<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Meet Chande on Medium]]></title>
        <description><![CDATA[Stories by Meet Chande on Medium]]></description>
        <link>https://medium.com/@meet.chande?source=rss-607ee1a13635------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*DCwmlGKyx0Mo6Ob9BTEkaA.jpeg</url>
            <title>Stories by Meet Chande on Medium</title>
            <link>https://medium.com/@meet.chande?source=rss-607ee1a13635------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 05:15:52 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@meet.chande/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[It Took Us a Year and 200,000 Emails to Get Outbound Right. Here’s the Playbook.]]></title>
            <link>https://medium.com/@meet.chande/it-took-us-a-year-and-200-000-emails-to-get-outbound-right-heres-the-playbook-7a299e60928a?source=rss-607ee1a13635------2</link>
            <guid isPermaLink="false">https://medium.com/p/7a299e60928a</guid>
            <category><![CDATA[sales]]></category>
            <category><![CDATA[outbound]]></category>
            <category><![CDATA[email]]></category>
            <dc:creator><![CDATA[Meet Chande]]></dc:creator>
            <pubDate>Tue, 10 Nov 2026 21:01:01 GMT</pubDate>
            <atom:updated>2026-11-10T21:01:01.934Z</atom:updated>
            <content:encoded><![CDATA[<p>If you’re about to start outbound email, I want to save you the year I lost.</p><p>For most of last year, our outbound didn’t work. We were doing too much at once: too many segments, three copies for each, different flows for each copy. Every campaign looked smart on paper and told us almost nothing.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_0oZ3spEpSMdTxoW9zUh2Q.png" /></figure><p>What fixed it was bringing in someone who had done outbound at scale, letting him lead, and paying attention to where we fell short. His approach came down to one idea: outbound is a game of scale, and scale needs simplicity.</p><p>Since then we’ve reached around 100,000 leads and here’s exactly how I’d do it if I were starting from zero tomorrow.</p><h3>Start with less than you think you need</h3><p>Pick a few segments. By segment I mean a mix of industry, company size and role.</p><p>Then write two or three versions of your email. Change the angle, the length or the call to action between them. That’s it.</p><p>It will feel too simple. That’s the point. Every extra segment or copy splits your volume, and your answers with it.</p><h3>Test before you trust anything</h3><p>Send three or four test campaigns of 2,000 to 3,000 contacts each. You’re looking for four answers:</p><ul><li>Which segment replies</li><li>Which copy works</li><li>Which day works best</li><li>Weekdays or weekends</li></ul><p>Judge all of it on one number: positive replies.</p><p>I’m strict about this. Outbound’s job ends when someone says they’re interested. Turning that into a sale is the sales team’s job. Blend the two and you’ll never know whether your emails or your follow-up is the problem.</p><h3>Then go big on what won</h3><p>Once one combination clearly wins, scale it exactly as it is. Same segment, same copy, same day. We took ours to 50,000 to 60,000 leads.</p><p>The hardest part is leaving it alone. You’ll want to tweak it. Hold off until it has proven itself at scale.</p><p>You know it works when it brings in revenue and a few paying clients.</p><h3>My 2 cents</h3><p>Keep it simple, test small, and scale what wins.It sounds obvious. It took us a year to actually do it.</p><p>I’m sharing the full story on Substack, including what we got wrong and how we run each test. Follow me there: <a href="https://substack.com/@meetchande">substack.com/@meetchande</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7a299e60928a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What does a GTM engineer actually do? The job comes down to revenue]]></title>
            <link>https://medium.com/@meet.chande/what-does-a-gtm-engineer-actually-do-the-job-comes-down-to-revenue-b3a912586286?source=rss-607ee1a13635------2</link>
            <guid isPermaLink="false">https://medium.com/p/b3a912586286</guid>
            <category><![CDATA[marketing]]></category>
            <dc:creator><![CDATA[Meet Chande]]></dc:creator>
            <pubDate>Sat, 10 Oct 2026 13:14:27 GMT</pubDate>
            <atom:updated>2026-10-10T13:14:27.241Z</atom:updated>
            <content:encoded><![CDATA[<p>I get job outreach every other day now, on email, LinkedIn, and sometimes even WhatsApp. That tells me the market for GTM engineers is clearly hot, and that a lot of people have jumped in over the last few months, each with a very different skill set.</p><p>For anyone new to the term, a GTM engineer builds and runs the systems behind how a company sells and markets, like outbound, enrichment, CRM automation, and lead routing. Some people say it’s just RevOps with a new name, and the two do overlap, but the day-to-day looks very different from one organization to the next, which is why the skill sets vary so much.</p><p>My 2 cents, a GTM engineer’s only job is to increase revenue. The tools, the automations and the dashboards are just ways of getting there, which means the real job is owning the whole funnel, finding where it leaks, and fixing those gaps before you try to expand it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iPQI0DSzwoflWcS9LKvR8Q.jpeg" /></figure><h4>The gap can be anything</h4><p>It isn’t always a missing system. Sometimes attribution doesn’t add up, so nobody knows what’s actually working. Sometimes lead response is slow and good leads go cold, sometimes the middle of the funnel is stuck, and sometimes the sales team simply needs more resources. Building something new is only one of the possible answers, and often it’s not the right one.</p><h4>What this looked like for me recently</h4><p>Over the last couple of months, most of my time went into creating and improving an outbound play and scaling it to 100k leads. Around it I built tools and automations, hired and set up the sales team, switch to new platforms, and etc.</p><p>Last week was about the middle of the funnel. It’s tempting to go straight for more leads at the top because that’s the visible part, but more leads into a slow middle just means more leads going cold. So I fixed the middle first and made it quick, before bringing more in at the top.</p><h4>Slow down for ten seconds</h4><p>The biggest lesson from these months is that when you’re moving fast, especially with AI, you still need to step back for ten seconds and check what’s missing, because the details are where things break. Most of the time should go into strategy rather than execution, since a flawed process executed faster is still a flawed process.</p><h4>Why skill sets will keep differing</h4><p>Skill sets will keep differing because every organization is different, with different systems and different plans. The one question that holds across all of them is whether the revenue is moving.</p><p>I’ll keep writing about what I’m building in GTM and what breaks along the way, so follow along if that’s useful.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b3a912586286" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Tool Count Is the Wrong Question. Ownership Is the Real Cost.]]></title>
            <link>https://medium.com/@meet.chande/tool-count-is-the-wrong-question-ownership-is-the-real-cost-6a60b1a2cd0a?source=rss-607ee1a13635------2</link>
            <guid isPermaLink="false">https://medium.com/p/6a60b1a2cd0a</guid>
            <category><![CDATA[cost-reduction]]></category>
            <category><![CDATA[gtm]]></category>
            <category><![CDATA[tools]]></category>
            <dc:creator><![CDATA[Meet Chande]]></dc:creator>
            <pubDate>Wed, 05 Aug 2026 12:11:01 GMT</pubDate>
            <atom:updated>2026-08-05T12:11:01.414Z</atom:updated>
            <content:encoded><![CDATA[<p>Here’s the point of this post in one line: the number of tools you run doesn’t cost you money. Two tools doing the same job.</p><p>I cut my own outbound stack by 79% this year, from $1,989 a month down to $417. It looks like a story about cutting tools. It isn’t. I kept the same number of jobs covered. What I removed was duplication, tools fighting over a job that only needed one owner. That distinction is the entire post.</p><figure><img alt="Tool Count Is the Wrong Question. Ownership Is the Real Cost." src="https://cdn-images-1.medium.com/max/1024/1*tMqVnLwCHYmdq-MiVtVhYQ.avif" /><figcaption>Tool Count Is the Wrong Question. Ownership Is the Real Cost.</figcaption></figure><h3><strong>Why this matters right now</strong></h3><p>Right now, the industry is having the wrong argument. BetterCloud’s 2026 State of SaaS report says company software stacks are growing again, up 11% in the past year, after two years of companies trying to shrink them. Zylo’s 2026 SaaS Management Index puts the average company at 305 tools and $55.7 million a year in spend, both still climbing.</p><p>Read that as: the “cut more tools” strategy already failed once. Stacks shrank, then grew right back. That’s what happens when the fix targets the count instead of the cause.</p><h3><strong>What I actually did</strong></h3><p>I have twelve tools, not 305. Small scale, same mechanism. I went through every one of them and asked a single question: does this tool own a job that nothing else here owns?</p><p>Five tools failed that test: ElevenLabs, Sending.ac, Saleshandy, LeadMagic, Hunter. Not bad tools. Each one was doing a job something else in the stack already did. Cut.</p><p>Two tools passed the test but were still bleeding money: Apify and Apollo. One had a pricing spike I hadn’t noticed. The other, I was paying for volume I didn’t use. Kept, renegotiated.</p><p>Five tools passed clean: HubSpot, Instantly, Zapmail, Unipile, SalesMsg. Each one owns a job nothing else touches. Untouched.</p><p>That’s the whole method. Not “which tools are expensive.” Which tools are duplicated.</p><h3>The number that explains why this is everywhere</h3><p>BetterCloud’s report has one stat that made this click for me: only 56% of the apps inside the average company’s stack have any formal approval at all. Nearly half got added because someone needed a job done fast, and nobody checked whether it was already covered. That’s duplication at enterprise scale. My twelve-tool version of the same mistake was smaller, but it was the identical mistake.</p><p>So the rule I use now, on anything new: every tool gets exactly one job. If two tools are competing for the same job, that competition is the cost. Not the invoice. The confusion.</p><h3>Why this only works in this order</h3><p>Once the five surviving tools had zero overlap between them, something else got easier: connecting them. I built a GTM agent that runs the sequencing and maintenance across all five through their APIs and MCPs. Most of my outbound runs itself now.</p><p>That only worked because I cut overlap first. If I’d tried to automate the mess, I’d have automated five tools arguing over the same job, just faster.</p><h3>The one-line takeaway</h3><p>Don’t ask how many tools you have. Ask how many of them are doing someone else’s job. BetterCloud and Zylo both show that question getting worse at enterprise scale, even after a serious consolidation push. It held true at the scale of my twelve tools too. Fix ownership, and the cost problem mostly fixes itself.</p><p>Full tool-by-tool, dollar-by-dollar breakdown of my own audit is on Substack.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6a60b1a2cd0a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[My Full GTM Engine and the Rules Behind It (Steal It)]]></title>
            <link>https://medium.com/@meet.chande/my-full-gtm-engine-and-the-rules-behind-it-steal-it-399e5e8a20ab?source=rss-607ee1a13635------2</link>
            <guid isPermaLink="false">https://medium.com/p/399e5e8a20ab</guid>
            <category><![CDATA[outbound-marketing]]></category>
            <category><![CDATA[growth-marketing]]></category>
            <category><![CDATA[go-to-market]]></category>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[gtm]]></category>
            <dc:creator><![CDATA[Meet Chande]]></dc:creator>
            <pubDate>Tue, 07 Jul 2026 12:16:00 GMT</pubDate>
            <atom:updated>2026-07-07T12:16:00.887Z</atom:updated>
            <content:encoded><![CDATA[<p>Google recently tightened its spam filtering, and bulk-send patterns now get flagged at a much lower threshold than before. Practical caps are down to 30–50 emails per mailbox per day, and once warmup, inbox tests, and actual replies are factored in, the real number drops to 10–20. That means more mailboxes per campaign and more infrastructure cost to match. If your outbound has quietly gotten worse recently, this is likely why.</p><p>Most outbound advice online is either too high-level to act on or too gated behind a course. Here’s the actual engine we run and the actual rules we operate by across email, LinkedIn, SMS, calling, and CRM. Use whatever’s useful.</p><h3>The Logic</h3><p>Keep it simple so it’s scalable. Run multi-touch, multi-platform outreach, and never rely on one medium. That single principle is why the stack below spans five channels instead of just email.</p><h3>The Stack</h3><ul><li><strong>Instantly.ai</strong> — email warmup and campaign launch</li><li><strong>Zapmail</strong> and <strong>Sending.ac</strong> — domain and mailbox management (Google and Microsoft/Azure)</li><li><strong>SecondBrainLabs (Sbl.so)</strong> — LinkedIn campaign sequences</li><li><strong>Unipile</strong> — LinkedIn one-on-one, personalized engagement</li><li><strong>SalesMSG</strong> — SMS and calling</li><li><strong>ElevenLabs</strong> — AI-generated voice for LinkedIn audio messages</li><li><strong>GlockApps</strong> — inbox placement and deliverability testing</li><li><strong>Google Postmaster</strong> — Gmail deliverability and domain reputation monitoring</li><li><strong>HubSpot</strong> — CRM and single source of truth</li></ul><p>No single tool here is the differentiator. The stack works because of the rules layered on top of it, which is really what this post is about.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*cQQgmY8pZNzSonrsLSc6HA.gif" /><figcaption>My Outbound GTM Stack</figcaption></figure><h3>1. Email</h3><h3>Warmup and Setup</h3><ul><li>Weekday warmup: 10 emails per mailbox. Weekend: half that.</li><li>Target reply rate during warmup: 30–70%.</li><li>New accounts warm up slowly through Instantly rather than jumping straight to outbound volume.</li><li>Prewarmed domains are an option when you need capacity fast and can’t wait out a standard 30-day warmup.</li><li>Minimum domain ownership before use: 1 month. Maximum before retirement: 3 months, unless deliverability stays at 95+, in which case keep going.</li></ul><h3>Health and Deliverability</h3><p>The deliverability score is the single most important number in the system, and it dictates one of three actions:</p><ul><li><strong>Below 90:</strong> discard the account entirely.</li><li><strong>90–94:</strong> isolate it. Pull it from outbound completely, focus 100% on warming it back up, and monitor daily.</li><li><strong>95+:</strong> healthy. Proceed with outbound.</li></ul><p>A few rules that make this actually work:</p><ul><li>Inbox placement tests run on alternating weeks, on a random weekday, never on weekends. But if deliverability drops at any point, that triggers an immediate off-cycle test. Don’t wait for the next scheduled run.</li><li>Placement tests must run through a third-party platform, never your domain provider or warmup tool. They have no incentive to surface bad news about your own placement.</li><li>Google Postmaster only becomes meaningful once a domain sends 100+ emails/day to Gmail. Below that, the signal is too noisy to act on.</li><li>Domains get checked against DBL, SURBL, and URIBL blacklists. A flag means immediate removal from outbound and a root-cause investigation. If it’s unresolvable, the domain gets discarded.</li><li>SPF and DMARC are verified on every domain before it goes live. Failed auth means immediate removal until it’s fixed.</li><li>Cold email never goes out from the root domain, only from dedicated outreach subdomains, each with its own SPF record rather than one inherited from the root. If a subdomain gets burned, the damage stays contained instead of taking the company’s primary domain down with it. Worth noting this isn’t a perfect firewall: some monitoring tools, Google Postmaster included, still track reputation signals at the root domain level in certain cases, so a subdomain buys containment, not full immunity.</li></ul><h3>Bounce and Reply Handling</h3><ul><li><strong>Hard bounce:</strong> permanent removal from every sequence, flagged in CRM, no retry.</li><li><strong>Soft bounce:</strong> retry up to twice, then remove permanently.</li><li><strong>Bounce rate threshold:</strong> 2% triggers an immediate stop on that domain and a full email verification pass before resuming.</li><li><strong>Spam rate threshold:</strong> 0.3% triggers an immediate stop and a full deliverability check before any further sends.</li><li><strong>Out of office:</strong> if a referral is mentioned, add them to the sequence. If it’s just a return date, pause and resume outreach after that date.</li><li><strong>Person no longer at the company:</strong> look for a referred replacement in the reply. If there’s one, reach out to them. If not, find a replacement contact yourself and remove the original from both the sequence and the CRM.</li></ul><h3>Capacity Planning</h3><p>Infrastructure sets the ceiling. Your total domains and mailboxes determine your outbound volume for the month, not the other way around. Plan the campaign around what your infrastructure can safely support, not what you wish it could support.</p><p>One routing rule worth calling out: Google domains should send to Google inboxes (Workspace to Gmail). Microsoft/Azure domains are more flexible and can send to any provider.</p><h3>2. LinkedIn</h3><ul><li><strong>Tool routing:</strong> campaign-level sends and sequences run through SecondBrainLabs. One-on-one engagement and replies run through Unipile.</li><li><strong>Connection requests:</strong> max 20 per account per day, max 100 per account per week, spread across at least 5 days.</li><li><strong>Messages:</strong> max 20 direct messages per account per day.</li><li><strong>Profile requirements before any account goes into outbound:</strong> complete profile photo, headline, and work history, minimum 2 company experiences listed, minimum 100+ connections, and consistent ongoing posting (not just during warmup).</li><li><strong>AI-assisted profiles:</strong> if you’re running AI-assisted LinkedIn accounts alongside your team’s real profiles, hold them to the same bar. Minimum 1 month of warm-up (connection requests plus posting to build credibility) before any outbound activity starts. Treat profile completeness and posting cadence as non-negotiable regardless of who or what is behind the account.</li><li><strong>Account health:</strong> if LinkedIn flags or restricts an account, stop all outbound on it immediately. Don’t try to work around the restriction. Wait 24–48 hours before resuming at a lower volume. A real team account that gets restricted twice gets retired from outbound. An AI-assisted account that gets restricted doesn’t get appealed at all — retire it and start a fresh one from warmup.</li></ul><h3>3. SMS and Calling</h3><ul><li>Follow TCPA regulations for US numbers and ICO/PECR regulations for UK numbers, full stop.</li><li>Maintain opt-out records and honor them immediately, every time.</li><li>Only contact leads within their local business hours (8 am - 9 pm), and never on Sundays or public holidays.</li><li>Always call at the lead’s local time. This one’s non-negotiable.</li><li>Always leave a voicemail if the call isn’t picked up, max 30 seconds, following the campaign’s script.</li></ul><h3>4. CRM</h3><p>HubSpot is the single source of truth. Every channel, email, LinkedIn, SMS, calling, syncs activity back to the contact record so there’s one complete view of every lead.</p><p>A few rules that keep it useful instead of just a data dump:</p><ul><li>Any lead that replies positively or books a demo becomes a hot lead immediately, and hot leads get a personalized, dedicated follow-up sequence. They don’t get treated like standard outbound.</li><li>The CRM gets audited every two weeks, checking for stale records, incorrect stages, missing activity, and leads that should’ve been moved or removed but weren’t.</li><li>All campaign-level and reporting decisions get made from HubSpot data, not from whatever a single tool’s dashboard shows.</li></ul><h3>The Actual Takeaway</h3><p>None of this is exotic. There’s no secret tool or hack in here. What keeps this system working is that the thresholds are hard rules, not suggestions: 90 for deliverability, 2% for bounce rate, 0.3% for spam rate, 20 connection requests a day. The temptation in outbound is always to push past these numbers for one more send, one more campaign, one more quarter of pipeline. The moment you do, deliverability degrades faster than it recovers.</p><p>Steal the stack, steal the thresholds, but mostly steal the discipline around them.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=399e5e8a20ab" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Google Just Killed FAQ Rich Results. The Same Cycle Is Now Starting in AI Search.]]></title>
            <link>https://medium.com/@meet.chande/google-just-killed-faq-rich-results-the-same-cycle-is-now-starting-in-ai-search-f0d929cf6607?source=rss-607ee1a13635------2</link>
            <guid isPermaLink="false">https://medium.com/p/f0d929cf6607</guid>
            <category><![CDATA[google]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[search-engine-optimizati]]></category>
            <category><![CDATA[seo]]></category>
            <category><![CDATA[digital-marketing]]></category>
            <dc:creator><![CDATA[Meet Chande]]></dc:creator>
            <pubDate>Mon, 01 Jun 2026 03:41:00 GMT</pubDate>
            <atom:updated>2026-06-01T03:41:00.592Z</atom:updated>
            <content:encoded><![CDATA[<h3>TL;DR</h3><p>On May 7, Google killed FAQ rich results in regular search. Seven years of FAQ schema being standard SEO advice, gone by June. The same cycle is now starting in AI search, just faster. This piece walks through how the Google cycle worked, what’s happening in GEO right now, and what to do this month if you have FAQ schema on your pages.</p><h3>What Google announced</h3><p>The deprecation notice on Google’s developer documentation is one paragraph long. As of May 7, 2026, FAQ rich results no longer appear in Google Search. By June 2026, the search appearance, rich result report, and Rich Results Test support are all gone.</p><p>Two things worth noticing in what Google didn’t say. There is no instruction to remove the markup. And there is no explanation for why this is happening now, after seven years of FAQPage being a routine SEO recommendation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*EIMTjXBb22ZTSr-1" /><figcaption>Photo by <a href="https://unsplash.com/@shutter_speed_?utm_source=medium&amp;utm_medium=referral">Shutter Speed</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3>The Google cycle</h3><p>The FAQ schema story is a seven-year arc, and the shape of it is the part to remember.</p><p>Google introduced the FAQPage schema in 2019. It was easy to add; Yoast made it a two-click setting, and most other SEO plugins followed. It produced a visible reward: those rich-result boxes in the SERP that looked great in screenshots and lifted click-through. Marketers added it everywhere. Product pages got fake FAQs, essays got fake FAQs, etc. The format was abused to the point of uselessness because almost nobody checked whether the questions on the page were actually ones the page answered.</p><p>By 2023, Google had cut the rich results back to government and health sites. Three years later, it removed the rest.</p><p>Lily Ray, who helped popularize FAQ schema with a 2019 Moz article, included a section in that piece warning about exactly this risk. Her phrasing aged well: <em>anything that can be spammed in SEO, will be spammed.</em></p><p>That’s the cycle. A platform introduces a markup signal that gives marketers a visible reward. Marketers add it everywhere, including pages where it doesn’t fit. The signal stops being a quality filter. The platform takes the reward away.</p><h3>The GEO replay in real time</h3><p>That same cycle is now starting in GEO, and unlike the SEO version, this one is compressing into months instead of years.</p><p>Open LinkedIn this week. You’ll find hundreds of posts asserting that AI search engines specifically reward FAQ schema. The deeper problem is structural. There is no primary source from OpenAI, Anthropic, Google AI Overviews, or Perplexity confirming that they specifically favor JSON-LD over rendered HTML.</p><p>It’s the same pattern that made FAQPage what it became. Unverified claim → industry consensus → mass adoption → platform takes the lever away.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*u8o6QOStZjQM2Xlz" /><figcaption>Photo by <a href="https://unsplash.com/@sanketgraphy?utm_source=medium&amp;utm_medium=referral">Sanket Mishra</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3>Content and markup are not the same thing</h3><p>Before the prescriptive part, one distinction worth holding in your head: FAQ content and FAQ schema are different things. Most coverage of the May 7 announcement has flattened them into one, and they aren’t.</p><p>FAQ content, Q&amp;A formatting, is good for human readers. It’s scannable, maps to how people search, and is often easier to write well than expository prose. None of that changes because Google removed a rich result.</p><p>FAQPage schema is the JSON-LD format for marking up that content. Google’s rich result for it is gone. Bing still uses it. The markup itself isn’t deprecated, it’s just unrewarded by the largest search engine.</p><p>Killing one doesn’t kill the other.</p><h3>What to actually do this month</h3><p>The practical guidance is narrow:</p><ol><li><strong>Keep your FAQ content</strong> where it serves readers. The rich result going away doesn’t change whether Q&amp;A is a good format for the topic.</li><li><strong>Use semantic HTML.</strong> Headings, paragraphs, &lt;details&gt; and &lt;summary&gt; for collapsible sections. That&#39;s HTML&#39;s job. Schema is a separate question.</li><li><strong>Remove FAQPage schema from pages where the FAQ isn’t the point.</strong> A 2,000-word essay with three FAQs at the bottom should never have had FAQPage on it. The mainEntity claim, that the Q&amp;A is what the page is about, was a misrepresentation in those cases. It was working only because Google rewarded it.</li><li><strong>Where the page genuinely is an FAQ, keep the markup.</strong> Bing still rewards it. Google isn’t asking you to remove it. If any AI consumer ever does start favoring structured data, you want to be the one whose markup is actually accurate.</li></ol><p>The unifying rule: stop letting reward signals dictate data hygiene. Schema is a description of what your page is. If the description is accurate, keep it. If it isn’t, the rich result was the only reason to keep it, and that’s gone.</p><h3>The structural fix nobody is talking about</h3><p>Here’s something almost nobody outside the schema community is discussing: the spec itself is part of the problem.</p><p>Schema.org defines the vocabulary you use to label your pages. It has one type for FAQ content: FAQPage. The catch is that FAQPage means &quot;this page is, primarily, an FAQ.&quot; It&#39;s a label for the whole page.</p><p>That works for dedicated help-center FAQs. It doesn’t work for the much more common case: a product page, a blog post, or a service page that has an FAQ section at the bottom.</p><p>When marketers tagged those pages with FAQPage, the label was technically a lie. The page wasn&#39;t primarily an FAQ. It was a product page that happened to include some FAQs. But there was no other option in the spec. You either misrepresented your page or you skipped the markup entirely. Most people chose misrepresentation, because the rich result reward was sitting right there.</p><p>I believe a new type schema called FAQSectionwill fix this. The name does the work: it&#39;s for marking up an FAQ as one section of a page, not the whole page. Publishers could finally describe what&#39;s actually on their page without lying.</p><p>Until my proposal lands ( which doesn’t look like it will anytime soon), the safer move on any page that isn’t primarily an FAQ is to stop using FAQPage and let the questions and answers live as plain HTML. The reward is gone anyway, while the cost of being accurate is zero.</p><h3>The boring rule that survives every cycle</h3><p>The operating principle of SEO, and now GEO, is unchanged. When ranking signals are gameable, exploitation will scale faster than usefulness. The cycle keeps repeating because the people inside it keep repeating it.</p><p>If you’re writing content this quarter, the question worth asking isn’t “what does ChatGPT reward this month?” The reward signal will move. The only question you need to ask is the boring one: Does my markup describe what my page actually is?</p><p>Get that right, and you stop having to chase the next reward signal.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f0d929cf6607" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Change in Google Isn’t Answering Your Query Anymore. It’s Answering Six]]></title>
            <link>https://medium.com/@meet.chande/query-fan-out-ai-overviews-31d2867298a0?source=rss-607ee1a13635------2</link>
            <guid isPermaLink="false">https://medium.com/p/31d2867298a0</guid>
            <category><![CDATA[ppc-marketing]]></category>
            <category><![CDATA[seo]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[digital-marketing]]></category>
            <dc:creator><![CDATA[Meet Chande]]></dc:creator>
            <pubDate>Sat, 28 Feb 2026 10:19:33 GMT</pubDate>
            <atom:updated>2026-02-28T10:21:01.212Z</atom:updated>
            <content:encoded><![CDATA[<p>I wrote a couple of LinkedIn posts recently about something I’ve been noticing in my work. Pages that rank well in regular search but are completely invisible in AI Overviews. I dug into it, restructured some content, and saw pickups improve almost immediately.</p><p>The mechanism behind it is called Query Fan-Out. And it’s quietly reshaping how both organic content and paid search work. This is a slightly deeper look at what I’ve been seeing.</p><p><strong>What Query Fan-Out Actually Does</strong></p><p>When someone types a query into Google, the search engine no longer matches that query to results. It expands it. Behind the scenes, Google’s AI breaks your single query into five, six, sometimes a dozen related sub-queries, each targeting a different angle of intent.</p><p>Google formally introduced this at <a href="https://blog.google/products-and-technology/search/google-search-ai-mode-update/">I/O 2025</a> alongside AI Mode, powered by a custom version of Gemini 2.5 built specifically for query decomposition. But query fan-out also runs under AI Overviews, which now appear on a significant chunk of all Google searches. Estimates range from <a href="https://searchengineland.com/google-ai-overviews-surge-pullback-data-466314">13% to 30%</a>, depending on who’s measuring and when.</p><p>Here’s the practical version. Someone searches “best CRM for SaaS startups.” Google doesn’t just look for pages about that. It simultaneously runs sub-queries like:</p><p>- CRM pricing for startups<br>- CRM onboarding and setup time<br>- CRM alternatives and comparisons<br>- CRM integrations with common SaaS tools<br>- CRM for small teams under 20</p><p>Each sub-query pulls from different sources. The AI Overview synthesises all of it into one response. Your page either answers enough of those sub-queries to get cited, or it doesn’t appear at all.</p><p>A December 2025 <a href="https://www.ekamoira.com/blog/query-fan-out-original-research-on-how-ai-search-multiplies-every-query-and-why-most-brands-are-invisible">study by Surfer SEO</a> analyzing over 173,000 URLs found that 68% of pages cited in AI Overviews were not in the top 10 organic results. That single stat says a lot. Ranking well for the main keyword doesn’t guarantee you show up when the AI is pulling from six different angles.</p><figure><img alt="Query fanout" src="https://cdn-images-1.medium.com/max/1024/1*LRflznlZjznCT61jahv4rA.png" /></figure><p><strong>What This Means for Content</strong></p><p>My content answered one question well. Google was asking six. Once I restructured the full cluster to cover the sub-intents, not just the primary keyword, pickups went up noticeably.</p><p>This isn’t about writing longer pages or stuffing in more keywords. It’s about coverage. The AI is evaluating whether your content addresses the full scope of what a user might actually need. If your page only handles one slice of the intent, the AI will find someone who covers more.</p><p>The shift is subtle but important: authority is now measured across topic clusters, not single pages. One thorough page can work. But a cluster of interconnected pages covering the full intent map tends to perform better.</p><p><strong>What This Means for Paid Search</strong></p><p>This is the part that caught my attention most, and what I wrote about in my second LinkedIn post.</p><p>AI Overviews are educating users before they ever see an ad. They’re answering the informational and consideration-stage questions that used to drive clicks on paid results. The data backs this up clearly.</p><p><a href="https://www.seerinteractive.com/insights/aio-impact-on-google-ctr-september-2025-update">Seer Interactive</a> tracked over 25 million organic impressions and 1.1 million paid impressions through September 2025. Paid CTR on queries with AI Overviews dropped 68%, from 19.7% down to 6.34%. Organic CTR fell 61% on the same queries. July 2025 was particularly brutal, with paid CTR collapsing from roughly 11% to 3% in a single month.</p><p>Meanwhile, <a href="https://skai.io/blog/ai-overviews-2025-update/">CPCs have hit a six-year high</a>. Advertisers are spending more to capture fewer clicks.</p><p>But here’s what’s interesting: <a href="https://searchengineland.com/paid-search-pivots-google-ai-overviews-470193">65% of industries actually saw improved conversion rates</a> despite lower click volume. The reason? AI Overviews are pre-qualifying users. By the time someone does click an ad, they’ve already done their research. They’re further down the funnel.</p><p>The net effect is a quiet redistribution of the funnel. Top-of-funnel queries are increasingly being answered by AI Overviews pulling from organic content. PPC is getting pushed into brand and transactional keywords at the bottom. The middle still has both competing, but the top is becoming an organic-only game.</p><p>If you’re planning budgets right now, that shift matters. Content covering intent clusters is picking up the exact queries that PPC is walking away from.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*f4U3hSsvJ9YEJ1nrrqnYtw.png" /></figure><p><strong>The Takeaway</strong></p><p>Query fan-out isn’t a future concern. It’s already running. The content side and the paid side are both feeling it, just in different ways.</p><p>For content, the move is straightforward: stop optimizing for a single keyword and start mapping the full cluster of intents Google is likely to expand your query into. Cover the sub-questions. Make your content extractable.</p><p>For paid, the move is more strategic: recognize that high-funnel informational spend is losing efficiency fast. Redirect toward bottom-of-funnel, brand, and retargeting. The places where AI Overviews haven’t eaten the intent yet.</p><p>Small shift. Big difference.</p><p>*<em>I also shared shorter takes on the content angle and the PPC angle on my LinkedIn. Check it out there if you want the quick versions.</em>*</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=31d2867298a0" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>