<?xml version="1.0" encoding="utf-8"?><feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en"><generator uri="https://jekyllrb.com/" version="3.10.0">Jekyll</generator><link href="http://garnaout.github.io/feed.xml" rel="self" type="application/atom+xml" /><link href="http://garnaout.github.io/" rel="alternate" type="text/html" hreflang="en" /><updated>2026-10-02T21:42:16+00:00</updated><id>http://garnaout.github.io/feed.xml</id><title type="html">Georges Arnaout</title><subtitle>Georges Arnaout advises SaaS leadership teams and investors on retention, expansion, and redesigning Customer Success around AI agents.</subtitle><author><name>Georges Arnaout</name></author><entry><title type="html">The CCO role is changing fast</title><link href="http://garnaout.github.io/blog/the-cco-role-is-changing-fast/" rel="alternate" type="text/html" title="The CCO role is changing fast" /><published>2026-06-01T00:00:00+00:00</published><updated>2026-06-01T00:00:00+00:00</updated><id>http://garnaout.github.io/blog/the-cco-role-is-changing-fast</id><content type="html" xml:base="http://garnaout.github.io/blog/the-cco-role-is-changing-fast/"><![CDATA[<p>The CCO role is changing fast.</p>

<p>Retention, expansion, churn, NPS, and capacity still matter. But the new questions are different:</p>

<ul>
  <li>Where should humans stay involved?</li>
  <li>Where should AI agents act?</li>
  <li>What decisions can systems make safely?</li>
</ul>

<p>How do we protect customer intimacy without adding headcount?</p>

<p>This is not a tools conversation. It is an operating model conversation.</p>

<p>The most interesting CS work happening right now is not “using AI.” It is redesigning Customer Success around what humans do best, and what agents can now do better.</p>

<p>That is where I believe the CCO role is going.</p>

<p>Curious how other CS leaders are thinking about this.</p>

<p><em>First published on LinkedIn.</em></p>]]></content><author><name>Georges Arnaout</name></author><summary type="html"><![CDATA[Retention, expansion, churn, NPS and capacity still matter. The new questions are about where humans stay involved and where agents act.]]></summary><media:thumbnail xmlns:media="http://search.yahoo.com/mrss/" url="http://garnaout.github.io/assets/img/social-card.png" /><media:content medium="image" url="http://garnaout.github.io/assets/img/social-card.png" xmlns:media="http://search.yahoo.com/mrss/" /></entry><entry><title type="html">Most Digital Customer Success models are about to break</title><link href="http://garnaout.github.io/blog/most-digital-customer-success-models-are-about-to-break/" rel="alternate" type="text/html" title="Most Digital Customer Success models are about to break" /><published>2026-02-01T00:00:00+00:00</published><updated>2026-02-01T00:00:00+00:00</updated><id>http://garnaout.github.io/blog/most-digital-customer-success-models-are-about-to-break</id><content type="html" xml:base="http://garnaout.github.io/blog/most-digital-customer-success-models-are-about-to-break/"><![CDATA[<p>Most Digital Customer Success models are about to break. Not because tools are getting better. Because work itself is getting automated at the execution layer.</p>

<p>Digital CS today exists to solve one problem: “How do you serve thousands of low-ARR accounts without blowing up cost-to-serve?”</p>

<p>The answer has been: scaled humans plus automation.</p>

<p>That answer is about to become obsolete. Here is why.</p>

<h2 id="1-execution-becomes-software">1. Execution becomes software</h2>

<p>When agents can run onboarding, renewal prep, expansion triggers, and risk mitigation on their own (and do it well), “digital CSM” stops being a role and becomes an exception handler.</p>

<h2 id="2-the-cost-curve-inverts">2. The cost curve inverts</h2>

<p>Today: cost-to-serve rises with account volume.</p>

<p>With agents: marginal cost trends toward zero.</p>

<p>That changes pricing, coverage, and who even deserves a human.</p>

<h2 id="3-coverage-ratios-stop-being-a-constraint">3. Coverage ratios stop being a constraint</h2>

<p>It is no longer 1 to 200 or 1 to 500. It is 1 to 2,000 or more, because most accounts never touch a human.</p>

<h2 id="4-digital-playbooks-stop-working">4. Digital playbooks stop working</h2>

<p>Quarterly check-ins, webinars, generic campaign emails, static health scores. These are batch systems.</p>

<p>Agentic CS is continuous and customized per each customer.</p>

<h2 id="5-two-cs-motions-emerge">5. Two CS motions emerge</h2>

<p>Machine CS: autonomous execution for the long tail.</p>

<p>Human CS: strategy, company politics, relationships, and expansion for the top tier.</p>

<p>Same org. Completely different economics.</p>

<h2 id="what-replaces-todays-model">What replaces today’s model</h2>

<p>I talk to CS leaders on a daily basis and unfortunately most teams are trying to make Digital CS more efficient. That misses the point.</p>

<p>Digital CS is not disappearing. It is evolving into something fundamentally different.</p>

<p>What replaces today’s model is a machine-led success motion, with humans operating as supervisors, designers, and escalation owners.</p>

<p>If you are building your Digital CS motion today, I’d love to hear your thoughts.</p>

<p><em>First published on LinkedIn.</em></p>]]></content><author><name>Georges Arnaout</name></author><summary type="html"><![CDATA[Scaled humans plus automation was the answer to serving thousands of low-ARR accounts. Agents change the economics, and the model with them.]]></summary><media:thumbnail xmlns:media="http://search.yahoo.com/mrss/" url="http://garnaout.github.io/assets/img/social-card.png" /><media:content medium="image" url="http://garnaout.github.io/assets/img/social-card.png" xmlns:media="http://search.yahoo.com/mrss/" /></entry></feed>