I was going to tell a founder to change his company’s name. I was going to be so helpful about it!!
The company is called Lebesgue, pronounced leh-BEG.
It’s the last name of Henri Lebesgue, a French mathematician who built a better way to measure things. Things that classical math whatever that is kept missing. Real & influential things that shaped outcomes but kept getting missed.
(Their AI is named Henri. Of course it is.)
I did not end up pitching a new name.
What I DID do was spend nearly an hour talking with their founder Josip Begić, who is also a literal published mathematician, about…growth problems.
Josip spent most of his career in growth hacking, managing millions in ad spend. Turns out he became a founder for the same reasons I started Go-to-Millions:
You can see what’s happening inside your own 4 walls. But you can’t always see what’s happening in the market. Or the full complexity of the buyer journey. Not when most of the channels telling us WHERE to put our $$$ also WANT those $$$.
This is a piece about solving that problem.
About how much more we can suddenly see, and what that should change for you, right now. Not to rush you.
Let’s get smarter, shall we? 👇
(I promise, no math ahead.)
PROBLEM #1: YOUR DATA HAS AN AGENDA
When you have budget to allocate, every solution has an agenda.
A Meta agency isn’t going to tell you to dial down Meta and move to TikTok or Snap, even if they’re working better. They’re going to tell you to buy more Facebook ads.
This is structural. It Is Known. It’s literally priced into our MTAs models.
This is why none of us add up conversions from channel reporting, lest we think we’re doing 3x our actual revenue. Lol.
This was the first problem Josip & I talked about, and it’s why I decided to table my name change pitch. For the moment.
Because we’ve all dreamed about an unbiased entity whose only job is to tell you where to put your money to make more of it.
And this bias is only part of the issue!
PROBLEM #2: YOU CAN ONLY SEE INSIDE YOUR OWN HOUSE
So, you have your incrementality testing on lock, your MMM is calibrated, your measurements cleaned & 3rd-party-verified. It’s possible you’re still flying with 1 eye closed. 🏴☠️
But even this clean data from your own account only shows your own numbers, not what’s working across the market.
Like, what competitors are truly outperforming you. What and where they’re spending. Who really “owns” the category you’re in, and WHY they own it.
And you’ll never be able to understand your marketing FULLY if you don’t know what competitors are doing.
And that’s what we’re all trying to do here, no?
Watching the market. Watching the best brands. Staying close to our customer. Not being so absorbed in our own motions that we lose the plot on what’s actually working out there.
I started a newsletter about it.
Josip built bots that followed everything brands were doing.
More on what they found later.
PROBLEM #3: THE FASTEST-GROWING CHANNEL IS INVISIBLE TO YOU
AEO/GEO is a problem I know well. We all know this is urgent, in theory.
I’d already hired an entire agency to tackle this problem for me. (I need to get cited in [REDACTED] like my life depends on it.) I retagged my entire site. Built a damn Reddit strategy.
…but I didn’t have the latest numbers.
Looking at new data from 35,000+ brands set my hair on fire:
- AI-referred traffic is growing 15x YoY.
- It converts at nearly 3x vs. Google Search, with 30% higher revenue per session.
- And it is VERY POSSIBLE you have NO IDEA this is happening.
Why: GA4 currently groups AI referrals into “direct” or “other” traffic. Shopify’s native analytics don’t surface them at all. Most of the traffic doesn’t come with cute little UTMs appended.
68% of ecommerce brands already have AI-attributed purchases in their data, but because standard dashboards can’t track it yet, it’s just thrown into the spin cycle with everything else. Invisible.
Some of you might remember me sharing the full report in a panic a few weeks ago. I even roped Josip into doing a whole workshop about it for me. (June 30th btw, you’re invited, anonymous Qs welcome.)
These shoppers are different because they arrive pre-decided.
A customer asked ChatGPT for the best SPF moisturizer for sensitive skin. ChatGPT gave them a shortlist. They showed up at your store having already chosen to consider you.
The AI shopper is more qualified, more educated. We already see it on Amazon Rufus Alexa, and everywhere the data exists to look. That’s only going to get more true.
It’s officially existential to know:
- Which prompts are sending customers to our competitors instead of us?
- What sources did AI read to make that call?
- What on-and-off-site actions can we take immediately to fix it?
PROBLEM #4: YOUR SIGNALS DON’T TALK TO EACH OTHER
I have a whole stack. 1 thing for creative analytics, another for MTA, a different 1 for MMM. They’re wonderful. I stand by them.
But none of them can show me the level of competitor intel I want, and give me the steps I need to take to see & change the future. 🔮
I want to know how they’re ranking in different AI engines vs. me, and why.
I want to know what they’re spending, where, and how it’s performing.
What they’re spending. How their CX compares to mine.
What they’re saying to MY customers, and whether it’s working.
An unbiased 3rd party that talks to everything on the internet and actually captures how fucking complicated the buyer journey really is.
All this is what’s being built.
Even more interesting than the AI visibility data was the feature they’re releasing soon: industry benchmarks that compare your blended CAC, MER, and core metrics against brands with the same AOV and spend level as yours. YES
OUI, HENRI
Henri Lebesgue (the dead mathematician, RIP 1941) basically said: classical measurement only works on clean shapes. The real world is messy. We need better instruments.
Lebesgue (the company) was founded by someone who felt lied to by his own marketing data for long enough that he built a company about it.
Henri (their new AI, keep up), connects attribution, MMM, competitor intelligence, AI visibility, and market benchmarks into a single interface you can actually ask questions of. Built on observed performance data from 35,000+ ecom brands and did I mention $10B in ad spend.
I came in to that call ready to bully them about the name. I left a Henri Lebesgue truther.
4 THINGS TO CHECK THIS WEEK
♟️: Pull your AI-referred traffic. If GA4 is showing you “direct” as a suspiciously large channel, some of that is almost certainly AI traffic being miscategorized. Start looking.
♟️: Open ChatGPT, Claude, Perplexity, and Gemini. Ask them each to recommend the best product in your category. Are you on the list? Who is? For which sources?
♟️: Dig into your Google sessions for AI-referred traffic, and compare conversion rates. It’s very different by category.
♟️: Think about what would make you more citable off-site. High-authority editorial. Reddit threads. Expert reviews. Press. The stuff that becomes someone else’s training data.
This should be enough to start understanding the outlines of where you stand.
NEW HORIZON, INCOMING
We always knew the buying journey was more complicated than our data showed. But in many ways, we had to build our strategies around what we couldn’t see or track. Try to react more than we anticipate. Fantasize about a crystal ball. Just me?
The good news is the ceiling on what’s measurable just went way up.
We can start to see all the many threads that play into every purchase. The paid channels, the organic ones, the intel across the board, the growing cohort of pre-considered AI shoppers. We can see that every touchpoint matters, and stop using “omnichannel” like just a buzzword.
How do you, as the marketer or CMO or chief digital officer (hello), make sure you’re ahead of it?
Well, I’d love to help you see around the corner.
- First up, come hang out with me and Josip on June 30th for our AI Visibility workshop. I’m bringing the smartest growth person I know to join us, too.
- If you haven’t already, grab the free 2026 AI Discovery and Ecommerce Visibility Report.
- Plug Lebesgue into Shopify and see what you learn.
The window of opportunity here is both real and shrinking. Let’s hop on through it together.
Yours from the future,
Ari
