TikTok as a Demand Generator vs Commerce Platform

Hi there,

One comment from a recent CMOWire interview has followed me into countless conversations with brand leaders since.

The quote was simple: "If you are only calculating the success of TikTok based on TikTok Shop sales, that is not thinking holistically."

The response I kept getting wasn't pushback. It was recognition. Like: yes, I've been feeling this too, I just didn't have the framing for it. So let me give you the fuller version of what I meant.

The pullback news is being read wrong

When Grande Cosmetics and a handful of other first-movers scaled down their TikTok Shop operations this year, the trade press covered it as a cautionary tale about TikTok's retail ambitions. And yes, when TikTok announced a mandatory fulfillment shift in early 2025, it forced brands to run the math on their TikTok Shop economics, and many found the unit economics didn't hold up. TikTok ultimately paused the mandate in February 2026 after seller backlash. But the math those brands did? That didn't change.

But here's what's getting lost: these brands aren't necessarily saying TikTok doesn't work. They're saying the checkout surface isn't worth the operational trade-off. Those are very different conclusions.

When a brand pulls back from TikTok Shop, it doesn't mean TikTok stopped moving their product. It means they stopped selling through TikTok's checkout flow specifically. The content that drove people to search, to go to Target, to add something to an Amazon cart -- that machine didn't stop running just because the Shop tab got turned off.

The misread happens when brands conflate the platform's sales surface with the platform's marketing value. TikTok Shop underperforming on unit economics is a logistics problem. TikTok underperforming as a demand channel would be a fundamentally different problem. Very few brands have the data right now to tell which one they're actually facing.

The number that reframed it for me

Here's the data point I keep returning to. Haus published research showing that brands measuring TikTok only by in-platform attribution are undervaluing its contribution by nearly 2x, based on their analysis of hundreds of TikTok incrementality tests. The actual downstream impact includes a 34% lift in Amazon sales and a 57% lift in physical retail -- none of which shows up in your TikTok dashboard [1].

The 2x multiplier isn't a small rounding error. It's the difference between a channel that looks like it's barely breaking even and a channel that's actually one of your best demand drivers.

What makes this hard is that the attribution gap is structural, not a reporting tweak you can fix in your dashboard settings. When someone discovers your product on TikTok and buys it at Walmart a week later, TikTok's algorithm doesn't know that happened. It didn't get the signal. It can't optimize future campaigns toward that outcome. From the algorithm's perspective, that viewer saw your content and did nothing.

That's the machine learning problem hiding inside the measurement problem. You're not just miscounting revenue. You're training TikTok's algorithm on partial data and then wondering why performance plateaus.

What I'm watching for in the brands getting this right

The brands that are going to win at TikTok over the next 18 months will be the ones that close the loop between offline demand and digital signal.

In practice: when someone registers a product post-purchase, scans a QR at shelf, or engages with packaging in any retail environment, that interaction becomes a first-party signal flowing back into their TikTok ad account. TikTok's algorithm then knows its content drove a real purchase, and it can optimize toward more of that.

Skullcandy became the first brand to build this through Brij, going live this month. It's early. But it's the right architecture. Every offline purchase moment becomes training data for TikTok's optimization engine. The shelf becomes a signal source, not a black box.

The broader point: we're in a phase where brands that treat first-party data capture at retail as a paid media strategy -- not just a CRM initiative -- will have a fundamentally different relationship with TikTok's algorithm than the ones who don't.

The TikTok question isn't "should we be on TikTok Shop?"

It's: are we giving TikTok the data it needs to know what it's actually driving for us?

Most brands aren't. Yet.

Talk soon,

Kait

A few extras:

Some of my favorite brands are hiring…

1. Glossier - Creative Director
2. OLIPOP - Director, eRetail & Retail Media Networks

References:

[1] Haus, "The TikTok Report" -- https://www.haus.io/blog/the-tiktok-report

[2] Beauty Independent -- https://www.beautyindependent.com/tiktok-shop-reverses-course-fulfillment-change-brand-backlash/