We spent on influencers and got nothing back

The videos people find when they search this have titles like "We Trusted Influencer Marketing, Then This Happened."

They are all about the same moment. The reel went up, the numbers looked fine on the screenshot, and nothing happened at the other end. The reel went up, the numbers looked fine on the screenshot, and nothing happened at the other end.

The pages that rank alongside those videos are agencies explaining why you should have hired an agency.

So here is the opposite. This is the protocol Atvis Media runs before a creator gets booked, including the signals that stop a booking, because an agency that will not publish its disqualification criteria has not got any.

One thing first. Vetting is disqualification.

It is not a process for finding reasons to book somebody. It is a process for finding the reason not to, as early and as cheaply as possible, because every check you skip at shortlist stage gets paid for later at campaign rates. Run it in this order, because the first two checks kill more shortlists than the last five combined.

Check 1: does the audience live where you sell

Ask for a screen recording of the creator scrolling their own insights. Country and city. Last 30 days and last 90.

A recording, not a screenshot. Screenshots are pictures and pictures are editable.

Read it against your own map. A clinic with one address is buying a radius, not a follower count, and a 15,000-follower creator whose audience lives on the right side of town beats a national account at ten times the size.

If you ship nationally, you still want a city spread that looks like a plausible customer base rather than a list of places your product has no reason to reach.

Check 2: does engagement match the follower count

Follower-to-engagement mismatch is the oldest signal and still the most useful, as long as you read it as a question rather than a verdict.

Low engagement on a large account has honest explanations. An older audience that drifted, a niche where people save and share instead of liking, a run of posts the algorithm ignored.

It also has one dishonest explanation, and you cannot tell which from the ratio alone. So treat it as the reason to run checks 3 and 4 rather than as the finding itself.

Check 3: do the comments say anything

Open the last ten posts and read the comments yourself. Not the count. The text.

Coordinated engagement reads like a form letter. Fire emojis, "amazing content", "love this", arriving in a cluster within minutes of the post. Real comments argue, ask the price, tag a friend, or mention something that only happens in the video.

This check takes four minutes and no tool, and it is the one that most often changes my mind about an account that passed the ratio test.

Check 4: does the follower graph have steps in it

Bought followers arrive in blocks. Organic ones arrive in a slope with occasional spikes where something travelled.

A vertical step of thousands over a day or two, with no post underneath it that explains the spike, is the clearest signal in the whole protocol. It is even clearer when the step lands just under a round number, because tier pricing pays for crossing it.

An old step with two honest years on top of it is a different situation from a step last month. Both change what your money is buying, and only the second is worth stopping a campaign over.

The reason the market keeps paying for this is that rates are quoted per thousand followers, so the one number a creator can buy outright is the one the rate card rewards.

Check 5: how many people saw the last story

Story views against follower count is the check creators do not expect, which is exactly why it works.

Stories are seen mostly by people who follow the account and open the app. Stories are seen mostly by people who follow the account and open the app. Purchased followers mostly do not, which is why the gap shows. So a large account whose stories reach a small fraction of its audience is telling you something the feed posts were hiding.

Ask for the view count on the last five stories, in the same screen recording as check 1. You are looking for a consistent range, not one good number.

Check 6: who else paid them, and did anyone come back

Scroll a year of posts and list the brands.

Then look for repeats. A creator who worked with fourteen brands in a year and none of them twice is a creator whose campaigns did not produce a reason to rebook. That pattern says more about likely outcome than any audience metric on the page.

Category churn matters too. Three competing products in one category inside two months means the audience has already been sold the same thing three times, and your version arrives fourth.

Check 7: what happens on a post nobody pushed

Find their last few posts with no brand tag and no promotion behind them.

That is the creator's floor, and it is the number your reel will land closest to. Sponsored posts are often boosted, sometimes by the creator quietly, which makes them the worst possible sample to price from.

None of this catches a creator who buys views on your reel specifically after the deal is signed. No pre-booking audit can, because it has not happened yet. That one belongs in the contract as a clause about paid promotion disclosure, not in the shortlist.

What disqualifies a creator here

Six signals end a conversation at Atvis Media rather than opening a negotiation.

They refuse to share insights, or send a screenshot after being asked for a recording twice.

The audience does not live in the market the brand sells to.

Engagement arrives in a block within minutes, with comments that could be pasted under any post in the country.

The follower graph has a vertical step with nothing underneath it.

Story views sit far under what the follower count implies, consistently.

A year of brand partners with no repeats, in a category that should produce repeats.

Publishing that list is the point of this post. A brand can now run the same six tests on any agency's shortlist, including ours, and ask why a creator survived them.

Where the money really goes

Now the uncomfortable part, and the reason those video titles exist.

Influencer budgets move through approval chains where follower count is the easiest number to defend on a slide, and Influencer budgets move through approval chains where follower count is the easiest number to defend on a slide, and the question of which cities the audience lives in rarely reaches that room..

So the creator is usually not the reason the campaign failed. The brief was, or the metric was never set, or the product was never something a reel could sell in the first place. Why D2C influencer ROI disappoints covers that half. The other half is the one where you paid too much for the right person.

And a creator can pass all seven checks and still make a reel nobody watches. When that happens the protocol worked and the campaign still lost money, which is the part no agency page will tell you.

The short version

1. Get the audience geography as a screen recording before anything else.
2. Treat a weak engagement ratio as a question, then go read the comments.
3. Look for steps in the follower graph, especially near tier boundaries.
4. Check story views against follower count.
5. List their past brands and look for repeats.
6. Price off their unpromoted posts, not their sponsored ones.
7. Ask any agency for its disqualification list. If it has none, it has no protocol.

What an influencer marketing agency does sets out the rest of the job, and if you want the seven checks run on a shortlist you already have, send it over.