Happy Thursday!

I was on stage at a founder event last week and somebody asked me why their brand was stuck at $7M. I told them the same thing I tell everyone: you've got a dentist problem. And I don't mean that in a good way. I pulled the biggest takeaways from the conversation below.

Vendor of the Week

Lebesgue - the platform watching how AI is rewriting ecommerce discovery in real time.

Here's a number worth sitting with. Across 35,000+ Shopify brands Lebesgue analyzed, sessions that came in through AI tools converted at 3.6%. Sessions from Google Search converted at 1.23%.

Almost 3x!!

That's not a tweak to your SEO strategy. That's a shopper showing up with their mind half-made-up because ChatGPT, Perplexity, or Claude already did the comparison work for them. They didn't browse. They got told what to buy.

Which means the next year of eCommerce growth isn't won by whoever ranks #1 on Google. It's won by whoever shows up inside the AI recommendations before search even happens.

Lebesgue is trained on $10B+ in tracked ad spend so it tells operators where their brand appears inside AI-driven discovery, which competitors get surfaced instead, and what signals push those outcomes.

Grab a free AI Visibility Audit for your brand now.

The $7M Ceiling (And How to Break Through It)

Here's the diagnosis I give almost every brand that comes to me stuck between $5M and $20M. You walk in and say, what's wrong? And I ask: are your flows built out? Is your retention system working? Is your landing page actually selling? And they look at me like the patient who hasn't flossed in three years.

The problem isn't the ads. It's that the ads are the only thing anyone's been touching.

Every brand that gets to $5-7M on a few strong creative wins makes the same mistake: they think the playbook that got them there will get them to $50M. So they push harder on creative, try new channels, add budget. But they've got Swiss cheese everywhere else. No real retention system. No landing page testing. No offer structure. No subscription layer. Meanwhile, they're burning $50K on programmatic display tests when they haven't even cracked a new Meta audience.

The first thing I look at is always: Meta, the website, merchandising, the offer, and what happens after someone buys. Most brands are ignoring at least three of those five.

On the landing page question specifically — test everything. Homepage, PDP, collections, advertorials, listicles, offer pages. You'll find that one audience of moms converts better on a collections page because that's how they're trained to shop. A different audience wants a listicle with 7 proof points before they'll click add to cart. The feedback loop is tight and cheap. There's no excuse not to run the tests.

On finding influencers: forget the platforms. I've done this manually for 10 years. Go on TikTok or Reels, find the person already talking about your category, whoever has genuine engagement, whoever's good at talking to a camera. Reach out. The best creator relationships I've ever built started with me shipping product to someone who didn't ask for it and watching what happened next.

Listen here: YouTube | Spotify | Apple Podcasts 

The Biggest Takeaways

• The dentist diagnosis. Every brand stuck at $5-20M has the same problem: too focused on acquisition, not enough on everything else. Flows, retention, offer structure, landing pages. Fix the Swiss cheese before you add more ad spend.

• Test every destination. PDP, homepage, collections, advertorials, listicles. One audience segment does not behave like another. The brands that win on Meta aren't guessing which page converts. They've run 20 tests and they know.

• Find creators manually, at least the first 50. Software platforms give you followers. Manual search gives you someone who genuinely loves your category. That authenticity is the entire point. Ship product cold. See who talks about it on their own.

• AI is rewriting what personalization means. The next wave isn't segments. It's one-to-one email copy based on real behavioral signals. The brands building these workflows today are setting a baseline everyone else will be chasing in 2026.

• The $5k test still works. Before you build anything, validate the signal. A few hundred dollars of ad spend tells you whether headline one, two, or three gets the click. Whether video or static moves the needle. Whether the product has legs at all.

That's all for today

It's Thursday, and it's all downhill from here. Spend the next few hours looking at your retention stack and asking yourself honestly: if I were the dentist, what would I tell this patient?

Nik