There was a stat from an article last week and my immediate reaction was there is no way that can possibly be true.
So obviously I kept looking.
Consumer Edge estimated that TikTok Shop accounted for roughly 2% of U.S. online retail spend in July. That exceeds the online businesses of retailers like Target, Costco and more.
That was a pretty loud signal.
Not just that TikTok Shop was important but the real signal for me was that I needed to get under the hood and understand what was actually going on.
I know enough about TikTok Shop to be dangerous. Perhaps even a little more. I spend enough time in client and industry discussions that I can usually hold my own.
But there comes a point when a trend gets big enough that knowing the headline is not good enough anymore.
When the signal gets loud enough
This connects pretty directly to what I wrote about in the first Field Note and the need to stay closer to where the work is actually happening.
With TikTok Shop, I felt like I was a fan of a sports team. I knew the players, I knew the rules, I knew the strategy, I could talk about the game but if you asked me to lace up, I’d be the worst player out there.
So I did what I do when I realize I do not understand something well enough. I found people smarter than me and started asking questions. Lots of them!
In this case, I reached out to Deren Baker and Molly O’Bryen, two people I have a ton of respect for personally and who also happen to know this space incredibly well.
I talked to Deren that morning. Molly that afternoon. There was no formal agenda, it was a Slack message asking if we could talk TikTok Shop.
The way I like to approach these conversations is to understand two things. What levers can I pull and how does that impact the recommendations I would ultimately make. Most platforms have lots of fancy stuff, my general priority is to figure out what I am eventually going to do with all of it so I usually start with a few high-level questions but from a specific lens.
Pretend I am a brand. Walk me through this thing from beginning to end. How do we win? What do I physically need to go do next? How do you know when it’s working?
I find that I learn the best when I pressure test a real scenario.
Okay, but how does this actually work?
This is one of my favorite ways to learn.
Deren did an incredible job of setting the landscape for me and updating me on the biggest factors in the space. Then it came time to go deep into the platforms and I got on the call with Molly.
I realize the likelihood that I start running the day-to-day TikTok Shop operations for a brand is highly unlikely but I do this type of thing because I want to understand how it all actually works.
I like to think about it from the perspective of the brand selling on the platform. If this was my first time launching a shop, how would I start? I was personally well versed on the product and assortment best practices so we were able to skip ahead and get into the demand generation. Shoutout to Molly here because this is where we went an entire hour over on our call.
How do I find creators? How many creators do I actually need? HOW DO YOU EVEN KNOW THAT!? How do I contact them? Wait, what!? How do we get them the product? How do we know when it’s working? How do we make that better?
I went into the conversation already having two primary mental models. I knew creator affiliates were the engine of TikTok Shop and I knew volume mattered. What I had not quite appreciated was what happens once you start finding the creators who actually work and how that scales (or doesn’t).
I learned that you have to start incredibly wide. You don’t necessarily know ahead of time who is going to sell (in fact, most creators don’t sell). The creator who looks perfect for the brand based on perceptions or what was in the brief may not be the one who actually moves product. I heard multiple stories of brands not wanting creators because they didn’t meet their perception of who the brand audience was but after results came back that they were top sellers, it became an interesting conversation. Specifically centered around the question, if sales are the goal, why not consider creators based on sales vs. perception of fit?
Once you find the creators who do work, the game took a turn I was not quite expecting.
It seems obvious now, but coming from a media and performance background, I expected something different. It really wasn’t about the slickest platform or AI tools. The difference between a good TikTok Shop program and a bad one was centered around relationship management.
Some teams move their best creators into Discord groups. They give them earlier access to products, deals and more direct communication. There are literally people now in our industry whose entire job is texting and messaging with creators. In some cases founders and brand leaders are getting involved directly in the chats and discussions.
My expectation from wanting a tech demo was that I assumed technology is what was required to find and test an enormous number of creators (that volume thing we keep hearing so much about). What surprised me was how much the game changes after you find the good ones. The clearest recipe for success also had very little to do with technology. It was more closely correlated to how human you were with your creator partners.
Then my brain went somewhere else
Once I understood that creator-affiliate engine better, I immediately started thinking about other marketplaces differently. Amazon in particular.
Amazon already has one of the largest affiliate networks in the world through Amazon Associates and Creator Connections. It has millions of products and more data points driving their commerce intelligence than anyone would practically know what to do with.
That is usually the moment I get excited. I love going deep because there are so many little nuggets, tips and tricks that are highly transferable from one specialty to another. In a lot of cases, the levers end up being pretty similar, just positioned in a slightly different way.
I set out to get under the hood a little more on TikTok Shop and left inspired by what we could potentially do to bring that to Amazon and other retailers.
Knowing enough to be dangerous has an expiration date
There are plenty of areas in my job where I know enough to be dangerous.
Honestly, I think that is unavoidable in our industry with the amount of fragmentation. As much as we all pretend to be the masters of all, nobody can possibly be an expert practitioner across commerce, media, retail, creators, AI, identity, measurement, marketplaces and every new capability appearing every single day.
The skill is knowing when surface-level understanding is enough and when it is time to go deeper.
When clients start asking the same questions over and over again, I pay attention. When other teams start asking questions around a particular topic. Or, as we saw in the case of TikTok Shop, when the growth numbers start looking legitimately insane.
That is my cue. Find the people closest to the work and get curious. Ask dumb questions and have somebody share their screen and walk you through the actual process.
If you can find the right people, don’t stop asking questions until you can fully understand what actually separates good from bad and understand the inflection points well enough that you’re confident calling something good or bad in the first place.
That is getting under the hood.
All this for what?
I am not responsible for running TikTok Shop every day. There are people who know more about the platform than I ever will. That is a bit of the point for me.
The goal of this Field Guide is not for me to pretend I have mastered every emerging part of modern commerce. It is to stay curious enough to know when something deserves a closer look.
Then go into the field. Learn from the people actually doing the work.
This week, TikTok Shop got loud enough that I wanted to look under the hood. I went in expecting the interesting part to be the technology required to manage thousands of creators. I came out thinking the competitive advantage is how you manage the human side of it once technology has already done its job.
Mike, my thoughts on your lates field note:
The "sports fan vs. player" line is memorable. That's basically the tax on this industry right now — everyone's fluent enough to sound credible in a room, but fluency and operating knowledge are not the same thing, and the gap between them is where the real risk sits. Most people never admit the gap exists, let alone go find two people smarter than them and just ask.
The part that actually surprised me wasn't the TikTok Shop stat — I'd seen versions of that number and had the same "no way" reaction you did — it was your finding on relationship management. That the unlock wasn't the platform or the AI layer, it was how human you were with creators once you found the ones who convert. That tracks with basically everything I've seen in retail media and commerce too. We spend so much energy building the tech stack and the measurement framework, and then the actual differentiator turns out to be something that scales terribly and can't be automated — a founder in a Discord DM with a creator.
And the jump to Amazon Associates at the end is the right instinct. The infrastructure's been sitting there for years; nobody's really applying the creator-affiliate playbook to it with the same intensity TikTok Shop forced brands into. That's a good one to keep pulling on.
The bigger thing though — "knowing enough to be dangerous has an expiration date" is a great way to frame the discipline of staying current in a fragmented industry. It's not about being the expert in everything. It's about having a trigger — clients asking the same question three times, numbers that don't pass the smell test — that tells you when it's time to stop pattern-matching and go get under the hood. I might steal that framing, credited of course.
Good one. Keep them coming.