You can vet an influencer in 48 hours by splitting the job in two: day one is desk research (audience fit, comment quality, engagement against a baseline) and day two is a small tracked test with a unique link or code. Make your decision on day two, not before, and never on follower count.
What predicts sales is what the audience does with a creator's posts, not how many people follow them.
TLDR: Day one: check audience fit, comment quality and engagement rate, and ask for screenshots of past analytics. Day two: run a small tracked test with a unique link or code. Commit budget only if the clicks and replies look like real buyers.
Day one: five checks from your desk
| Check | What to look at | Red flag |
|---|---|---|
| 1. Audience fit | The creator's audience location and age, which you can ask them to share | Most followers live where you do not sell |
| 2. Comment quality | Read the comments on the last 10 posts | Generic emoji, repeated phrases, or comments with nothing to do with the post |
| 3. Engagement rate vs baseline | Likes plus comments divided by followers, on recent posts | Far below the normal range, or suspiciously identical from post to post |
| 4. Past campaign proof | Ask for analytics screenshots from the last two sponsored posts | They cannot, or will not, share any |
| 5. Fit with your product | Do they already talk about things your customers care about? | Sponsored posts for unrelated brands every week |
For check 3, use a realistic baseline. In Mydrop's September 2026 data, the median Instagram post earned a 3.1% engagement rate and TikTok 3.0%. The middle half of Instagram posts ran from 1.1% to 6.7%. A creator inside that range is normal. One far below it deserves a question.
If something feels off, run the quick checks in our guide to spotting fake influencer followers before you go any further.
Day two: a small tracked test
Desk research can only tell you that the audience looks right. A test tells you whether it buys.
Ask the creator for one story or post, and give them:
- A unique link (a link-in-bio page for each creator works well, so every click is theirs alone)
- A unique discount code, if you sell directly
- A fixed time window, such as 24 hours, so you can compare the clicks with the same window the week before
Keep the first deal small. You are paying for information, not reach.
Day two checks: what to read from the results
| Signal | What it suggests |
|---|---|
| Link clicks in the first 24 hours | The audience trusts the creator enough to act |
| Code redemptions | Real purchase intent, the closest thing to a sales prediction |
| Saves and shares on the post | The content mattered beyond a tap |
| Comments asking about price, sizing or availability | Buyer questions, not fan reactions |
| Lots of likes, no clicks, no questions | An audience that watches but does not buy |
Rank them in this order. Likes come last because they are the cheapest action.
The decision gate
At hour 48, ask three questions:
- Did clicks or redemptions reach a level that would pay back a bigger fee? You choose the level, and you write it down before the test starts, so you are not making excuses for the result afterwards.
- Did the audience ask buyer questions?
- Did the creator stick to the brief and deliver on time, without being chased?
Three yeses mean a bigger deal. Two means a second small test. One or none means move on, and you will have spent very little to find that out. For pricing the bigger deal, see how much to pay influencers.
Track it in one place
Keep every creator's link and result in one sheet or dashboard. After a few tests, you can then compare creators on cost per click and cost per redemption. Mydrop's link-in-bio pages can give each creator their own trackable link, and the analytics show how the posts you publish yourself perform over the same dates.
If the test goes well and you want to run something longer, our 30-day sales prediction test is the next step.
Next step: pick two creators you are considering, run the five day-one checks on each tonight, and send the shortlist winner a request for a tracked test tomorrow.
Google review
Trustpilot review
Product Hunt comment
