Gisela Picon

In the ChatGPT era, feature-specific pages win customers for you. Your prospects might have previously searched for the “10 best [category] tools” and found listicles. They still might. === But if they use ChatGPT or Claude, the results look very different from Google’s top 10 links. Try searching for your own category, and you’ll instantly notice how feature-heavy the summaries are. Even without the user asking for anything specific, AI guides them to what the product does and what makes it different. === Feature pages matter because: ✅ Your sales team needs them to convince leads. ✅ Your website needs them to guide visitors from an overview to details. ✅ Now, ChatGPT needs them to decide what to say about you. === With the SaaS GTM pro Alexander Estner, we put together a comprehensive guide on how to create the best feature pages. I’ll drop the link to the guide in the comments 👇

122

45 comentarios

trying to survive the ai search era?? here’s the playbook you need: ai search is where visibility now lives. chatgpt, perplexity, gemini, etc. they’re becoming the core signal for marketing performance. (thanks for the great insights here Malte Landwehr!) traditional seo = keywords + 10 links. ai search = answers first, clicks second. if your business relies on website traffic, this is huge. some publishers already see 30%+ drops in traffic from ai summaries alone. the new playbook: 1. accept the shift 2. create content ai can’t summarize (guides, interactive tools, unique experiences) 3. rethink metrics: clicks aren’t enough.  reach, influence, and brand signals matter. founders, adapt now or watch visibility disappear!

102

73 comentarios

🚨 Confirmed - ChatGPT uses Google SERP Snippets for its Answers! Here's a test I've run with proof: I’ve run a simple but straight-forward to follow test that confirms the reliance of ChatGPT on Google SERPs snippets for its answers, showing that: 1. ChatGPT pretty much was able to give an answer *only* after Google had started to show the page in its SERPs, while the page wasn't being indexed in Bing yet 2. ChatGPT confirmed it was providing an answer (which was incomplete) by "fetching a snippet from the publicly indexed version of the page...  via web search that previews the page content – likely from search engine indexing“. 3. When comparing the snippet answer given by ChatGPT vs the one of Google SERPs: I confirmed it was the same! Here's my full article with the steps and outcome of the test: https://lnkd.in/dWBCa9HX PS: I guess this is yet another reason why traditional SEO is not dead and shows once again how ahead Google is not only with their search index, but the way Gemini is able to just go and fetch the page to give an accurate answer even when the page wasn’t yet indexed by Google still, rather than a incomplete, inaccurate one like ChatGPT, relying on search results snippets from… Google. 😂

2173

167 comentarios

HubSpot released an actually solid agent builder, OpenAI is releasing MCPs + agent building, Outreach has solid agents + orchestration and Clay now can sync with your whole CRM. AI in GTM is maturing, and the exact layers you need to build an AI enabled GTM engine are stabilizing. I see 11 layers you need to stay ahead of the curve... most of which will eventually go into 1 platform. 1. Company data Apollo.io and ZoomInfo company lists usually aren't enough... We like custom business directoriesor DiscoLike lists, push to Clay to qualify your TAM > SAM 2. Contact job data Most data providers still don't have up to date employment info. Our solution: Clay lists to pull contacts, then check Linkedin to find current job 3. Validated contact data The gap is widening between regular database quality and actually CLEAN contact data. LeadMagic and TitanX are really the only names worth mentioning here 4. Signals Break it in 3 categories: qualify ICP, score fit, score timing Our list: Clay again, testing Octave for account qualification Worth a look: Common Room (lots of fit and timing signals in one place), vector.co, RB2B (OGs in person level website visitors), Starbridge Apollo.io, Zoominfo, Hubspot agents are interesting too 5. Product marketing Messaging varies based on account segment, use case, persona, product, competitor etc. Octave is really the only player here that gets the problem, or you need docs & decks Shoutout for Julia Hugg for helping unpack this for me. 6. Channel content Make the messaging work where prospects see it. Octave yes, Outreach, AirOps, HubSpot, Copy.ai, and now Clay (doc uploads) and even Lovable 7. Channel system of action Getting messaging in front of prospects Nooks, Orum 🥇, Outreach, HeyReach.io, Letterdrop, Salesloft, Trellus (YC W22), FrontSpin, Smartlead, Instantly.ai, Warmly, Unify, lemlist even Webflow and Lovable 8. Orchestration Right message, Right prospect, Right time. All 7 previous steps together. Clay, Common Room, Warmly, Unify all are gunning for this space. Critical b/c many channel actioning tools don't have proper CRM integrations 9. Data warehouse This data needs to live somewhere: it's the CRM today but... Hubspot and Salesforce don't have the data objects today to support all the previous layers. Someone will figure it out 10. Reporting What signals, messaging, channels, playbooks, reps, workflows etc etc are working, when, at what stage in the funnel Pocus and Common Room are probably most solid from what I've seen here... but not totally. 11. The human in the loop No messaging today is really good enough to NOT have a smart human in the loop somewhere. No, you don't need to do all this at once 1-3 in each section, 1 in each section to start, get it working In the words of Joey Gilkey "simple scales, complex fails" "Isn't this whole thing too complex?" No, skip a step here, run into problems Comment with questions below, I'll do my best to answer.

181

75 comentarios

This 7-step ABM campaign closed $472,000 in 180 days for an 8-figure SaaS: Here’s what Ivan Falco did for AirOps 👇 1. Find existing high-ROI clients → We looked at who had already closed and generated lots of revenue. And we found more like them. 2. Create ABM target → We mapped their entire TAM. And started by priorizing lookalike accounts (which we assumed were worth the extra effort). 3. Segment personas → We identified various buying groups. Made some additional research on each persona. The goal: to speak their language and use their words in our messaging. 4. Personalize touchpoints → Whether it’s Ads, emails or landing pages. We tailored each touchpoint to the individual persona and account. 5. Sync all Channels → Make sure all channel engagement is synced together. Every ads, outbound or content interaction is tracked inside the CRM. 6. Personalize messaging at every stage. → We adjusted messaging based on our buyers’ specific pain points and where they stood in the buying journey. 7. Optimise & Scale → We cut what didn’t convert. Doubled down on what did. Want to set up something similar? Book a meeting on our site: coldiq.com/get-in-touch ✌️ Tools used: - LinkedIn Ads: Fibbler, Canva, Supermetrics, Looker, HubSpot, HockeyStack. - Outbound: Clay, Semrush, LeadMagic, Smartlead, AirOps. - Intent Data: PhantomBuster, Trigify.io.

254

139 comentarios

Most B2B decks don’t “inform.” They kill momentum. Counter‑intuitive truth: static slides make buyers do the work. Short, personal video does the work for them. What’s working now (steal this): 1️⃣ Ditch the deck → send a 90–120s VSL. One person. One problem. One next step. 2️⃣ Talk to one buyer, not “the committee.” Win one human, then arm them with your clip. 3️⃣ Show ONE workflow, not your product. Pick the 45 seconds where you remove pain. 4️⃣ Lo‑fi > studio. Phone cam + screen share + captions beats “marketing gloss.” 5️⃣ No “let me know.” Use a binary, low‑risk ask: “Wire this to your ATS for a 10‑day test?” 6️⃣ Record after discovery. Use their words, then show the 30‑second fix. 7️⃣ Proof = mechanism, not brag. “35% faster because scheduling happens in Slack. No tab‑hopping.” 8️⃣ Drop VSLs where deals die: • Post‑cold nibble (pattern break) • Post‑discovery (recap + one clip) • Post‑demo (re‑anchor outcome + micro‑ask) 9️⃣ Measure replies, not views. If it doesn’t book step 2, it’s content—not a VSL. 🔟 Tiny case that beat a deck: Swapped a PDF for a 2‑min VSL: buyer’s bottleneck → 45‑sec screen → one test ask. Same‑day reply. Second call booked. +$70k pipeline. No new features—just the format. Copy‑paste script (fill the blanks): ➤ “Hey [Name], you mentioned [pain in their words]. Here’s [specific view] removing [manual step / delay]. Teams see [credible result] because [mechanism]. If helpful, I can [low‑risk next step: sandbox / connect to X / 10‑day pilot]. Want me to [specific ask + time window]?”

139

Static ICP is dead. I heard this 7 times last week from some of the smartest GTM + AI leaders. And they’re right. The way we define ICPs is evolving fast – and most teams are already behind. 𝐒𝐭𝐚𝐠𝐞 1️⃣ – 𝐓𝐡𝐞 𝐏𝐞𝐫𝐬𝐨𝐧𝐚 𝐄𝐫𝐚 The old playbook: branding consultants drawing personas like “50–500 employees, $10M revenue, US/CA.” Zero real data. Zero connection to actual users. If your consultant never met your customers, you weren’t building strategy – you were daydreaming. 𝐒𝐭𝐚𝐠𝐞 2️⃣ – 𝐓𝐡𝐞 𝐁𝐞𝐡𝐚𝐯𝐢𝐨𝐫 𝐄𝐫𝐚 We got slightly smarter. We reverse-engineered ICPs from product usage. Spot PQLs. Ask marketing for “more of them in the funnel” and call them MQLs. Better than guessing – but still reactive. We only learned who was a fit after they engaged. 𝐒𝐭𝐚𝐠𝐞 3️⃣ – 𝐓𝐡𝐞 𝐒𝐢𝐠𝐧𝐚𝐥 𝐄𝐫𝐚 AI flips the script. We can now predict behavior before leads even show up in our martech stack. Accounts reveal buying signals early. And we can create mass-personalized journeys at scale. Two prospects. Two completely different buying experiences. Both tailored to their exact needs. The age of static ICPs is over. Signal-based, dynamic ICPs are the new gold standard. Where’s your GTM team? Still stuck in personas and PQLs – or already surfing the new wave?   --   💡 Want more go-to-market insights? Opt in for a weekly dose of free actionable tips here: https://lnkd.in/dHQ-GBFZ

294

75 comentarios

“GEO” was coined by a research paper about “how to appear more often in LLM answers.” But that same paper suggested keyword stuffing as a valid method. The so-called strategies were nothing new — just outdated SEO tricks we left behind years ago. Still, hype-marketers got excited and started pitching GEO to CMOs and COOs with FOMO. Today at WordCamp US 2025, Danny Sullivan said it directly: 👉 There is no fundamental difference between GEO and SEO. Two things matter here: Google is a “Kremlinologist” entity. They say one thing, but act another way.(The term “Kremlinology” comes from the Cold War, to analyze the lies of the Kremlin.) I never take Google’s words at face value. When a Googler says “GEO is SEO” — I agree, because it’s simply the truth. What’s being called “Generative Engine Optimization” mostly relies on: Open Information Extraction Semantic Similarity with embedding algorithms But here’s the risk: while chasing AI Overview or “AI Mode” visibility, you can lose your page-level rankings. 🔑 SEO today has three layers — what I call the SERP Triad: Without ranking a web document, your passages won’t rank. Without ranking passages, RAG systems won’t cite you. These are tightly bound in a diverse, granular system. To go deeper, join our community/course: https://lnkd.in/dp8mGy8E

130

5 comentarios

Ranking on AI isn’t the same as ranking on Google. Well, almost. Nick Lafferty from Profound analyzed 30 million citations from ChatGPT, Google AI Overviews, and Perplexity. What they found should change how you think about search visibility. → These platforms don’t just pull from high-authority websites → They pull from Reddit → From YouTube → From Quora → From LinkedIn → From Wikipedia We turned Profound’s findings into a carousel you can act on: → What each AI model prioritizes → Where community and social content wins → What to do right now to show up in AI-driven search All credit to Nick Lafferty and the Profound team for the research. We just added the strategic lens and action steps! If AI visibility matters to your brand, this is worth your time. Slides below ↓

54

21 comentarios

How it started: - Website identification will change cold outbound forever! - Signals are the future for all SDRs! - AI agents can handle campaign creation and management end-to-end! How it's going: - Website identification still struggles to properly identify users (it does it off of IP - meaning it can grab wrong people from the same office/home) - Signals often work, but actual volume of people visiting your CEOs profile is maybe a few hundred a month - and not nearly enough volume to be significant for 99% of companies. - AI can help with campaigns, but still hallucinates to the point where not having a human handle certain parts of it is a campaign death sentence. We'll get there one day, but don't conflate social media hype with actual value.

76

37 comentarios