit we position Instagram steps for people who follow you, engaged with your content, or were imported with consent, and we don’t recommend it for cold outreach.</p>
<h3 id="email">Email</h3>
<p>GraphKit does not send email itself. Email steps hand a lead to a campaign in your connected email platform (Warmbly, Instantly, Smartlead or lemlist), and that platform sends it under its own deliverability and unsubscribe handling. How the handoff and the stop rules work together is in <a href="https://graphkit.com/blog/add-email-to-linkedin-sequences/">adding email to LinkedIn sequences</a>.</p>
<h2 id="sequencing-escalate-on-signals-not-on-silence">Sequencing: escalate on signals, not on silence</h2>
<p>The spammy pattern is “no reply on channel A, so try channel B, then C”. Silence is an answer. A better rule is to add a channel only when something the person did makes it appropriate.</p>
<p>Patterns that hold up:</p>
<ul>
<li><strong>LinkedIn first, email second.</strong> Invite, message after acceptance, and hand the lead to an email campaign only if you have a lawful basis to email them and they haven’t replied on LinkedIn. Many teams already run this as two separate tools; running it as one sequence is what makes the stop rules work.</li>
<li><strong>Messaging apps for people already talking to you.</strong> A WhatsApp follow-up for contacts who opted in at an event or through a form. GraphKit ships a template for exactly this.</li>
<li><strong>Instagram after engagement.</strong> A message to someone who follows the account or engaged with a post, rather than to a list of handles.</li>
</ul>
<p>A pattern to avoid: the same message pasted into four channels on consecutive days. If the content is identical, the extra channels add pressure, not information.</p>
<h3 id="pace-the-person-not-just-the-account">Pace the person, not just the account</h3>
<p>Per-account limits protect accounts. They do nothing to protect the person on the other end from hearing from you on three channels in one afternoon. Our recommendation, not a platform rule: no more than one touch per business day across all channels for a given lead, and four to six touches in total over about three weeks before the sequence ends. Use wait steps generously; GraphKit’s can be set in hours, days or business days.</p>
<p>It also helps to keep one sender per person. GraphKit’s default rotation is sticky: once an account has messaged a lead on a channel, that lead stays with that account on that channel, so nobody gets the same opener from two different people at your company.</p>
<h2 id="what-stops-everything">What stops everything</h2>
<p>A multichannel sequence is only as polite as its stop rules. In GraphKit, four checks run before every step, including while a lead is sitting in a wait, in this order:</p>
<ol>
<li><strong>Suppression.</strong> The lead, one of their identities, their company domain or company name is on a suppression list for the workspace or the whole organisation.</li>
<li><strong>Opt-out keyword.</strong> The lead sent a message that starts with an opt-out word (such as “stop” or “unsubscribe”, plus equivalents in several languages), or a message of four words or fewer that contains one.</li>
<li><strong>Unsubscribe.</strong> The lead unsubscribed in the connected email platform.</li>
<li><strong>Reply.</strong> The lead replied on any channel, including by email through a connected platform.</li>
</ol>
<p>The same decision runs again with fresh data immediately before a message is sent, so a reply that arrives between scheduling and sending still prevents the next message. Stop on reply is on by default.</p>
<p>Opt-out keywords are always honoured on WhatsApp, Telegram and Facebook, and can be switched on for LinkedIn and Instagram. A match stops the lead and suppresses them on every channel in one step, and the inbox then refuses manual replies to that person too. Workspaces can add their own keywords but can never remove the defaults, so “stop” always works. We deliberately left out words like “end”, “cancel” and “quit”, which often begin ordinary replies. More on this in <a href="https://graphkit.com/blog/opt-outs-whatsapp-telegram-messenger/">opt-outs on WhatsApp, Telegram and Messenger</a>.</p>
<h2 id="identity-dont-guess-handles-or-numbers">Identity: don’t guess handles or numbers</h2>
<p>Before any step runs on a channel, GraphKit needs a verified identity for the lead on that channel. It does not guess a WhatsApp number from a name or match an Instagram handle by similarity. Unverified leads skip the step. That is slower than guessing, and it is the reason messages don’t land with the wrong person.</p>
<h2 id="a-checklist-before-you-launch">A checklist before you launch</h2>
<ul>
<li>For every channel in the sequence, can you say why this person expects to hear from you there?</li>
<li>Is there a wait of at least a business day between touches on different channels?</li>
<li>Does a reply anywhere stop everything, including email?</li>
<li>Are opt-out keywords on for every channel, not only the ones that require it?</li>
<li>Are customers, competitors and previous opt-outs on a suppression list? <a href="https://graphkit.com/leads/">Lists and exclusions</a> can be shared across workspaces in the same organisation.</li>
<li>Have you run it as a dry run and read the messages as the recipient would?</li>
</ul>
<p>If any answer is no, fix that before adding another channel. Replies from every channel should land in one place so nobody answers the same person twice; <a href="https://graphkit.com/blog/unified-inbox-for-outbound-replies/">a unified inbox for outbound replies</a> covers how to run that.</p>]]></content:encoded>
      <pubDate>Sun, 15 Nov 2026 14:00:00 GMT</pubDate>
      <dc:creator>GraphKit team</dc:creator>
      <category>Multichannel</category>
      <category>Consent</category>
      <category>Sequencing</category>
      <category>Stop on reply</category>
    </item>
    <item>
      <title>A unified inbox for outbound replies, and how to triage it</title>
      <link>https://graphkit.com/blog/unified-inbox-for-outbound-replies/</link>
      <guid isPermaLink="true">https://graphkit.com/blog/unified-inbox-for-outbound-replies/</guid>
      <description>Why replies from LinkedIn, WhatsApp, Telegram, Instagram and Facebook belong in one shared inbox, plus a triage routine for owners, labels and reply times.</description>
      <content:encoded><![CDATA[<p>Outbound across several channels has a predictable failure mode. The campaign side works: invitations go out, messages are sent, the sequence advances. Then the replies scatter. One is in a LinkedIn tab on a sales rep’s laptop, one is in a WhatsApp chat on someone’s phone, two are in Telegram, and the interested reply from Tuesday is still unread on Friday because it arrived on the account of the person who was off that week.</p>
<p>The fix is not more notifications. It is putting every reply in one shared place, with owners, labels and rules about who sees what. This post covers why that matters and a triage routine that works for small teams and agencies alike.</p>
<h2 id="why-replies-scatter">Why replies scatter</h2>
<p>Each platform assumes one person per account. LinkedIn is built around a person’s own inbox, WhatsApp around their phone, Telegram around their apps. That is fine for personal use and awkward for a team, because outbound replies are team work:</p>
<ul>
<li><strong>Accounts and people are not one to one.</strong> Five sales reps may share ten connected accounts. A reply belongs to the conversation, not to whoever’s phone it landed on.</li>
<li><strong>Coverage has gaps.</strong> Holidays, sick days and time zones mean the account owner is often not the best person to reply in the next hour.</li>
<li><strong>Context is spread out.</strong> Someone who replies on WhatsApp may have ignored three LinkedIn messages and opened an email. Without seeing all of that, the reply that comes back is generic.</li>
<li><strong>Managers can’t see the state of things.</strong> “How many people are waiting on us?” has no answer when the answer lives in a dozen apps.</li>
</ul>
<h2 id="what-a-shared-inbox-has-to-do">What a shared inbox has to do</h2>
<p>A unified inbox is more than a list of messages. For outbound, it needs to do a few specific things well.</p>
<p><strong>Show every channel in one list.</strong> GraphKit’s <a href="https://graphkit.com/inbox/">inbox</a> brings in conversations from LinkedIn, Instagram, WhatsApp, Telegram and Facebook, including each Facebook Page you manage. When an account is connected, it backfills the last 90 days (up to 200 messages per chat), then stays current through real-time updates from the platforms plus a catch-up every 15 minutes in case anything was missed.</p>
<p><strong>Reply from the right account.</strong> A reply goes out from the account the conversation is on, as the Page for Page conversations, so the other person sees a normal reply in their app. Sends carry an idempotency key, and a send whose outcome is unclear is never repeated, so a flaky connection doesn’t produce a double message.</p>
<p><strong>Keep the person’s whole history next to the conversation.</strong> The lead panel shows the person’s identities per channel, which lists and campaigns they are in, and a timeline of messages and campaign actions, including events from a connected email platform. That is the context you need before typing.</p>
<p><strong>Respect privacy.</strong> Not every message on a connected account is outbound. More on this below, because it is the part teams most often get wrong.</p>
<p><strong>Stay honest about what the platform reports.</strong> Telegram does not provide read receipts, so GraphKit doesn’t show them for Telegram. An inbox that invents a “seen” state leads people to the wrong conclusions.</p>
<h2 id="privacy-who-sees-what">Privacy: who sees what</h2>
<p>Connecting a personal LinkedIn account to a shared tool raises an obvious concern: can the whole team now read my private messages? It should not be able to, and in GraphKit it can’t by default.</p>
<p>Each connected account has a privacy setting:</p>
<table>
<thead>
<tr>
<th>Setting</th>
<th>What the team sees</th>
<th>Default for</th>
</tr>
</thead>
<tbody>
<tr>
<td>Share all</td>
<td>Every conversation on the account</td>
<td>Instagram, WhatsApp, Telegram, Facebook</td>
</tr>
<tr>
<td>Campaign only</td>
<td>Conversations that came from a campaign</td>
<td>LinkedIn</td>
</tr>
<tr>
<td>None</td>
<td>Nothing; the account only sends</td>
<td>(opt in)</td>
</tr>
</tbody>
</table>
<p>LinkedIn starts as campaign only because LinkedIn accounts belong to people. Conversations outside that scope are private: only roles with permission to read private conversations (client admins and agency roles) can open them, and for everyone else they do not exist. An admin can also limit a team member to specific connected accounts, which suits teams where each rep only handles their own accounts and a manager covers everyone.</p>
<p>Agree on these settings before connecting accounts, not after someone finds their personal messages in a shared view.</p>
<h2 id="a-triage-routine-that-works">A triage routine that works</h2>
<p>Tools help, but the routine is what makes replies get answered. This is the one we recommend.</p>
<h3 id="1-work-from-saved-views-not-the-full-list">1. Work from saved views, not the full list</h3>
<p>GraphKit has saved views for All, Unread, Assigned to me, Unassigned, Awaiting reply, Private and Archived. Most of the day should be spent in two of them:</p>
<ul>
<li><strong>Unassigned</strong>: new replies nobody owns yet. Someone sweeps this a few times a day and assigns or answers.</li>
<li><strong>Assigned to me</strong>: your queue.</li>
</ul>
<p><strong>Awaiting reply</strong> is the safety net: conversations where the other person spoke last. Anything that sits there too long is a dropped ball.</p>
<h3 id="2-give-every-conversation-an-owner">2. Give every conversation an owner</h3>
<p>Assign each reply to one person. The assignee must be able to see the conversation, so assignment never hands work to someone who can’t open it. Ownership ends the “I thought you had it” problem, and it gives managers a way to rebalance when one person’s queue grows.</p>
<h3 id="3-label-by-outcome-consistently">3. Label by outcome, consistently</h3>
<p>Labels are only useful if everyone applies them the same way. A small set is enough:</p>
<ul>
<li><strong>Interested</strong>: wants to continue the conversation.</li>
<li><strong>Meeting booked</strong>: the reply turned into a meeting.</li>
<li><strong>Not now</strong>: polite no, possibly later.</li>
<li><strong>Not a fit</strong>: wrong person or wrong company.</li>
<li><strong>Referral</strong>: pointed you to someone else.</li>
</ul>
<p>GraphKit’s <a href="https://graphkit.com/analytics/">analytics</a> counts conversations labelled Interested, Positive, Positive reply, Meeting booked, Meeting, Hot lead or Qualified as positive replies, so using one of those names for good outcomes makes reporting work without extra setup. Labelling is manual; GraphKit does not classify replies automatically today.</p>
<h3 id="4-set-a-response-time-target-and-look-at-it">4. Set a response-time target and look at it</h3>
<p>A reply on a messaging app expects a faster answer than an email. As a working target (ours, not a rule), answer interested replies within one business hour and everything else the same business day. The Awaiting reply view, sorted by oldest, shows how you are doing.</p>
<h3 id="5-write-private-notes-not-side-chats">5. Write private notes, not side chats</h3>
<p>Internal notes sit in the conversation and are never sent. Use them for “spoke to her on a call, follow up in March” or “this is the CFO, loop in Sam”. Notes in the conversation beat messages in a separate team chat because the next person to open it sees them.</p>
<h3 id="6-use-saved-replies-for-the-repetitive-parts-only">6. Use saved replies for the repetitive parts only</h3>
<p>Saved replies support the same template variables as campaigns (<code>first_name</code>, <code>company</code> and so on, with fallbacks). Use them for the parts of a reply that are always the same, such as a booking link or a pricing summary, and write the rest. A saved reply sent unchanged to an interested person reads like what it is.</p>
<h3 id="7-clean-up-in-bulk">7. Clean up in bulk</h3>
<p>Out-of-office replies, “thanks, not interested” and auto-responses pile up. Select up to 500 conversations at once to archive, mark read, label, assign or mute them. Marking a conversation read in GraphKit also marks it read on the network.</p>
<h3 id="keyboard-shortcuts">Keyboard shortcuts</h3>
<p>For anyone who triages daily: <code>j</code> and <code>k</code> move between conversations, <code>r</code> replies, <code>e</code> archives, <code>u</code> marks unread, <code>l</code> opens labels, <code>a</code> assigns, <code>/</code> searches, and Cmd or Ctrl+Enter sends. <code>?</code> shows the full list.</p>
<h2 id="when-a-reply-should-stop-the-campaign">When a reply should stop the campaign</h2>
<p>Every reply on any channel should stop the sequence for that person, and in GraphKit it does by default. The check runs before every step and again right before each send. That is what makes a shared inbox safe: the person answering doesn’t have to remember to pause anything. Opt-out messages such as “stop” go further, suppressing the person on every channel and blocking replies to them; <a href="https://graphkit.com/blog/opt-outs-whatsapp-telegram-messenger/">opt-outs on WhatsApp, Telegram and Messenger</a> explains how.</p>
<h2 id="what-we-havent-built-yet">What we haven’t built yet</h2>
<p>To set expectations: GraphKit does not draft replies with AI, does not tag replies automatically, has no mobile app yet, and does not send in-app notifications for new replies. If your team depends on any of those today, factor that in. We would rather you know before you connect accounts.</p>
<h2 id="the-short-version">The short version</h2>
<ul>
<li>Put every outbound reply in one place, with privacy settings agreed up front.</li>
<li>Sweep Unassigned, work from Assigned to me, and watch Awaiting reply.</li>
<li>One owner per conversation, a handful of outcome labels, notes in the thread.</li>
<li>Let replies stop campaigns automatically so nobody has to remember.</li>
</ul>
<p>For agencies running this across many clients, <a href="https://graphkit.com/blog/white-label-outbound-for-agencies/">white-label outbound for agencies</a> covers workspaces and roles.</p>]]></content:encoded>
      <pubDate>Sun, 15 Nov 2026 12:00:00 GMT</pubDate>
      <dc:creator>GraphKit team</dc:creator>
      <category>Inbox</category>
      <category>Triage</category>
      <category>Multichannel</category>
    </item>
    <item>
      <title>Opt-outs and consent on WhatsApp, Telegram and Messenger</title>
      <link>https://graphkit.com/blog/opt-outs-whatsapp-telegram-messenger/</link>
      <guid isPermaLink="true">https://graphkit.com/blog/opt-outs-whatsapp-telegram-messenger/</guid>
      <description>Practical guidance on consent and opt-outs for B2B messaging on WhatsApp, Telegram and Facebook Messenger: what platforms require and how to honour it.</description>
      <content:encoded><![CDATA[<p>Messaging apps feel personal because they are. A WhatsApp message lands next to someone’s family group, and an unwanted one gets blocked or reported far faster than an email. That is why WhatsApp, Telegram and Facebook Messenger are good channels for conversations people have agreed to, and poor channels for anything else.</p>
<p>This guide covers what each platform says about consent and opting out, what “consent” looks like in practice for B2B teams, and how to make sure an opt-out is honoured everywhere, not just in the chat where it was typed. It is practical guidance, not legal advice. Laws on electronic marketing differ by country, and sometimes depend on whether the person is an employee of a company or a sole trader; take advice for your own situation. The UK regulator’s guidance on <a href="https://ico.org.uk/for-organisations/direct-marketing-and-privacy-and-electronic-communications/business-to-business-marketing/">business-to-business marketing</a> is a useful starting point for UK senders.</p>
<h2 id="what-the-platforms-say">What the platforms say</h2>
<h3 id="whatsapp">WhatsApp</h3>
<p>WhatsApp’s <a href="https://whatsappbusiness.com/policy/">Business Messaging Policy</a> is the clearest of the three. You may contact people only if:</p>
<ul>
<li>they have given you their mobile phone number or username, and</li>
<li>you have received opt-in permission confirming they wish to receive subsequent messages from you.</li>
</ul>
<p>On opting out, it says you “must respect all requests (either on or off WhatsApp)” to block, discontinue or opt out of communications, including removing the person from your contacts list. It also says users can block or report businesses, and that WhatsApp may limit how many messages a business can send based on that feedback.</p>
<p>WhatsApp’s <a href="https://www.whatsapp.com/legal/terms-of-service">Terms of Service</a> separately prohibit bulk messaging and auto-messaging. Put plainly: WhatsApp is not a cold outreach channel, and treating it as one puts the account at risk regardless of what any law allows.</p>
<h3 id="telegram">Telegram</h3>
<p>Telegram’s <a href="https://telegram.org/tos">Terms of Service</a> ask users to agree not to “use our service to send spam or scam users”, and state that breaking the terms may result in a temporary or permanent ban. Telegram’s <a href="https://core.telegram.org/api/terms">API terms</a> also prohibit actions on a user’s behalf without their knowledge and consent. There is less detail than WhatsApp’s policy, but the direction is the same: message people who expect to hear from you.</p>
<h3 id="facebook-messenger">Facebook Messenger</h3>
<p>Meta’s <a href="https://developers.facebook.com/docs/messenger-platform/policy/policy-overview">Messenger Platform policy</a> governs businesses using its official messaging APIs. It limits when a business can message someone to a window (24 hours for most actions) opened by the person, and notes that users can block or mute a conversation with a business at any time. GraphKit sends Facebook messages from personal profiles rather than through that API, and Facebook Pages are inbox-only in GraphKit for the same reason: Pages can only answer people who wrote recently. The principle carries over either way. Message friends, people already in a conversation with you, and people who asked to hear from you.</p>
<h2 id="what-consent-looks-like-in-practice">What consent looks like in practice</h2>
<p>For B2B teams, consent on these channels usually comes from one of three places:</p>
<ol>
<li><strong>They started the conversation.</strong> Someone messages your WhatsApp number from your website, or writes to you on Telegram after a talk.</li>
<li><strong>They opted in somewhere you can point to.</strong> A form that asked for a WhatsApp number and said what you would send, a sign-up sheet at an event, a checkbox in a trial flow.</li>
<li><strong>They are an existing contact who agreed.</strong> A customer who asked to be updated on WhatsApp, a partner who prefers Telegram.</li>
</ol>
<p>What does not count: a number found on a website, a number from a data vendor, or a phone number on someone’s LinkedIn profile. Having a number is not the same as having permission to use it for messages.</p>
<p>For each opted-in contact, keep a record of where the consent came from, when, and what the person was told. If a contact ever asks why you messaged them, you want to answer in one sentence.</p>
<h3 id="how-graphkit-records-consent">How GraphKit records consent</h3>
<p>Every lead identity on WhatsApp, Telegram and Facebook carries a <strong>consent basis</strong>: opt-in, existing conversation, imported with consent, or (on Facebook) friend. Campaign steps on those channels only run when one is present; without it, the step is skipped or the lead stops, depending on how you set the step up.</p>
<ul>
<li>When someone you have never spoken to writes to you, GraphKit creates the lead with an existing-conversation basis automatically.</li>
<li>When you import a CSV, you can mark the contacts as collected with consent, per channel. Do this only for lists where it is true.</li>
<li>Only workspace admins can change a consent basis, and every change is written to the audit log.</li>
</ul>
<p>When you connect a WhatsApp or Telegram account, GraphKit also asks you to confirm that you will only message people who gave you their number or username and agreed to hear from you, and that you won’t send bulk or cold messages. It is a checkbox, but it is worth reading.</p>
<h2 id="honouring-opt-outs-everywhere">Honouring opt-outs everywhere</h2>
<p>An opt-out that only works in one chat is not an opt-out. If someone says “stop” on WhatsApp, they mean stop, not “switch to email”.</p>
<h3 id="keywords">Keywords</h3>
<p>GraphKit watches every inbound message on WhatsApp, Telegram and Facebook for opt-out keywords, whether or not the person is in a campaign. The defaults include “stop”, “stopall”, “unsubscribe”, “opt out” and “optout”, plus equivalents in German, the Nordic languages, French, Spanish, Portuguese, Italian, Dutch and Russian. A message matches if it starts with a keyword, or if it is four words or fewer and contains one. So “STOP” and “please stop” match; a long message that happens to mention “stop” halfway through does not.</p>
<p>Some words are left out on purpose. “End”, “cancel” and “quit” often start ordinary replies (“Cancel that, Thursday works better”), and treating them as opt-outs would silence real conversations. Workspaces can add their own keywords, but can never remove the defaults.</p>
<p>On LinkedIn and Instagram, keyword detection is off by default and can be switched on. We recommend switching it on.</p>
<h3 id="what-happens-on-a-match">What happens on a match</h3>
<p>In one step:</p>
<ul>
<li>the lead’s campaigns stop,</li>
<li>the lead is suppressed on <strong>every</strong> channel, including email handoffs,</li>
<li>the conversation is marked as opted out, and the inbox and API refuse further replies to that person.</li>
</ul>
<p>That last part matters more than it sounds. Most accidental “one more message” incidents are manual, from a teammate who did not see the opt-out. Blocking the composer removes that possibility.</p>
<h3 id="requests-that-arrive-somewhere-else">Requests that arrive somewhere else</h3>
<p>People also opt out by replying to an email, telling a colleague, or calling. WhatsApp’s policy explicitly covers requests made off WhatsApp. Treat these the same way: add the person to a suppression list.</p>
<p>GraphKit’s suppression lists work at workspace or organisation level, and an entry can match a single identity, a lead (all of their identities), a company domain or a company name. Suppression is checked at import and before every campaign step, including while a lead waits between steps. Identity-level entries survive deleting the lead, so re-importing the same person later does not quietly bring them back. Suppression lists are never shared between different customers on GraphKit.</p>
<h2 id="making-it-easy-to-say-no">Making it easy to say no</h2>
<p>The platforms expect you to honour opt-outs; good practice is to make them easy to find in the first place.</p>
<ul>
<li><strong>Say who you are and why you are writing</strong> in the first message. An opted-in contact may not remember the form they filled in three weeks ago.</li>
<li><strong>Mention how to stop</strong>, especially for anything recurring. “Reply STOP and I won’t message you here again” is enough.</li>
<li><strong>Don’t argue with a no.</strong> No “just one last thing”, no follow-up asking why.</li>
<li><strong>Don’t move channels after an opt-out.</strong> If you also have their email, the opt-out still applies unless they say otherwise.</li>
<li><strong>Keep volume low.</strong> GraphKit’s defaults are 20 messages a day per WhatsApp or Telegram account and 15 per Facebook account, with minimum gaps of five and four minutes. These are our recommendations, not platform figures. Messaging apps are for conversations, and conversations don’t scale like campaigns.</li>
</ul>
<h2 id="a-checklist">A checklist</h2>
<ul>
<li>Every WhatsApp, Telegram and Facebook contact has a consent source you could explain in one sentence.</li>
<li>Imported lists are only marked as consented when they were.</li>
<li>Opt-out keywords are on for every channel you use.</li>
<li>Off-platform opt-outs go onto a suppression list the same day.</li>
<li>Existing customers and previous opt-outs are suppressed at organisation level, so new workspaces inherit them.</li>
<li>Nobody on the team can reply to an opted-out conversation by mistake.</li>
</ul>
<p>For the sequencing side of the same problem (which channel, in what order, and what stops it), read <a href="https://graphkit.com/blog/multichannel-outbound-without-spamming/">multichannel outbound without spamming</a>. For how each channel works in GraphKit, see the <a href="https://graphkit.com/channels/#whatsapp">channels page</a>. GraphKit is not affiliated with, endorsed by or approved by Meta, WhatsApp or Telegram.</p>]]></content:encoded>
      <pubDate>Sun, 15 Nov 2026 11:00:00 GMT</pubDate>
      <dc:creator>GraphKit team</dc:creator>
      <category>Multichannel</category>
      <category>Consent</category>
      <category>Opt-outs</category>
      <category>WhatsApp</category>
      <category>Telegram</category>
      <category>Facebook Messenger</category>
    </item>
    <item>
      <title>Adding email to LinkedIn sequences with Warmbly and others</title>
      <link>https://graphkit.com/blog/add-email-to-linkedin-sequences/</link>
      <guid isPermaLink="true">https://graphkit.com/blog/add-email-to-linkedin-sequences/</guid>
      <description>How to add email to LinkedIn sequences through Warmbly, Instantly, Smartlead or lemlist: handoffs, field mapping, stop rules and reply and unsubscribe sync.</description>
      <content:encoded><![CDATA[<p>Many outbound teams already run LinkedIn and email side by side, usually in two tools that don’t talk to each other. The result is familiar: someone accepts your LinkedIn invitation and replies there, and two days later your email platform sends them “just bumping this to the top of your inbox”. Each tool did its job. Together they spammed someone who had already answered.</p>
<p>Adding email to a LinkedIn sequence is mostly about fixing that: one sequence decides when email starts, and a reply or unsubscribe anywhere stops everything. This post covers how that works with an email platform you connect to GraphKit, with Warmbly as the featured option and Instantly, Smartlead and lemlist also supported.</p>
<aside class="callout callout-note" aria-label="Available now"><p class="callout-title">Available now</p><div class="callout-body"><p>Connect Warmbly, Instantly, Smartlead or lemlist from Integrations → Email in GraphKit. Platform differences, such as which ones support pausing a lead, are on the <a href="https://graphkit.com/integrations/#email">integrations page</a>.</p></div></aside>
<h2 id="graphkit-doesnt-send-email-on-purpose">GraphKit doesn’t send email, on purpose</h2>
<p>Email deliverability is its own discipline: domains, authentication, warm-up, inbox rotation, bounce handling, unsubscribe links. Dedicated email platforms do this well, and most teams already have one. So GraphKit never sends email itself. Instead, an <strong>email handoff</strong> step adds the lead to a campaign in your email platform, and that platform sends the emails on its own schedule, with its own unsubscribe handling.</p>
<p>What GraphKit adds is coordination:</p>
<ul>
<li><strong>A reply on email stops the lead on every channel</strong>, under the same stop-on-reply rule as LinkedIn and the messaging channels.</li>
<li><strong>An unsubscribe in the email platform suppresses the lead on every channel</strong>, not just email.</li>
<li><strong>A reply on LinkedIn (or any social channel) pauses the lead in the email platform</strong>, so the email sequence doesn’t carry on regardless.</li>
<li><strong>A suppressed lead is removed from the email campaign.</strong></li>
</ul>
<p>Those last two are settings that are on by default, and a sync job applies them every minute.</p>
<h2 id="the-two-email-steps">The two email steps</h2>
<h3 id="email-handoff">Email handoff</h3>
<p>The handoff step names a connection (your email platform account) and one of its campaigns, and maps fields from the GraphKit lead into the platform’s fields. Mapping uses the same template syntax as messages, including fallbacks:</p>
<pre class="astro-code github-dark" style="background-color:#24292e;color:#e1e4e8;overflow-x:auto" tabindex="0" data-language="plaintext"><code><span class="line"><span>first_name    → {{first_name}}</span></span>
<span class="line"><span>company_name  → {{company}}</span></span>
<span class="line"><span>job_title     → {{job_title}}</span></span>
<span class="line"><span>industry      → {{industry|SaaS}}</span></span></code></pre>
<p>The email address is always sent. If you map nothing, GraphKit sends first name, last name, company and job title. You can map up to 40 fields, including custom ones your email templates use.</p>
<p>The handoff needs an email address that has not bounced. If the lead has none, the step is skipped by default, or the lead stops if you configured the step to stop. Email steps don’t use LinkedIn’s daily caps or the campaign’s sending window, because the email platform runs its own schedule; they do respect the platform’s rate limits and wait when it asks them to.</p>
<h3 id="email-stop">Email stop</h3>
<p>The stop step pauses or removes the lead in the email platform, for one campaign or for every campaign the lead was handed to on that connection. Remove is the default. Use it on paths where email should end, for example after a “connected” branch where you’d rather continue on LinkedIn.</p>
<h2 id="a-sequence-that-combines-the-two">A sequence that combines the two</h2>
<p>A pattern that works for many B2B teams:</p>
<ol>
<li><strong>LinkedIn invite</strong>, with a short note or none. See <a href="https://graphkit.com/blog/linkedin-connection-notes-that-get-replies/">connection notes and follow-ups</a>.</li>
<li><strong>Wait</strong> up to a week, then <strong>condition: connected?</strong></li>
<li><strong>Yes:</strong> LinkedIn message one business day later, a second message after four business days, then stop.</li>
<li><strong>No:</strong> <strong>email handoff</strong> to a short email campaign (two or three emails), only for leads you have a lawful basis to email.</li>
<li>Any reply, on LinkedIn or email, stops the lead everywhere.</li>
</ol>
<p>The point of the branch is that each person hears from you on one main channel at a time. Someone who connected gets LinkedIn messages; someone who didn’t gets email. Running email in parallel for everyone is possible, but it doubles the touches for the people who engaged, which is the opposite of what you want. Our broader view on this is in <a href="https://graphkit.com/blog/multichannel-outbound-without-spamming/">multichannel outbound without spamming</a>.</p>
<p>Before launching, run the campaign as a dry run. GraphKit makes the same decisions without sending anything, which shows you which leads would reach the handoff and what would be mapped.</p>
<h2 id="choosing-an-email-platform">Choosing an email platform</h2>
<p>All four platforms support handing a lead to a campaign, removing it, and sending reply, unsubscribe and bounce events back to GraphKit. They differ in the details.</p>
<table>
<thead>
<tr>
<th></th>
<th>Warmbly</th>
<th>Instantly</th>
<th>Smartlead</th>
<th>lemlist</th>
</tr>
</thead>
<tbody>
<tr>
<td>Connect with</td>
<td>API key, or OAuth sign-in (Cloud)</td>
<td>API v2 key</td>
<td>API key (Pro plan or higher)</td>
<td>API key</td>
</tr>
<tr>
<td>Self-hosted address</td>
<td>Yes</td>
<td>No</td>
<td>No</td>
<td>No</td>
</tr>
<tr>
<td>Pause a lead</td>
<td>Yes</td>
<td>No, falls back to remove</td>
<td>Yes</td>
<td>Yes</td>
</tr>
<tr>
<td>Remove a lead</td>
<td>Yes</td>
<td>Yes</td>
<td>Yes</td>
<td>Yes</td>
</tr>
<tr>
<td>Reply, unsubscribe, bounce events</td>
<td>Yes</td>
<td>Yes</td>
<td>Yes</td>
<td>Yes</td>
</tr>
<tr>
<td>Signed webhook deliveries</td>
<td>Yes (HMAC)</td>
<td>No, secret URL token</td>
<td>Optional signing</td>
<td>No, secret URL token</td>
</tr>
<tr>
<td>Safe automatic retry of a handoff</td>
<td>Yes</td>
<td>Yes</td>
<td>Reconciled instead</td>
<td>Reconciled instead</td>
</tr>
</tbody>
</table>
<p>Where a platform doesn’t support safe retries, GraphKit doesn’t retry a handoff whose outcome is unclear. It looks the lead up instead where the platform allows it, and otherwise marks the step failed rather than risking a lead being added twice.</p>
<h3 id="warmbly">Warmbly</h3>
<p><a href="https://warmbly.com">Warmbly</a> is our sister product: open-source email outreach from the same company as GraphKit, Mindroot Ltd. You can use Warmbly Cloud or run it yourself; the code is on <a href="https://github.com/warmbly/warmbly">GitHub</a>. It is the most complete fit for GraphKit’s handoff:</p>
<ul>
<li><strong>Sign in with Warmbly</strong> (OAuth) on Cloud, so nobody copies API keys around. Self-hosted installations connect with an API key.</li>
<li><strong>A self-hosted base URL</strong>, so the integration works against your own Warmbly server. Plain http is only accepted for local addresses.</li>
<li><strong>Signed webhooks</strong>, so reply and unsubscribe events can be verified as genuinely coming from Warmbly.</li>
<li><strong>Pause and remove</strong>, plus a lead lookup, so an unclear handoff can be checked rather than repeated.</li>
</ul>
<p>If you are choosing an email platform from scratch and care about owning your sending infrastructure, Warmbly is worth a look. If you already run Instantly, Smartlead or lemlist, keep it; the handoff works with all of them.</p>
<h3 id="instantly-smartlead-and-lemlist">Instantly, Smartlead and lemlist</h3>
<ul>
<li><strong>Instantly</strong> needs an API v2 key (v1 keys don’t work). Instantly’s API has no lead pause, so “pause” falls back to removing the lead from the campaign. Its webhooks are unsigned; GraphKit authenticates them with a secret URL and header.</li>
<li><strong>Smartlead</strong> needs an API key from a Pro plan or higher. Webhook signing is optional and can be turned on with a signing secret.</li>
<li><strong>lemlist</strong> supports pause and remove. Its webhooks are unsigned and authenticated with a secret, and each webhook receives every event type.</li>
</ul>
<h2 id="what-none-of-the-platforms-tell-us">What none of the platforms tell us</h2>
<p>One gap applies to all four: none of them documents an event for “this lead finished the email sequence”. Campaign-level completion events exist, but not per lead. In practice, that means GraphKit can’t automatically resume LinkedIn steps the moment a lead’s emails run out. Plan sequences so that email is the last stage for a lead, or add a fixed wait after the handoff that matches the length of the email campaign.</p>
<h2 id="reporting-without-double-counting">Reporting without double counting</h2>
<p>Email handoffs and email replies are reported separately from social replies in GraphKit’s <a href="https://graphkit.com/analytics/">analytics</a>, and email replies are attributed to the GraphKit campaign that handed the lead off. A lead who replies on LinkedIn and by email isn’t counted twice as two “replies”. Your email platform remains the place for opens, clicks and deliverability metrics.</p>
<p>For teams that build their own automations, GraphKit also emits outbound webhooks for <code>email.replied</code>, <code>email.bounced</code> and <code>email.unsubscribed</code>, signed and retried. See <a href="https://graphkit.com/developers/">API and webhooks</a>.</p>
<h2 id="checklist">Checklist</h2>
<ul>
<li>Decide which leads get email, and confirm you have a lawful basis to email them.</li>
<li>Use a branch so each lead hears from you on one main channel at a time.</li>
<li>Map the fields your email templates use, with fallbacks for anything optional.</li>
<li>Leave “pause on social reply” and “remove on suppression” switched on.</li>
<li>Check that the email platform’s own unsubscribe handling is set up; GraphKit relies on it.</li>
<li>Dry-run the sequence before it touches a real inbox.</li>
</ul>]]></content:encoded>
      <pubDate>Sun, 15 Nov 2026 10:00:00 GMT</pubDate>
      <dc:creator>GraphKit team</dc:creator>
      <category>Email</category>
      <category>Warmbly</category>
      <category>Sequencing</category>
      <category>Stop on reply</category>
      <category>LinkedIn</category>
    </item>
    <item>
      <title>How agencies run white-label outbound for clients</title>
      <link>https://graphkit.com/blog/white-label-outbound-for-agencies/</link>
      <guid isPermaLink="true">https://graphkit.com/blog/white-label-outbound-for-agencies/</guid>
      <description>How outbound agencies structure client work: a workspace per client, pooled connected accounts, white-label branding, custom domains, logins and reports.</description>
      <content:encoded><![CDATA[<p>Running outbound for one company is a campaign problem. Running it for fifteen clients at once is an operations problem: keeping their data apart, keeping their accounts healthy, letting them see what is happening without seeing each other, and presenting all of it under your agency’s name rather than a vendor’s.</p>
<p>This post walks through how agencies structure that work: workspaces, seats, roles, branding, domains and reports. It uses GraphKit’s <a href="https://graphkit.com/agencies/">agency features</a> as the worked example, and is explicit about what is and isn’t built today.</p>
<h2 id="one-workspace-per-client">One workspace per client</h2>
<p>The basic unit is a workspace per client. Everything that belongs to a client lives inside it: their connected accounts, leads, lists, campaigns, inbox and analytics. Nothing crosses between workspaces unless you deliberately put it at organisation level.</p>
<p>In GraphKit an agency organisation can create as many client workspaces as it needs. Workspaces are archived, never hard-deleted, so ending an engagement doesn’t destroy the history you may need for a dispute or a final report. Data from another workspace isn’t just hidden; requests for it return “not found”, the same as if it didn’t exist.</p>
<p>A few habits make workspaces easier to run:</p>
<ul>
<li><strong>Name them for the client, not the project.</strong> Projects change; the client relationship is what you report on.</li>
<li><strong>Set the time zone per client.</strong> Working hours and quota days follow each connected account’s time zone, so a client in Sydney and one in London both send during their own business day.</li>
<li><strong>Keep shared assets at organisation level.</strong> Exclusion lists can live at organisation scope, so “never contact our own clients” or “never contact this competitor” applies to every workspace at once.</li>
</ul>
<h2 id="pooled-seats">Pooled seats</h2>
<p>Clients grow and shrink. An agency that has to buy seats per client ends up with idle accounts in one workspace and none to spare in another.</p>
<p>GraphKit sells connected accounts once per organisation and lets you allocate them to workspaces. The organisation’s total sets the volume tier, so every account is billed at the same per-account price, and the allocations can never add up to more than you bought. An account holds a seat while it is connecting, active, paused, restricted or waiting to reconnect; disconnecting or removing it frees the seat. Teammates and client users never use seats.</p>
<p>At current <a href="https://graphkit.com/pricing/">pricing</a>, an agency with 30 connected accounts across four clients is in the 25–49 tier at $35 per account per month, so $1,050 a month, billed monthly, whichever clients the 30 accounts are allocated to. Moving one client’s accounts up or down changes only the allocation, not the price per account, until the total crosses a tier boundary.</p>
<h2 id="roles-that-match-how-agencies-work">Roles that match how agencies work</h2>
<p>Agencies need two kinds of access: their own staff across every client, and each client’s people inside their own workspace only. GraphKit has fixed roles for this:</p>
<table>
<thead>
<tr>
<th>Role</th>
<th>Who it’s for</th>
<th>Scope</th>
</tr>
</thead>
<tbody>
<tr>
<td>Agency owner</td>
<td>Agency leadership, billing</td>
<td>All workspaces</td>
</tr>
<tr>
<td>Agency manager</td>
<td>Account managers</td>
<td>All workspaces</td>
</tr>
<tr>
<td>Client admin</td>
<td>The client’s lead contact</td>
<td>Their workspace</td>
</tr>
<tr>
<td>Campaign operator</td>
<td>Whoever builds and runs campaigns</td>
<td>Their workspace</td>
</tr>
<tr>
<td>Inbox agent</td>
<td>Whoever answers replies</td>
<td>Their workspace</td>
</tr>
<tr>
<td>Viewer</td>
<td>Stakeholders who only need to look</td>
<td>Their workspace</td>
</tr>
</tbody>
</table>
<p>Roles are fixed rather than built from per-screen toggles. That is less flexible, and it is deliberate: a fixed matrix is easy to audit and hard to misconfigure. Someone with too little access gets “forbidden”; data from another client’s workspace returns “not found”.</p>
<h3 id="client-logins">Client logins</h3>
<p>Clients sign in to the same application your team uses, on your domain once white-label is set up, with a role in their own workspace. There is no separate client portal to maintain. Invitations are links you copy and send yourself, so they come from you, not from a vendor’s email address.</p>
<h3 id="who-connects-which-account">Who connects which account</h3>
<p>This is the most important operational rule in agency outbound: <strong>the person who owns an account connects it.</strong> A client’s LinkedIn, WhatsApp or Instagram account should be connected by that client, signed in to their own workspace. Agency staff should never ask for, store or type in a client’s password.</p>
<p>GraphKit’s account connection uses a hosted sign-in run by our connection provider under its own name. Before it opens, GraphKit shows a notice saying so. That screen stays under the provider’s name even under white-label, because hiding who operates a sign-in page is exactly what people are trained not to trust. Expect clients to ask about it, and explain it during onboarding.</p>
<h2 id="white-label-branding">White-label branding</h2>
<p>From 10 purchased connected accounts, GraphKit unlocks white-label. White-label changes how the product looks; it doesn’t unlock features, because nothing else is gated. Every feature, including integrations, works the same either way.</p>
<p>What you can set:</p>
<ul>
<li>product name, description, logo and favicon</li>
<li>accent colour and font</li>
<li>support URL and email, terms and privacy URLs</li>
<li>label overrides, for example renaming “Campaigns” to “Sequences”</li>
<li>hiding GraphKit branding</li>
</ul>
<p><strong>One brand per organisation, today.</strong> Your brand shows on every domain you add, but GraphKit does not yet support several different brands inside one organisation. An agency that sells under two names needs two organisations, and their seats don’t pool. Multiple brands per organisation is on our roadmap. If you run more than one brand, plan for it now.</p>
<p>If an organisation drops below 10 accounts, a warning appears straight away. After 14 days, hidden GraphKit branding returns as a small “Powered by GraphKit” footer, and verified domains keep working. That grace period exists so a client churning at the end of a month doesn’t break the experience for every other client overnight.</p>
<h2 id="custom-domains">Custom domains</h2>
<p>Clients should log in at something like <code>outbound.youragency.com</code>, not at a vendor’s address. In GraphKit, adding a domain takes two DNS records: a TXT record to prove you control the domain and a CNAME pointing it at GraphKit. Certificates are issued and renewed automatically, and pending domains are checked automatically until the records resolve.</p>
<p>Use a subdomain rather than your root domain; a CNAME on the root conflicts with your website and email records. Each hostname belongs to one organisation, and you can point a domain at a default workspace so a client who signs in there lands in the right place.</p>
<p>One security detail worth knowing: the domain decides which brand someone sees, never what they can access. Access always comes from the signed-in session and its role.</p>
<h2 id="reporting-to-clients">Reporting to clients</h2>
<p>Clients pay for outcomes and want to see them without logging in to a tool they didn’t choose. Two pieces cover this.</p>
<p><strong>Agency analytics.</strong> Your team sees totals across every active client workspace plus a per-client table, and each workspace’s report can be exported as CSV grouped by day, channel, account or campaign. It is the view you use to spot a client whose acceptance is sliding before they notice.</p>
<p><strong>Shareable report links.</strong> For each client you can create a read-only report link that shows aggregate metrics for their workspace, branded with your white-label settings when those are on. The link needs no login, shows no messages, lead names or account errors, and can be revoked at any time. Only roles that manage the team can create them.</p>
<p>Positive replies in those reports come from labels your team applies in the inbox (such as Interested or Meeting booked), so agree on labels with each client up front. <a href="https://graphkit.com/blog/unified-inbox-for-outbound-replies/">A unified inbox for outbound replies</a> covers a labelling scheme that works.</p>
<h2 id="billing-your-clients">Billing your clients</h2>
<p>GraphKit bills your agency: one subscription per organisation, monthly, priced per connected account, excluding tax. It doesn’t invoice your clients on your behalf. Reseller billing is designed but not built, so you keep using whatever invoicing you have. Most agencies price outbound as a retainer that includes accounts, which pooled seats make easy to cost.</p>
<h2 id="onboarding-and-offboarding">Onboarding and offboarding</h2>
<p>A repeatable checklist saves most of the problems agencies run into.</p>
<p><strong>Onboarding a client</strong></p>
<ol>
<li>Create the workspace and set its time zone.</li>
<li>Allocate seats for the accounts the client will connect.</li>
<li>Invite the client admin with a copied link; they connect their own accounts.</li>
<li>Agree inbox privacy per account (LinkedIn defaults to campaign-only).</li>
<li>Import the client’s existing customers and competitors into a suppression list.</li>
<li>Run the first campaign as a dry run and review the messages with the client.</li>
<li>Create a report link and agree how often you’ll review it together.</li>
</ol>
<p><strong>Offboarding a client</strong></p>
<ol>
<li>Stop campaigns and export what the client is entitled to keep (leads export as CSV, analytics as CSV).</li>
<li>Revoke report links and client logins.</li>
<li>Disconnect accounts, which frees their seats.</li>
<li>Archive the workspace.</li>
</ol>
<h2 id="what-to-check-in-any-agency-platform">What to check in any agency platform</h2>
<p>If you are comparing tools, these questions separate agency-ready platforms from single-team tools with a workspace switcher: Are seats pooled across clients? Can clients connect their own accounts? Is white-label self-serve, including domains and certificates? Are client reports shareable without a login? Can inbox access be scoped per account? Do all integrations still work under white-label? We compare one popular option in <a href="https://graphkit.com/blog/heyreach-alternatives/">HeyReach alternatives in 2026</a>, including where it is ahead of us.</p>]]></content:encoded>
      <pubDate>Sun, 15 Nov 2026 09:00:00 GMT</pubDate>
      <dc:creator>GraphKit team</dc:creator>
      <category>Agencies</category>
      <category>White-label</category>
      <category>Workspaces</category>
      <category>Reporting</category>
    </item>
    <item>
      <title>HeyReach alternatives in 2026: an honest comparison</title>
      <link>https://graphkit.com/blog/heyreach-alternatives/</link>
      <guid isPermaLink="true">https://graphkit.com/blog/heyreach-alternatives/</guid>
      <description>An honest look at HeyReach alternatives in 2026, from published pricing and help pages: where HeyReach is strong, where it falls short, and where we are weaker.</description>
      <content:encoded><![CDATA[<p>We build GraphKit, which competes with HeyReach, so read this with that in mind. To keep it useful anyway, every statement about HeyReach below comes from its own pricing page and help centre as checked on 15 November 2026, and we include the places where HeyReach is ahead of us or cheaper. If something here is out of date, <a href="https://graphkit.com/contact/">tell us</a> and we will correct it.</p>
<p>HeyReach is a trademark of its owner. GraphKit is not affiliated with HeyReach.</p>
<h2 id="what-heyreach-does-well">What HeyReach does well</h2>
<p>HeyReach is a mature LinkedIn outbound tool for teams and agencies, and several parts of it are genuinely strong:</p>
<ul>
<li><strong>A capable LinkedIn inbox.</strong> One inbox across all senders, with filters by sender, campaign and tag, bulk actions, and an automatic sentiment tag (Interested, Not Interested or Generic) applied 15 minutes after a lead’s first reply. Agencies with several workspaces get a “Master View” across all of them.</li>
<li><strong>Inbox privacy per LinkedIn account.</strong> A “Track only HeyReach-started” setting keeps personal LinkedIn conversations out of the shared inbox.</li>
<li><strong>Voice notes</strong>, including transcripts of incoming voice notes.</li>
<li><strong>Notifications</strong> in the app and by email for events such as new replies.</li>
<li><strong>A broad integration surface.</strong> A public API with 88 endpoints at 300 requests a minute, a free MCP server, a CLI, Clay, a native (beta, one-way) HubSpot integration, and two-way integrations with Instantly, Smartlead and EmailBison that pause the other channel when a lead replies.</li>
<li><strong>Agency features.</strong> Up to 30 workspaces, per-workspace seat limits, moving senders between workspaces, granular per-screen permissions, and a Manager role for people from outside your organisation.</li>
<li><strong>Unlimited users at no cost</strong>, and a 14-day free trial with three LinkedIn accounts and no card.</li>
</ul>
<h2 id="where-heyreach-falls-short">Where HeyReach falls short</h2>
<p>These gaps come from HeyReach’s own documentation, either because a feature is listed as a roadmap idea or because it is not documented anywhere:</p>
<ul>
<li><strong>No inbox assignment or internal notes.</strong> There is no conversation owner or “assigned to me” view, and notes are a roadmap idea.</li>
<li><strong>The inbox cannot start new conversations.</strong> HeyReach’s help centre says you cannot send direct messages to leads that haven’t already been contacted from your LinkedIn inbox or a HeyReach campaign.</li>
<li><strong>No native AI reply drafting</strong> (only through MCP or workflows you build), and auto-tags cannot be edited.</li>
<li><strong>No mobile app</strong> is documented.</li>
<li><strong>White-label is not self-serve.</strong> It is set up by HeyReach rather than by you, no DNS or certificate steps are documented, and Instantly, Smartlead and credit purchases are unavailable to white-label users.</li>
<li><strong>No client billing and no shareable client reports</strong> are documented.</li>
<li><strong>Webhook delivery details</strong> (retries, signing, timeouts) are not documented.</li>
<li><strong>LinkedIn only.</strong> Beyond LinkedIn, email comes through integrations; Instagram, WhatsApp, Telegram and Facebook are not published.</li>
</ul>
<p>It is also worth knowing that HeyReach is candid about platform risk. Its blog states, “There is no approved outreach automation tool — ours included.” That is true of every tool in this category, GraphKit too.</p>
<h2 id="pricing-side-by-side">Pricing, side by side</h2>
<p>HeyReach prices per LinkedIn sender, with plan bundles for agencies. Monthly list prices from its <a href="https://www.heyreach.io/pricing">pricing page</a>:</p>
<ul>
<li><strong>Growth:</strong> $79 per sender for 1–9 senders, $59 per sender for 10 or more, up to 50 senders. Includes a dedicated residential proxy per sender. No white-label.</li>
<li><strong>Agency:</strong> $999 a month for 25 senders or $1,399 for 50. One white-label brand. <strong>Bring your own proxies.</strong></li>
<li><strong>Unlimited:</strong> $2,999 a month, with a fair-use cap of 300 senders. Bring your own proxies. Additional white-label brands are $500 per add-on.</li>
</ul>
<p>HeyReach also offers quarterly and yearly billing at lower effective prices. GraphKit bills monthly only, per connected account on any channel, with volume tiers from $49 down to $19 per account (see <a href="https://graphkit.com/pricing/">pricing</a>). Here is the cheapest published HeyReach option against GraphKit at monthly prices:</p>
<table>
<thead>
<tr>
<th>Accounts</th>
<th>GraphKit</th>
<th>Cheapest HeyReach option</th>
<th>Difference</th>
</tr>
</thead>
<tbody>
<tr>
<td>5</td>
<td>$245</td>
<td>$395 (Growth)</td>
<td>GraphKit $150 less</td>
</tr>
<tr>
<td>10</td>
<td>$420</td>
<td>$590 (Growth)</td>
<td>GraphKit $170 less</td>
</tr>
<tr>
<td>25</td>
<td>$875</td>
<td>$999 (Agency 25, own proxies)</td>
<td>GraphKit $124 less</td>
</tr>
<tr>
<td>50</td>
<td>$1,450</td>
<td>$1,399 (Agency 50, own proxies)</td>
<td><strong>HeyReach $51 less</strong></td>
</tr>
<tr>
<td>100</td>
<td>$2,400</td>
<td>$2,999 (Unlimited, own proxies)</td>
<td>GraphKit $599 less</td>
</tr>
<tr>
<td>300</td>
<td>$5,700</td>
<td>$2,999 (Unlimited, own proxies)</td>
<td><strong>HeyReach $2,701 less</strong></td>
</tr>
</tbody>
</table>
<p>Two caveats make this less tidy than a table suggests. First, from 25 accounts upward the cheapest HeyReach option requires your own proxies, which is a real cost in money and time that the table leaves out; GraphKit’s price includes the connection infrastructure for LinkedIn, Instagram, WhatsApp and Facebook. Second, at large volumes HeyReach Unlimited is simply cheaper: above about 125 LinkedIn accounts, if you can supply proxies and need one brand, it will usually cost less than GraphKit. Our <a href="https://graphkit.com/compare/heyreach/">GraphKit vs HeyReach</a> page lets you check other account counts.</p>
<h2 id="other-alternatives">Other alternatives</h2>
<p>From the vendors’ own pricing pages, as summarised on our <a href="https://graphkit.com/compare/">comparison page</a>:</p>
<ul>
<li><strong>Expandi</strong> charges $99 per seat a month (Business) and includes a dedicated country-based IP. A unified multi-account inbox is not published as a feature.</li>
<li><strong>Dripify</strong> is $79 (Pro) or $99 (Advanced) per seat, includes a unique local IP, and publishes a cap of 75 connection requests per day per account.</li>
<li><strong>Waalaxy</strong> prices per user in euros, from €19 to €69, with monthly invitation allowances of 300 (Pro) or 800 (Advanced and Business).</li>
<li><strong>La Growth Machine</strong> prices per “identity” (a LinkedIn account plus email inboxes) and includes a dedicated mobile proxy per identity.</li>
<li><strong>lemlist</strong> is email-first, with LinkedIn on its Multichannel plan.</li>
</ul>
<p>None of them publishes Instagram, Telegram or Facebook outbound, and only lemlist lists WhatsApp, as an add-on.</p>
<h2 id="where-graphkit-is-different">Where GraphKit is different</h2>
<ul>
<li><strong>Five channels in one sequence and one inbox.</strong> LinkedIn, Instagram, WhatsApp, Telegram and Facebook, with email handed to Warmbly, Instantly, Smartlead or lemlist. WhatsApp, Telegram and Facebook are for existing contacts and opt-ins only, which GraphKit enforces with a consent basis per lead.</li>
<li><strong>Pooled seats with volume pricing.</strong> One price per account at your organisation’s tier, allocated across client workspaces.</li>
<li><strong>Self-serve white-label from 10 accounts</strong>, including custom domains with automatic certificates, and every feature (integrations included) still available under white-label.</li>
<li><strong>Inbox ownership and privacy.</strong> Assignment, internal notes, per-account privacy, and per-member account scoping. More in <a href="https://graphkit.com/blog/unified-inbox-for-outbound-replies/">a unified inbox for outbound replies</a>.</li>
<li><strong>Shareable, revocable client report links</strong>, branded when white-label is on.</li>
<li><strong>Documented webhooks</strong>: HMAC-signed, with up to eight delivery attempts over about 18 hours.</li>
</ul>
<h2 id="where-graphkit-is-weaker">Where GraphKit is weaker</h2>
<p>This list is longer than most vendors would publish. It is accurate as of the date above.</p>
<ul>
<li><strong>No free trial</strong>, and <strong>monthly billing only</strong>: no quarterly or yearly discount.</li>
<li><strong>More expensive at high volume</strong>, as the table shows, if you can run your own proxies.</li>
<li><strong>No dedicated or static IP promise.</strong> We say “no proxies to buy” for LinkedIn, Instagram, WhatsApp and Facebook, not “dedicated residential IP”. Telegram has no managed proxy.</li>
<li><strong>GraphKit never sends email itself.</strong> Email runs through a connected platform (Warmbly, Instantly, Smartlead or lemlist), so you need one of those for email steps.</li>
<li><strong>Fewer native integrations.</strong> No MCP server, and HubSpot and Clay are planned; Zapier, Make and n8n work through webhooks.</li>
<li><strong>No AI.</strong> No automatic reply tagging and no AI drafting; positive replies are labelled by people.</li>
<li><strong>No notifications or mobile app yet.</strong> HeyReach has in-app and email notifications; we don’t.</li>
<li><strong>No voice notes</strong>, no A/B variants, and campaigns can’t be duplicated or edited after launch yet.</li>
<li><strong>One brand per organisation.</strong> Running several agency brands needs separate organisations today.</li>
<li><strong>No automatic withdrawal of pending invitations and no automatic warm-up</strong> for new accounts.</li>
<li><strong>We are newer.</strong> HeyReach has years of production use behind it.</li>
</ul>
<h2 id="how-to-choose">How to choose</h2>
<ul>
<li><strong>Mostly LinkedIn, more than about 125 accounts, proxies already in place:</strong> HeyReach Unlimited is likely cheaper.</li>
<li><strong>Need AI auto-tagging, voice notes or a native HubSpot sync today:</strong> HeyReach has them; we don’t yet.</li>
<li><strong>Running outbound across LinkedIn and messaging apps, or an agency that needs self-serve white-label, pooled seats, client report links and inbox assignment:</strong> GraphKit is built for that.</li>
<li><strong>Want to try before paying:</strong> HeyReach has a free trial; GraphKit doesn’t, but bills monthly with no commitment.</li>
</ul>
<p>Prices change. Check each vendor’s page before you decide.</p>]]></content:encoded>
      <pubDate>Sat, 14 Nov 2026 10:00:00 GMT</pubDate>
      <dc:creator>GraphKit team</dc:creator>
      <category>Comparisons</category>
      <category>HeyReach</category>
      <category>Pricing</category>
      <category>Agencies</category>
    </item>
  </channel>
</rss>
