Short answer: there is no official number. Post enough different videos per product that one outlier cannot decide the result, and pick that number before you start. Many operators work with somewhere between three and ten videos per product; treat that as an assumption to tune against your own results, not a rule. More variation between videos matters more than more copies of the same idea.
Any page that tells you "exactly 7 videos" is guessing. Here is how to reason about it for your own store.
Why one video is not a test
Short-form reach is uneven. The same product can get very different views from two videos posted a day apart, for reasons that have nothing to do with the product: the hook, the first frame, the time, who happened to see it first. One video tells you about that video. Several videos with different hooks and formats start to tell you about the product.
Think of each video as one draw. The fewer draws, the more a single lucky or unlucky one swings your conclusion.
Choosing the number
| Situation | Videos per product (assumption) | Why |
|---|---|---|
| Quick screen of many products | 3 | Enough to catch a clear dud or a clear signal |
| Standard test | 5 | Room for a few hooks and two formats |
| Promising product, mixed results | 8–10 | Separates a good angle from a good product |
| Retest after a change (price, page) | 3 | You are checking the change, not the product |
Adjust with three questions:
- How different are your videos? Five near-identical clips count as roughly one. Five different hooks and formats count as five.
- How much time do you have? A test you cannot finish is worth nothing. Batch the videos first; see batching.
- How many accounts fit the niche? Spreading videos over two niche accounts shows whether a result depends on one audience.
Spread across accounts and days
| Day | Account A | Account B |
|---|---|---|
| 1 | Video 1 (problem hook) | — |
| 2 | — | Video 2 (result hook) |
| 3 | Video 3 (unboxing) | — |
| 4 | — | Video 4 (photo slideshow) |
| 5 | Video 5 (demo) | — |
Staggering keeps each account's feed varied and stops the test from resting on one posting slot. If both accounts post about the product, give each its own angle, as covered in one product, many accounts. Photo posts are a cheap way to add variation; see TikTok slideshows for products.
Decide the read-out rule first
Before the first post, write the rule you will use: for example, "the product passes if at least two of five videos beat my median link-click rate." Your median comes from your own previous tests. Without a rule set in advance, it is easy to keep posting until something looks good. The full test design is in organic product testing.
Posting the test videos with Farmero
When a test means dozens of videos across several accounts, the posting is the tedious part. Farmero posts video and photo slideshows to your TikTok accounts from your own iPhones, confirms each post on the profile, and warns you if the same clip is about to go to the same platform twice. Automating these apps may breach their terms, and nobody can promise an account will never be restricted. See posting.
FAQ
Is there an official number of videos to test a product?
No. TikTok does not publish one. Pick a number you can produce with real variation and keep it the same across products so results compare fairly.
Should the test videos all be different?
Yes. Change the hook, format or length between videos. Reposting the same clip tells you little and unoriginal reposts are not eligible for the For You feed.
When should I add more videos to a test?
When results are mixed, such as one strong video and several weak ones. Retest the strong angle with new videos before calling the product a winner.
Can I test more than one product per account at a time?
Yes, if they fit the account's niche. Keep a log of which video belongs to which product so you can read the results cleanly.