One viral video proves one video. A product worth sourcing shows a different pattern in short-form: many unrelated creators showing it, buyers using the same buying words in the comments, a benefit you can see in five seconds, and recent uploads still outperforming. This guide turns those observations into a scorecard you can fill in before you spend on samples.
What short-form can tell you, and what it cannot#
Short-form video is strong evidence for demand language, demonstration clarity, creator spread and timing. It says nothing reliable about your margin, supplier quality, return rate, conversion rate or seasonality. Treat the scorecard as a filter that decides what deserves a supplier quote, not as a sales forecast.
The seven-signal scorecard#
Score each signal 0, 1 or 2. The thresholds below are starting points, so adjust them to the size of your category.
| Signal | 2 points | 1 point | 0 points |
|---|---|---|---|
| Creator spread | 5+ unrelated creators posted it in the last 14 days | 2 to 4 creators | One creator or a single network of creators |
| Platform spread | Appears on TikTok and on Reels or Shorts | Appears on two platforms with few videos | One platform only |
| Velocity | Recent videos beat each creator’s own median | Mixed | Recent videos below median |
| Demo clarity | Benefit is visible in 5 seconds, no sound needed | Needs a voiceover to land | Benefit cannot be shown |
| Comment intent | Many comments ask where to buy, price, or fit | Some | Mostly jokes and tags |
| Objections | Top objections can be answered with a feature or a guarantee | Some can | Objections are about safety, legality or durability |
| Saturation | Few sellers, varied ad creative | Several sellers | Many sellers using the same creative |
A total of 10 or more goes on the shortlist. 7 to 9 goes on a watch list for next week. Below 7 is a skip.
How to collect each signal#
Creator spread and platform spread. Search the product’s category words and the buyer’s problem words, not just the product name. Many items circulate without a stable name. Count distinct creators, then check whether the videos show the same SKU. Look at how each video was made, too. Ten near-identical videos from creators in one agency’s roster count as one source.
Velocity. For each video, divide views by the creator’s median views on recent posts. A ratio above 2 in the last two weeks suggests the product, not the creator’s audience, did the work. Daily snapshots show whether it is still rising. Compare day-3 views across videos posted a week apart.
Demo clarity. Mute ten videos. If you can still say what the product does, it passes. Products that need a story to make sense are expensive to advertise.
Comment intent. Sample around 100 comments on each of five videos and tally them:
| Comment type | Example | What it tells you |
|---|---|---|
| Purchase intent | “link?”, “where do I get this” | Demand exists and the path is unclear |
| Price | “how much”, “is it worth it” | Price sensitivity band |
| Fit | “does it work with iPhone 15?”, “what size” | Product questions your page must answer |
| Skepticism | “does this really work?” | Proof beats you need in the ad |
| Social | “tag your friend” | Awareness and shareability, weaker buying intent |
| Joke | emojis, memes | Noise |
A product with 20% or more purchase intent and fit comments is easier to sell than one with the same views and mostly social comments. Those cutoffs are working rules, so calibrate them against products you have sold.
Saturation. Check the public ad libraries for the same product. Meta runs an Ad Library and TikTok’s Creative Center lists top ads. Count active advertisers selling the exact item, and compare their creative. If twenty sellers run the same unboxing, you will need a better angle than the product.
Worked example (illustrative)#
Take a made-up product, a magnetic cable organizer for desks. These numbers are invented to show the method.
| Signal | Observation | Score |
|---|---|---|
| Creator spread | 9 unrelated desk-setup creators in 14 days | 2 |
| Platform spread | TikTok plus a few Reels, no Shorts | 1 |
| Velocity | 7 of 9 videos above creator median | 2 |
| Demo clarity | Cables snap into place, visible with no audio | 2 |
| Comment intent | About 12% buying or fit questions | 1 |
| Objections | Mostly “does it hold heavy cables” (testable) | 1 |
| Saturation | 14 advertisers, same unboxing creative | 0 |
| Total | 9 of 14 |
At 9, it goes on the watch list. The demand evidence is strong and the saturation score is the problem. The plan is to order a sample, test whether it holds a charging cable for a laptop, and look for an angle the 14 advertisers are not using, such as a travel kit or a cable-per-device bundle. If the angle fails, the product waits.
The kill list#
Skip or investigate further before sourcing when any of these apply:
- Health and skin claims. Products that sell through “cures”, “detox” or medical outcomes carry claim risk. Videos can say things your ads cannot.
- Other people’s IP. Branded characters, logos and lookalikes.
- Regulated hardware. Batteries, chargers, heaters and anything marketed to children usually need compliance work.
- High return risk. Size-sensitive, fragile or fit-dependent items.
- Restricted categories. Check the marketplace’s restricted-products policy before you plan around it.
- Fake proof. The FTC’s final rule on fake reviews and testimonials bans fake and AI-generated reviews, and buying fake social metrics. Build proof from real customers.
Running the scorecard in ContHunt#
A Deep Search can fill most of the scorecard in one request, because it searches TikTok, Instagram Reels and YouTube Shorts together:
Find products that appear in the “made me buy it” style on TikTok, Reels and Shorts in the last 14 days in the home-office niche. For each product, count distinct creators and platforms, say whether the benefit is visible without sound, and summarize the buying questions in the comments. Rank by creator spread and recent velocity.
ContHunt returns the analyzed videos with links, so you can open the evidence behind every score. Save the shortlist to a board, then use account tracking (an experimental feature) on the two or three creators who started the trend, and check next week whether velocity held. A weekly trend radar turns this into a routine.
From shortlist to test#
Order samples before inventory. Shoot your own demo, and brief creators with the proof beats the comment analysis revealed. Judge the test by your own numbers: cost per purchase, return rate, refund reasons. The scorecard only earns the right to run it.
Get started with one Deep Search in your category and score your first three candidates today.
