Hi Reader,

It's Josh from AmpliSell.

Last Friday I was on a call with a senior e-commerce leader at a global brand. He said:

How we measure success on TikTok is the BIG question right now, so let's answer it.

First, it's not Amazon or DTC.

At AmpliSell we've run the Amazon channel for brands for almost a decade. Success there is clear and easy to measure: revenue, profit, TACOS, MER, etc. Ad spend is directly attributable and customers are on there because they want to buy. Amazon is generally profitable and scalable with the right playbook (humble brag: 93% of our clients are up YoY #callme).

TikTok is completely different. People are there to be entertained. Buying is an afterthought.

But that doesn't mean it isn't a powerful revenue-driving channel.

You just have to know what to measure.

And how.

Let's start ON TikTok and expand out from there.

GMV is the obvious score, but that is an output.

Let's take a look at the chain of purchase and (inputs) that drive it:

Creators make videos (video count). People watch videos (video views) and buy products (GMV). Ads (ad spend) make the videos appear to more people (video views) and more people buy.

Or to put more simply:

We need A LOT of videos to sell products, we need A LOT of people to see those videos, and the videos have to be compelling.

*I'm ignoring LIVES for now, because most people aren't at that point, but if you're there, email and I'll share what we track.

So we track every link in that chain:

Affiliate performance: videos per sample sent and GMV per sample. This tells if you creators are posting videos and if the videos are actually selling. BONUS: Track L3 new content -> this is your leading indicator of GMV. The more your top creators pump out the more GMV you get. Read last week’s episode for a strategy to get more and better videos out of your creators.

Now we have an idea of what is driving GMV on TikTok and we can pull the respective levers to improve it.

But as a channel, TikTok PNL is going to look terrible next to Amazon or DTC.

You have to send free products to creators (cogs + shipping), convince them to post a lot about it (time management), most of them suck (wasted cogs + shipping), and the ones that are good want a lot of money (commission rates + incentives). On top of that you have to run ads so their videos show up more.

You face A LOT of costs just to drive a TikTok sale.

So why are we investing in TikTok?

Let’s expand out from TikTok and look at the Halo.

Because TikTok PNL is missing most of the revenue.

Videos on TikTok drive people to buy on Amazon, DTC, and in-store.

You have 3 additional channels capturing demand from TikTok.

Connor at Ridge, the guys that sell $70 wallets (in this economy?!), shared geo-lift study results recently. 88% of the lift showed up outside of TikTok. 49% on Amazon, 39% on DTC, and just 12% on TikTok Shop. Smells like marketing!

Haus analyzed hundreds of incrementality experiments across brands and found TikTok delivers roughly 1.9x the value when you measure omnichannel impact. For brands selling across DTC, Amazon, and retail, every $1 of TikTok lift came with an extra $0.34 on Amazon and $0.57 in physical retail (Target, Sephora, Ulta, Best Buy).

Two more Haus findings worth knowing:

62% of TikTok-driven lift hit new-customer KPIs. TikTok is finding buyers these brands were missing everywhere else. This is why I keep harping on it as a channel. Gen Z is on here and discovering new products. If you're not on here, you're missing out on the next generation of buyers. [end soap box].

The lift keeps building after spend stops. Brands saw 68% additional lift in the post-treatment window, amplified even further heading into promo periods.

AmpliSell Data:

Our clients are averaging a 38% lift on Amazon and 22% lift on DTC from TikTok Shop. Branded search on Amazon ALWAYS goes up with TikTok video views.

Measuring the halo

First, some of you guys are big enough to run geo-lift studies, some aren't. I'm going to share practical and easy ways to measure halo that don't require an agency - it won't be exact, but it will be strong evidence that justifies investment.

Your Free Signals

1. Build a Simple Graph. One chart, four lines: TikTok video views, Amazon branded search, Amazon ordered revenue, DTC revenue. Amazon & DTC typically rise within 2 to 3 days of TikTok video view spikes and branded search lift on Amazon usually lags consistent posting by 5 to 14 days, with the correlation becoming clear after roughly 60 days of 3 to 5 posts per week.

We use Cruva to do this for our clients. It's easy to explain, shows multiple correlations, tracks Amazon and Shopify, and gives you a pretty graph which makes it easy to share. Look at the July 11 and July 30 spikes driving Amazon revenue for this client.

BONUS: Every time a video pops, snapshot the 7 days before and after: Amazon sessions, branded search, DTC sessions, total revenue. You can see 2 or 3 in the chart above. Five spikes with the same lift is evidence.

Your Measurable Inputs

Amazon Attribution + Brand Referral Bonus. You can tag every link you control that points to Amazon: your TikTok bio, your link-in-bio page, creator bios where you place the link. Attribution is free for brand-registered sellers and tracks clicks through to purchases on a 14-day window.

Enroll in the Brand Referral Bonus and Amazon pays you back an average of 10% of those attributed sales.

You get a floor for your Amazon halo and you get money back from Amazon.

Note: this measurement does impact the outcome because we're actively directing to Amazon.

BONUS: Ask the customer. Add a post-purchase survey. "Where did you first hear about us?" with a TikTok option. Knocommerce does this (no affiliation) and shared results from one of their clients "surveys revealed that TikTok was an underreported but significant acquisition channel, with 32% of new customers coming from TikTok."

Your Definitive Test

Halo effects can vary wildly as you saw above with the Haus and Ridge studies.

Let's narrow down yours with a test. This the most time consuming, but delivers a definitive result.

The idea is simple. You can't rerun the last 90 days without TikTok to see what would have happened. So you use the next best thing: products you never put on TikTok Shop. They lived through the same season, the same promos, the same economy. The only difference between them and your TikTok products is TikTok. So if your TikTok products pull ahead on Amazon and DTC, TikTok did that.

Here's how to run it:

Split your catalog. Products on TikTok Shop are your test group. Similar products you kept off are your control group. Same seasonality, same promo calendar, similar buyer. Launching soon? Hold a few products back to create a control group.

Clean the data. One product, one name, one SKU across every channel. Each product needs sales history from before TikTok. Start measuring once videos go live, not when you create the listing.

Do the subtraction. For each group, add up Amazon and DTC sales only. Leave TikTok GMV out. You already know what TikTok sold. You're hunting for what TikTok caused everywhere else. Compare equal windows before and after, at least 4 weeks each. Test group change minus control group change is your halo.

Example: your test products' Amazon plus DTC sales went from $100K to $130K a month (+$30K). Your control products went from $100K to $110K (+$10K). That $10K gain is your baseline. Your halo from TikTok is the $20K.

Test it. Swap different products into the control and re-run. Subsequent tests strenghten the evidence. Just remember at some point we're here to make money, not test all day, so accept the results and move on.

Read the result. Your halo multiplier is your halo revenue ÷ TikTok GMV. In the example: $20K of halo on $10K of GMV = 2.0. Every $1 TikTok reported came with $2 more landing on Amazon and DTC. Same math as the Cruva number above, measured on your own brand.

Your NEW Halo Adjusted PNL

You can run two PNLs for TikTok.

The first is on-platform only. GMV minus costs = profit.

The second applies your multiplier from your tests to GMV.

Until you've run one, you can use the benchmarks from this email (Haus at ~1.9x omnichannel, our clients at +38% on Amazon and +22% on DTC) as placeholders.

When you outgrow all this, graduate to geo-lift studies.

Until then, a bunch of signals pointing the same direction beats a perfect study that was never run. Trust me on this one as I agonize over every sentence in this newsletter instead of just sending it.

So, back to the question

Sales or marketing? Revenue channel or marketing spend?

Yes.

TikTok is a revenue driving channel that behaves like marketing. Ridge’s study put just 12% of the lift on TikTok itself. The other 88% landed on Amazon and DTC. Measure TikTok on platform GMV alone and you will conclude it loses money, underinvest, and hand the next generation of your customers to whichever competitor measured it properly.

Here's what you do

On TikTok, manage the inputs. Videos per month, video views, GMV per sample. Video views will drive GMV, especially video views from your top creators. Keep the videos pumping!

Off TikTok, measure the halo. Branded search, cross channel revenue correlations, spike windows, and post-purchase surveys.

TikTok looks like an expensive experiment when measured by itself.

But when you measure the omnichannel impact, it’s the best way to buy your next five years of customers.

We help ecommerce leaders who don't have the time and team to scale TikTok Shop.

• Sampling and Creator Community Building
• Content Strategy
• Advertising Campaigns
• LIVE stream operations & scheduling
• Halo sales tracking
• Full-funnel analytics & optimization

We plan, execute, and deliver omnichannel growth.

Book a call here -> https://calendar.app.google/SjUfGyLrPdH9g7fS9​
​
Cheers,