About personas

Good: a lot more job applicants are seemingly proactive about reaching out with ideas and plans. Bad: AI slop that makes sure their application gets discarded asap.

One of the many hard things about being in my business is that a key thing we help customers do is qualitative research - but most customers don't know that it's qualitative research. So they're not looking for qualitative research. Nor primary market research (since they've never heard of that phrase). They don't talk about it that way, and there's no uniform language to describe it.

Don't pay too much attention to or worry too much about other smaller brands. Companies don't get killed by small startups, even if identical. It's always the Salesforces of your category or overall weak demand (resulting from bad market, marketing, strategy, or all of the above).

Voice of customer plays a role in how companies ship vague, me-too messaging. Your customers already understand the value of your product, have learned your kinks, and see past the vague claims (because their mind fills in the blanks). Customers are great at telling you how to be a better version of the category. They parrot the obvious value their getting from you. What they can’t tell you is how to stand out in the market. They already picked you. Voice of Customer = the people who already bought. Voice of Market = everyone you still need to win. When companies test messaging with target buyers on Wynter, the most common reaction is some version of “I’ve seen these claims 1000 times.” Or “how’s this different from [category leader]?” Your customers would never tell you that. Voice of market does. You need input from both.

AI paradox - everyone's output is up, but impact isn't. Production bottlenecks went away - we can ship way more campaigns, content, product, etc. In product, building the right thing was always more impactful than building more things, but nobody believed it. AI has now made it evident that this is indeed the case. In marketing, everyone's ability to ship more has shifted the bottleneck back to attention. The idea - what to do to begin with - matters the most. Shipping AI content is not grabbing anyone's attention.

Outsourced research is often wasted research. How often does this happen: leadership orders a study from a third party vendor (or internal insights team), and most of the insights are not implemented? All the time. The research is presented, the team nods and ships on gut feel and HiPPO. People ignore research they didn’t run themselves. Research from outside the team = someone else telling you your baby is ugly. Easy to dismiss. People who actually ran the study are not in the room where decisions happen. So the group that makes decisions treats it as "mildly interesting," one input among many, and nothing changes. Product figured this out a decade ago. PMs doing the learning/research is way more common than it was years ago. In marketing, outsourcing is still the default. Someone else is delivering decks, and marketers still ship on gut. This has to change. Marketers must experience the research firsthand. They still need help designing the studies and so on, but they need to own and execute the research themselves. Only then the insights live in the room where it happens, and actually get implemented.

Assuming that your customers speak for the market is a major error. Most companies are doing some form of voice of customer: sales call recordings, customer support tickets, customer surveys. But your customers aren't your market. To grow faster, you also need to learn what your target market (where future revenue comes from) wants and thinks. That's voice of market. VoC alone means you're optimizing for people you already won. Add VoM and you see what matters to everyone else. You need both. 1+1>2.

Learning why something worked is more important than whether something worked. - Tom Chi

11 years ago I started an experimentation conference in Austin TX that was called ConversionXL Live. I was the MC and opening keynote speaker at the first event (first photo). Today that conference is well and alive - now called Circus. My role has changed from being on the stage to sitting in the back row as a regular attendee (second photo). @Speero_agency team has been doing a beautiful job with the event for multiple years already, bringing together this fantastic community of experimentation folks from around the world. It's a beautiful thing to witness that something you started is strong and amazing without you.

- People never do something only once. That lack of attention to details you witnessed or someone delivering a project significantly past its deadline? It’s gonna happen again and again. - Some people are slow and some people are fast. When you ask the slow people to be faster, they mostly can't. I've seen too many people who don't have a "right now" gear - If you have a person in a leadership role that’s not proactively questioning the status quo, suggesting new and better ways to do things, and constantly coming up with ideas for their domain - they likely never will. - A-players have strong intrinsic motivation. They want to win and succeed - for themselves. The mediocre folk want endless kudos for everything. - If a new team member is not kicking ass in 3 months, they likely never will. (But they can still be useful). - Drama kings and queens will bring a lot of social energy and excitement to a team at first. Over time it will inevitably turn into poison and gossip, which can cast a shadow on the whole culture. You have to get rid of them. - If you the leader don’t actively manage the expected standards and pace, both will drop - and the gravity is strong. Lead by example on both fronts. Lessons learned on the road over the last 20 years.

I pinned Anthony Pierrii down for 45 minutes to talk about using AI in positioning. Here are 10 things I learned: 1. You’ve got 3 options. You can pivot into a new AI category, use AI to stand out inside your current category, or not lead with AI at all. Pick one on purpose instead of drifting into it. 2. Only pivot into a new AI category if one of these is true. You believe SaaS UI is dying, or your buyers live in Claude and the terminal, or your AI product is clearly beating your core product. If need you have a strong (actual) point of view that you're betting your company's future on. 3. A new category needs a job to be done. If it isn’t on anyone’s job description or calendar, nobody will see themselves in it. You’ll be pushing a boulder uphill forever. 4. Don’t walk away from what people already buy you for. If buyers still shop for your category, renaming yourself “AI agents” just sends your inbound to your competitors. 5. If you go “AI-native [category],” your first fight is with the incumbent, not the other startups. Win the “why not HubSpot” argument first. Looking like the other challengers is fine. Beat them on distribution. 6. Your reputation decides whether anyone believes the AI claim. If you’re seen as a legacy player, “agentic” reads as lipstick on a pig. 7. Your segment decides how loud you go. AI-forward buyers want more of it. Traditional industries hear “AI” and think “hallucination.” 8. Don’t sell the vibe, sell the use case. Say exactly what the AI does and when you’d use it. Most people won’t figure it out on their own. 9. Treat it as a bet. You can’t A/B test a whole company’s positioning. Commit for 6-12 months, then read the signal. 10. Leaving AI out of the headline is a legit choice. You can still mention it lower on the page so LLMs know you do it.

The words 'agentic' and 'AI-powered' are top contenders for the most overused terms in B2B. I analyzed 323 category leaders (top 4-5 brands across 77 categories). October 2024 the word 'agentic' was on just 1% home pages. October 2025 - 14.9%. Today? 32.2%. If we combine 'agentic' with AI agents and autonomous agents, we're at 53% of home pages. Some examples: * Webflow went from Visual website builder to The agentic web platform for modern businesses. * Attio went from The next gen of CRM to The CRM for agentic revenue. * GitLab went from AI-powered DevSecOps platform to agentic software engineering. AI-powered is also still going strong at 58% of home pages. Certain categories are bigger adopters/offenders than others. "Agentic" in sales enablement went from 0% to 67%, and in A/B testing it went from 13% to 63%. Some categories are still holding out. Payroll, billing, e-signature, and survey tools have zero agentic homepages. Only 20 of the 323 homepages (6%) don't mention AI anywhere. When they zig, zagging is a good idea.

We surveyed 100 B2B marketing leaders (VPs, CMOs, heads of marketing at B2B tech companies) about how they actually figure out what buyers think. Not what they'd like to do. What they did last time. The answers were... honest. Here are the five things that stuck with me. 1. Gut feel is the most popular research method 35% said their last big marketing decision was mainly a gut call from leadership. 13% based it on research with target buyers. The rest went with internal data, sales feedback, analyst reports, etc. Somebody senior felt strongly. Done. 2. Market research is the flossing of B2B marketing Everybody says they do it. Your dentist knows the truth. 76% told us yes, we research buyers who aren't our customers. Then we asked them to describe the last time their team tried to understand what buyers think. Only 37% mentioned a single non-customer. And the #1 way they learn about non-customers? Chatting at conferences and trade shows. One AVP of Product Marketing summed up the industry better than any report could: "Asked sales what they think, we argued about our opinions, and no one bothered to ask anyone in our target audience, we defaulted to guessing like we usually do." Like we usually do. 3. AI made us better at listening to the same people 92% mine sales calls. 85% interview customers. That part's handled. About a third mention AI in how they understand buyers. Mostly running Claude or ChatGPT over Gong calls, CRM notes and transcripts. Which is great. But it's the same folks. People who already bought from you, or are already in your pipeline. Meanwhile only 53% talk to buyers who aren't customers once a year or less. Or never. This is where the next growth lever is. 4. Enterprise marketers have the nicest kitchens and eat cereal The 1,000+ employee companies checked more boxes for how they research non-customers. Surveys, message tests, perception studies, the works. But 68% of them actually do it once a year or less. For mid-market companies it's 47%. Bigger budgets, in-house research teams, fewer conversations with strangers. 5. The #1 thing these marketers wish they knew about their market? Why buyers choose, switch, or just stall and do nothing. About a third said some version of it.

Observation about messaging: AI is the new table stakes. It's not impressive, it's assumed. And if you're not specific about the exact capabilities, people will roll their eyes.

Went to ACL this weekend. When registering the wrist band, they ask for your age bracket: 25-29, 30-35 etc The top bracket? 50+ Not 50-59, 60-... Lol "Old people"