51,288 founders and marketers are getting this newsletter today.
Welcome to the 295 new operators who joined us this week! 🤯
Hey Fastlane Insiders! 👋
Across 90 days of tracked Shopify category questions, AI search cited YouTube 10,640 times. My own site, after a decade of publishing, got 4,847. Better than two to one against me. The strange part is where that gap actually lives, because on one engine the whole result flips.
Every brand I talk to is running the same two experiments right now. Working out which AI surfaces actually send buyers, and what to do with the tools already in the stack. Shopify quietly shipped the outbound half of its MCP protocol, which connects Klaviyo, Triple Whale, and your survey data straight into Claude or ChatGPT, and almost nobody has touched it. Meanwhile, TikTok Shop keeps driving revenue nobody's dashboard is built to see.
Here's what's inside:
🎧 Podcast: The TikTok Shop halo effect, with Joybyte president Myriah Castillo.
💡 Knowledge Drops: The outbound half of Shopify's MCP nobody's using, and why YouTube beats your site in AI search.
🔥 Tool of the Week: Redo Recover, for the 70% of abandoned carts your SMS flows never reach.
📡 Industry Pulse: new ecommerce tools, DTC growth data, and an omnichannel framing move.
Let's dive in. 👇
Your next product photoshoot is five words long.
Richer listing content is worth up to 8% more sales.* Building it used to mean a photographer, a studio, and half a day you didn't have. Now you type "create three candle product photos", and Agentic AI writes the full prompt, picks the model, and sizes each shot. Ask for more and it makes more, until the listing is full.
* Source: Amazon, A+ Content, accessed August 2026
Use code FASTLANE for 25% off any subscription, up to 44% if you get yearly.
🎧 New Podcast Episode! 🎧
The TikTok Shop Halo Effect Nobody's Measuring
Your CAC keeps climbing, and the old playbook of handing the budget to a few creators and hoping is no longer bailing anyone out. Most Shopify founders can't tell you whether creator spend actually moved revenue. They can only tell you what the dashboard says happened inside the app.
This week I sat down with Myriah Castillo, President of Joybyte, one of the fastest-growing TikTok Shop certified agencies in the country. Myriah built and sold her own guacamole CPG brand before running creator programs for more than 30 brands, from challenger startups to names like Eggland's Best and Duluth Trading Company. One launch her team ran drove more than 25,000 TikTok Shop orders.
A few things we got into:
Creators are a signal layer, not a reach play. A creator's existing audience tells you who your real customer is, often before you spend a dollar on paid.
The stoplight system that decides what earns ad spend. Three signals in the comment section tell Myriah whether to scale a creator, wait, or walk away. Most brands read all three wrong.
The halo nobody's dashboard shows. Her team tracks a 20 to 40% lift on Amazon sales, plus website spikes, every time a TikTok video takes off.
Where you start depends on your stage. Early founders use nano and micro creators one way. Mid-market brands use whitelisting another way. Both beat the thing most brands are still doing by default.
If you're judging TikTok Shop purely on in-app sales, you're undervaluing the channel that's quietly feeding the rest of your funnel.
LISTEN NOW → Hear the full stoplight framework and the whitelisting play Myriah says most brands leave on the table.
💡 Knowledge Drops of the Week 💡
The Half of Shopify's AI Protocol You're Actually Supposed to Use
Every readiness checklist circulating this year is selling merchants on work Shopify has already done for them. The shopper-facing half of MCP switched itself on back in March. There is nothing left to buy your way into.
The half nobody's talking about points the other direction: connecting the tools you already pay for, Klaviyo, Triple Whale, Fairing, straight into Claude or ChatGPT so you can ask questions across your own stack in plain English. A question that used to cost your analyst two hours now takes about ninety seconds, and none of it requires you to be on Shopify. If you want the wider context first, here's how the three shopping protocols relate to each other. This one is about what MCP does for you rather than to you.
Three things worth knowing before you connect anything:
One connection is not a test. A single tool just gives you a faster dashboard. The payoff starts at the second and third, when you get answers no single vendor's reporting can produce on its own.
Start read-only on everything. There is no version of an agent misreading a variant against live inventory that you want to live through. Two of the three tools make this easy. The third needs you to set it up deliberately.
Your sample size will lie to you. A model asked to rank five discovery channels will confidently rank them based on eleven responses. There's a one-line guardrail you can build into the question itself to stop it.
The honest trade-off. This is a distraction for anyone under $500K. Fix your product data first. Above that, the stack audit only pays off if you pick one recurring question and run it the same way every week.
READ THE FULL BREAKDOWN → The full stage-by-stage rollout, from $500K to $10M and beyond.
Why YouTube Is Beating Your Site in AI Search (And What to Do About It)
A promotional podcast trailer I uploaded years ago and completely forgot about is now one of the most cited YouTube videos in my entire category. That is not a win. That's an open door with the wrong thing standing in it.
Across 90 days of tracked Shopify category questions, YouTube pulled 10,640 AI search citations to my own site's 4,847, a gap of better than two to one. The number that actually matters, though, is which engine you're testing on. 91% of those YouTube citations came from just two of them, and on a third my own site beat YouTube by nearly three to one. Ahrefs' monthly index shows the same pattern holding at web scale, so this isn't a Shopify quirk.
Here's what the data says is working:
AI cites procedures and comparisons, not opinions. Every one of the ten most cited videos in my category is a how-to or a versus breakdown. Not a single founder interview made the list out of 482 published episodes.
Views don't predict citation. Structure does. Four in ten cited videos had under 1,000 views. Two structural signals did the work instead, and both take minutes to add to a video that's already live.
Chapters multiply one upload into several citable units. Timestamped videos get cited across two to five different chapters, and YouTube's spec for making that work is three lines long and free.
The trade-off worth naming. Test your five highest-intent questions on the two engines that matter before you touch a camera. I got into the methodology behind this data on a past episode with Searchable's product lead.
READ THE FULL BREAKDOWN → The engine-by-engine split, and how to fix video you've already published.
🔥 Tool of the Week 🔥
Your SMS flows reach 30% of carts. Recover gets the other 70%.
Your abandoned cart flows only text shoppers who opted into SMS marketing. That's about 30% of your carts. The other 70% never subscribed, so your flows never reach them.
Recover does. Trained agents text those shoppers one-to-one in real conversations. They answer sizing questions, handle shipping objections, send a code, and win the sale back. No automation, no templates, no blasts.
It runs alongside the flows you already have. Nothing to rebuild, nothing to rewrite, five minutes to install on Shopify.
You only pay for results, backed by a 10x ROI guarantee.
⚡ This Week’s Industry Pulse ⚡
A few updates actually move the needle. Here's what made the cut. 👇
A commodity category just got a beauty makeover: Oral care brand Boka posted 481% year-over-year retail growth and took the number-one Amazon spot in its category by borrowing the aesthetic playbook from skincare rather than playing the usual fluoride-percentage, dentist-endorsement game. If you're stuck competing on commodity specs, the unlock is often sitting in a category you'd never think to copy.
Retail media just got three new dance partners: GrowthLoop embedded The Trade Desk, Moloco, and Audience Acuity directly into its composable commerce media network this week, so retail teams can run on-site and off-site activations through one connected stack instead of stitching partners together themselves. The tooling layer behind retail media, not just the networks themselves, is where the next round of competition is happening.
A big-box chain just got an unusual retail partner: Michaels launched an omnichannel custom framing experience with Architectural Digest, pairing a craft retailer's in-store service with a design magazine's editorial authority to sell a purchase most shoppers would never associate with a big-box store. Borrowed credibility, built fast.
Before September Gets Here
This is the last Thursday of August. Next time this lands, it's September, the quarter is nearly done, and the calendar stops asking politely.
I like this particular week of the year. It's the last one where the work is still yours to shape. Nothing is on fire. You can still change your mind about almost anything, and it costs you a conversation. Four weeks from now, most of what you do will be a reaction to something you already decided.
So do less than you think you should this coming week. Pick the one thing you've been meaning to look at properly and go look at it. Not a project. A look.
And thank you for the ten minutes. There are nearly 50,000 people opening this newsletter today, at wildly different stages of this, and most of you had somewhere else that time could have gone. I haven't stopped noticing that.
Hit reply and tell me what you're heading into September with. I read every one, and they're most of where this thing comes from.
Tail end of summer here in BC. I'm trying to enjoy what's left of it, and I hope you get some too.
Cheers!
Steve
