OpenAI just updated its dataset on how people use ChatGPT. I went through every file and found a few surprises: 1. Classic Google query types are now about half of all messages - 48% of all messages were Google-style questions in September 2026, up from 43% in July 2024. - The 48% breaks down into 20% simple searches (facts, products, recipes) and 28% Practical Guidance (how-to, tutoring, health, creative ideas). - OpenAI sorts every message into work or personal. Google-style questions make up 53% of personal messages and 38% of work messages (September 2026). - Personal messages grew from 48% to 68% of all messages since July 2024. - Most of the rise comes from personal messages taking a bigger share, because they're search-heavy. 2. Commercial research at work grew 75% Among US work messages (Q1 2025 vs Q3 2026): - "Obtain information about goods or services": 1.7% → 2.9% - "Determine values or prices of goods or services": 0.3% → 0.7% - "Gather information from physical or electronic sources": 1.2% → 1.8% Surprised me: researching a subject for work fell by up to two thirds (technology -66%, history and social issues -65%, health -42%). Product and price research grew. 3. Countries that use ChatGPT more use it more for search - The quarter of countries with the highest ChatGPT use per capita put 24% of messages into simple searches in Q3 2026, vs 16% for the lowest quarter. - Across ~150 countries, the correlation grew from 0.30 to 0.74 (Q1 2025 vs. Q3 2026). 4. Personal use grows faster than work - OpenAI's classifier tags every sampled message as work-related or not (consumer plans only). - Work went from 52% of messages in July 2024 to 32% in September 2026. The data only shows shares, so work messages could still be growing in absolute numbers. In September 2026, the share of messages that were work-related was: - Free: 25% - Go: 42% - Plus: 51% - Pro: 68% Source: OpenAI Signals v3.0, announced in "ChatGPT's expanding world of work" 159 countries, 27 million sampled messages, July 2024 to September 2026

So, why should I not get a new iPad mini plus a norma iphone?

Godspeed, cancer research

AI is killing SEO traffic is not completely wrong, but definitely not completely right. I a new analysis of Similarweb data (10k domains, US + France, 3 years), I found Organic Search fell -13.2% in the US and 24.4% in France over the last 2 years. HOWEVER, not all of that is due to AI (Overviews). Paid Search has grown strongly over the same time horizon, probably due to more real estate in the SERPs and harder differentiation from organic results. Also, organic traffic losses are more concentrated in industries like publishing, which lost lost 41.5% in the US and 42.5% in France over the last 2 years. Lastly, Google's updates are aggressive and often reduce rank performance that can be, at least somewhat, recovered. So, as much as I see AI reducing clicks left and right, we should not throw the baby out with the bath water. Full analysis in the new Growth Memo.

ad load on youtube has gone up signifcantly. wouldn't be surprised to see a jump in youtube ad revenue next quarter.

I’d be careful about using ChatGPT as a proxy for all AI Search. Look at Gemini and Claude in this chart. ChatGPT is still ahead, but Gemini gained adoption through most of the period. Claude started climbing much faster in early 2026. This is a good reason to check what happens to the same question across different tools. Take a question that came up on a sales call. Ask it in ChatGPT, Gemini and Claude. See who gets recommended and read the sources behind those recommendations. You might find that a competitor barely shows up in ChatGPT but gets recommended regularly in Gemini. I’d want to understand that before telling the team our AI visibility looks good. You’ll need more than one run to see whether that difference holds. But it gives you something concrete to work with. I wouldn’t split the budget according to this chart. Visiting an AI tool once a month says little about whether someone uses it to research your product. I would use it to question how much our reporting misses when we only look at ChatGPT. Source: Datos / SparkToro, Q2 2026 State of Search. US desktop panel, April 2025–June 2026.

Do increases in AI visibility show up in branded search or Direct traffic a few weeks later? Across 632 tests on four brands, I couldn’t establish a reliable relationship. A few correlations looked promising at first, including a 0.72 between AI visibility and Direct traffic. A consistent relationship would give us another way to investigate AI’s influence when people see a recommendation and visit later, without clicking a trackable link. But the data didn't hold. I’d tested seven possible delays and kept the strongest correlation. To check whether it was unusual, I shifted the AI data against the traffic data, deliberately pairing different weeks. Three of the 19 mismatched versions matched or beat the real correlation. At that selected delay, I had only 17 usable weeks. The apparent relationship also wasn’t consistent across the two halves of the timeline. An effect could exist, but I couldn’t isolate it in this sample. I walk through the analysis in this week’s Growth Memo. If you’re running similar tests, report how you chose the delay and give the comparison data the same opportunity to find a strong result.

Hi Kopenhagen

I'd easily pay $20/month to see only tweets and comments from real name, verified X accounts.

How much AI traffic is hiding in the Direct column in Google Analytics? I looked across 12 sites and couldn’t reliably isolate AI’s contribution. Between January and September, total Direct sessions on the pages I analyzed rose 141%. But engaged Direct sessions were roughly flat! That matters because there’s a plausible explanation for missing AI referrals. I've seen this in my own studies: someone gets a recommendation, remembers the brand, and visits the website later. Finding it in our analytics would help us understand how much those referral reports leave out. But the data doesn't give that away - yet. Full write-up: growth-memo.com/p/does-ai-dr…

I found the whole experience incredibly frustrating. First, it didnt work for 20 min. Then, I couldn't apply gift cards without first redeeming them in the app store 🙄. Now, it won't arrive before January. I want to go back to @Apple , but not looking at the Z Fold.

paid subscribers of @Kevin_Indig & @thegrowthmemo just got access to the SEO Planning Workbook, a skill you can add to Codex/Claude/AG right now the smartest planning framework I’ve read with honest risk management; if you work in SEO you need it like you need tea/coffee & love

Get ready folks, direct checkout is here!

Seeing the impact of Russian bots on western democracy is making me sick.

👀

Big news! It's launch day for my AI search book. The book includes difficult stories from our time in the trenches, long-term strategies that outlive citation weather changes, and three in-the-weed case studies from Zapier, BlueVine and Siege. Order it here: amazon.com/Generative-Engine…

Cya in the morrow, Spotlight

I think about Organic Growth in ecommerce as a loop: AI Shortlist → visit → packaging → relationship → name → AI shortlist 1. First, you build visibility when AI recommends products. 1. Make the most of the visits you still get with quizzes, guides and other tools. 3. Invest in outstanding packaging to give customers something to talk about. Packaging and first use can produce the reviews, unboxing videos and word of mouth that influence future discovery. 4. Grow the relationship between purchases. Helpful service, a timely check-in or a replenishment reminder gives customers a reason to return. 5. Drive brand recall with repeat exposure, distinctive cues and strong experiences. All details in today's Growth Memo.