TikTok metrics: how to analyze organic, ads and Shop

A cosmetics shop in Recife, Brazil, posts an unboxing video on a Tuesday night. By Wednesday morning it has 480,000 views. The team celebrates in the group chat, someone screenshots it, and the owner asks how many sales came from it. Silence. Seven orders, and three of them came from customers who already bought through Instagram.

That scene plays out at a lot of companies that joined TikTok in the past two years. The problem is rarely the content. It’s the reading: TikTok’s dashboards show dozens of numbers, and the most visible ones are exactly the ones that explain the least about revenue.

The TikTok metrics that matter depend on which of three layers you’re analyzing: the organic profile, ads or TikTok Shop. Each layer has its own question and its own decision metric. Blending the three is the fastest way to reach the wrong conclusion.

Short answer: to analyze TikTok metrics, split them into three layers. For organic, look at retention (average watch time and the share who finish the video), shares and traffic source. For ads, look at hook rate, cost per result and ROAS. For TikTok Shop, look at GMV net of returns and margin after fees.

Why TikTok needs a different kind of reading

On Instagram, a good part of reach still flows through followers. TikTok works differently: the For You feed pushes each video to people who have never heard of your brand, tests their reaction and decides whether to widen distribution. Followers matter less; viewer behavior matters more.

And the audience is too big to treat as a side channel. According to DataReportal’s Digital 2026: Brazil report (published November 2025, based on the platform’s own ad tools), TikTok had 131 million users aged 18 and over in Brazil at the end of 2025, and its ads reached 80.3% of the country’s adults.

In practice, that means views on TikTok are cheap and abundant. If you’ve read our guide to Instagram metrics, the rule here is even stricter: reach numbers say little, interest signals say a lot.

The three layers and what to watch in each

Before opening any dashboard, be clear about which question each layer answers. Here’s what we usually recommend:

Layer	Question it answers	Decision metric	Treat with suspicion
Organic	Does the content hold the attention of people who don’t know me?	Retention, shares, saves	Total views, followers
Ads	Am I buying attention and sales at a cost that adds up?	Hook rate, cost per result, ROAS	Reach and CPM on their own
TikTok Shop	Do in-app sales actually make money?	Net GMV, product conversion, margin	Gross GMV

Which organic TikTok metrics actually matter?

Retention: the number the algorithm watches too

Average watch time and the percentage of people who watched the full video are the most honest organic metrics. They show whether the video holds people who landed on it by accident. A 30-second video with a 4-second average watch time was, for most viewers, a 4-second video.

The retention chart in each video’s analytics is where the answer shows up. Compare two videos on the same topic and the difference almost always sits in the first few seconds.

Shares and saves are worth more than likes

A like costs one tap. A share means someone put their own reputation behind your video; a save means they plan to come back. For brands, these two signals tend to predict branded search and profile visits better than any heart count.

If you want an engagement rate, calculate it on views, not followers: (likes + comments + shares + saves) ÷ views × 100. A video with 40,000 views, 2,100 likes, 90 comments, 260 shares and 350 saves has a 7% engagement rate. But notice that shares and saves are only 22% of interactions; track that slice separately.

Traffic source and search

Video analytics show where views came from: For You, followers, profile, search, sound. A video that still gets search traffic weeks after posting is answering a real question, and that’s the kind of content worth turning into a series. A spike that is 100% For You and dies in 48 hours is entertainment, not an asset.

How to read TikTok Ads metrics without fooling yourself

In Ads Manager, the first thing to get right is what counts as a view. According to TikTok Ads’ video play metrics documentation (updated November 2025), a basic video view counts any playback start. There are separate metrics for people who watched at least 2 seconds, at least 6 seconds, and the 25%, 50%, 75% and 100% marks. Comparing ad “views” with organic “views” means comparing different things.

Two derived ratios cover most creative analysis:

• Hook rate: 2-second views ÷ impressions. It tells you whether the opening stops the thumb.
• Ad hold rate: 6-second views ÷ 2-second views. It tells you whether people stayed once they stopped.

Here’s a worked example (in Brazilian reais). A campaign spent R$3,000 and got 200,000 impressions (R$15 CPM). It had 52,000 two-second views (26% hook rate) and 15,600 six-second views (30% hold rate). It drove 1,800 clicks (0.9% CTR) and 45 purchases, a cost per purchase of R$66.67. With an average order value of R$180, revenue was R$8,100 and ROAS was 2.7.

Does that ROAS pay the bills? It depends on margin. If the product has a 40% contribution margin, break-even ROAS is 2.5, and the campaign is right at the edge. The full math is in our piece on how to calculate ROAS. The point of the example is that a cheap CPM and lots of views tell you nothing about that answer.

A common trap: the conversions TikTok reports use the platform’s own attribution window. Someone who buys days later through Google may show up in both reports. Use UTMs and compare against what actually landed in your store before scaling budget.

TikTok Shop: GMV isn’t revenue, and revenue isn’t profit

TikTok Shop launched in Brazil in May 2025 and grew fast. According to the first-year update TikTok published in June 2026, average daily GMV grew 102-fold since launch, the average daily number of live streams grew 20-fold and GMV from live streams rose 161-fold. In an in-app survey of 17,000 users, 57.8% said they complete the purchase in TikTok Shop itself after discovering the product on the platform.

These are the company’s own figures, so read them as a sign of direction, not as a benchmark for your store. The practical point is different: GMV (gross merchandise value) includes cancelled, returned and unpaid orders. To make decisions, you need three cuts:

1. Net GMV: GMV minus cancellations and returns.
2. GMV by source: your own videos, live streams, affiliates or the showcase. Each has a different cost.
3. Margin per order: price minus product cost, platform commission, affiliate commission, subsidized shipping and coupons.

A simulation with hypothetical rates: a R$100 product, R$40 cost, 12% platform commission, 15% affiliate commission and R$8 of subsidized shipping. That leaves R$25 per order. If 10% of orders come back, real margin drops to around R$20, depending on who pays return shipping. An affiliate who sells a lot with aggressive coupons can be generating impressive GMV and negative margin.

How to combine the three layers into one decision

Splitting the layers is how you measure properly. Combining them is how you decide. Every TikTok analysis ends with one question: where does the next dollar earn the most margin?

In practice, that means matching TikTok against what lives outside it: store revenue, CRM and other paid channels. This is where many companies get stuck, because every piece of data sits in a different dashboard. The data silos problem hits hardest in new channels.

What to ignore (or read with suspicion)

• Total profile views: one viral video distorts the whole month.
• Followers: on TikTok, they explain little about distribution.
• Likes on their own: a weak signal, easy to inflate.
• Gross GMV: without returns and fees, it’s a shop-window number.

These numbers aren’t useless. They work as context, but they become vanity metrics when they sit alone at the top of the report.

A 30-minute weekly routine

1. List the week’s 5 organic videos and compare average watch time and shares. Note what the top two do differently in the first seconds.
2. In Ads Manager, sort creatives by hook rate. Pause the ones that have been below the account average for more than seven days.
3. Check cost per purchase against break-even ROAS, not against last week.
4. In the Shop, pull net GMV by source and see which affiliate or live stream has negative margin.
5. Write down one decision. If the analysis didn’t end in a decision, it was just reading.

Frequently asked questions

Where can I see TikTok metrics?

In the app, with a business or creator account, metrics live in the Analytics area of your profile tools; on desktop, in TikTok Studio. Ads have their own reports in TikTok Ads Manager, and the Shop has a data dashboard in Seller Center. That’s three different places.

How do I calculate engagement rate on TikTok?

Add likes, comments, shares and saves, divide by the video’s views and multiply by 100. Calculating it on followers makes little sense on TikTok, because most views come from people who don’t follow the profile.

What are the main TikTok Ads metrics?

For creative, hook rate (2-second views over impressions) and retention to 6 seconds. For the business, cost per result, conversion rate and ROAS compared with the break-even ROAS of your margin.

How do I see TikTok Shop metrics?

In Seller Center’s data analysis area, with GMV, orders, conversion by product and performance by source (video, live, affiliate). To make decisions, subtract returns and fees; the dashboard shows gross figures.

TikTok doesn’t need one more dashboard

The people who analyze TikTok well aren’t the ones who look at more numbers. They know which question each layer answers and match the answer against cash. The hard part is almost never having the data; it’s pulling together what sits in the app, in Ads Manager, in the Shop and in the store. Sherlok does that matching: you ask in plain language, for example “which channel had the best margin per order this month?”, and get the analysis ready, using data from your connected sources.