Blueprint
Strategy
Brands
Testing many creators is a sampling problem, not a budget problem. The affordable way to do it is to make the guaranteed portion small but real, put the upside in commission, and accept that you are buying information rather than a campaign. A realistic structure is a modest fixed fee plus commission per creator, run across enough creators to produce a usable signal, with the explicit intention of concentrating budget on the few who perform. What makes it work is deciding the success criteria before you start.
The reason most brands work with three creators instead of twenty is not that they prefer three. It is that twenty standard fees is a budget nobody has. So they pick three carefully, hope, and learn almost nothing generalisable when it does not work.
This is a sampling problem. Creator performance varies enormously and is only weakly predicted by anything visible in advance, which means a small sample is mostly noise. This article sets out how to run a wide test on a budget that exists.
If creator outcomes vary widely — and they do — then three results tell you very little. You cannot distinguish a bad offer from three unlucky picks, and you cannot tell a good creator from a good week.
The consequence matters commercially: brands regularly conclude "influencer marketing does not work for us" on a sample too small to support the conclusion. The honest version of that sentence is almost always "these three did not work, and we do not know why."
A wider test does not make each result more reliable. It makes the distribution visible, which is what you actually need in order to decide where to put money next.
The affordability comes from changing what you guarantee, not from paying creators badly.
| Component | What to do | Why |
|---|---|---|
| Guaranteed fee | Small but real — enough to respect the work | Zero guarantees filter out the creators worth testing |
| Commission | Set generously, near your ceiling | This is where the creator's upside lives; be visibly fair |
| Product | Provided, not counted as payment | They need it to make the content; it is a cost, not a fee |
| Deliverables | Minimal and fixed | You are testing the creator, not a production brief |
| Duration | Enough for at least two posts each | One post is an anecdote; two starts to be a signal |
| Rights | Optioned, priced in advance | You will want the winners later — agree the price now |
The last row is the one brands skip and regret. If a creator performs, you will want to run their content as ad creative. Agreeing the licence price before anyone knows who won is cheaper and far less awkward than negotiating afterwards. The reasoning is set out in buying UGC upfront vs after it has performed.
The figures below are hypothetical and used to illustrate the method. They are not benchmarks.
Say you have DKK 30,000 and a product costing you DKK 150 to make.
| Line | Calculation | DKK |
|---|---|---|
| Guaranteed fees | 20 creators × 1,000 | 20,000 |
| Product cost | 20 × 150 | 3,000 |
| Reserve for rights on winners | 3 × 2,000 | 6,000 |
| Buffer | — | 1,000 |
| Total committed | 30,000 |
Commission sits outside this budget by design: it is variable, it only occurs alongside revenue, and it should be forecast separately against margin. If your margin cannot support the commission rate you are offering, that is a pricing problem to fix before the test, not during it — work out the ceiling as described in how much commission influencers should get.
If DKK 20,000 of guarantees is too much, the lever is the number of creators, not the fee per creator. Ten creators at DKK 1,000 is a real test. Twenty at DKK 250 is an insult that will be declined by exactly the people you wanted.
A small guarantee is easier to accept when the rest of the offer is good. Things that cost little and matter:
That last point is the difference between a low offer and a low first offer, and experienced creators respond very differently to the two. More on the framing in writing outreach that gets a reply.
The point of a wide test is to learn what works, so the selection should deliberately vary the things you are uncertain about: audience size band, content style, category adjacency, platform.
Twenty creators who all look like your existing customer profile will produce twenty similar results and teach you nothing you did not already believe. Deliberately include two or three you are unsure about — the surprises are the point. Baseline screening still applies: check the audience is real before you pay, per how to verify an influencer's audience and how to spot fake followers.
Write these down before the first post goes live, because deciding afterwards is how motivated reasoning gets in.
That last criterion is the most valuable output of a wide test. If twenty different creators all send traffic that does not convert, the common factor is not the creators — see why influencer campaigns get engagement but no sales.
Twenty creators is roughly where manual coordination stops being viable — twenty briefs, twenty links, twenty codes, twenty payment reconciliations, twenty follow-ups. Practical guidance is in coordinating 10 or more creators and scaling from 10 to 100.
Three things that specifically prevent a wide test from collapsing:
Expect a skewed distribution: a few clear performers, a larger middle, and some who produce almost nothing. That is the normal shape and it is the information you paid for.
The action is concentration. Raise the guarantee for the performers, run a second round with the promising middle, and thank the rest properly. How to read the results is covered in how to identify which influencers to scale; renegotiating upward is covered in renegotiating with a top performer.
This section is Make Influence's own operational view, not an industry standard.
We think the wide-test-then-concentrate pattern is the single most underused move available to a brand with a modest budget, and that the reason it is underused is administrative rather than financial. Twenty relationships is not expensive. Twenty relationships managed by hand is exhausting, and that is what actually stops people.
Two things about our own platform are relevant to this specific use case. Creators are manually approved rather than indexed from public profiles, which raises the floor of a wide cohort — the main risk in a broad test is that widening it means lowering quality. And our service fee is a percentage of the commission paid to the creator — published as 30% on Entry, 25% on Basic, 20% on Growth and 15% on Full Service Pro — so adding creators to a test does not add fixed cost.
The mistake we would most want a brand to avoid is running the wide test and then not acting on it. The test only pays for itself in the second round, when the budget moves to the people who earned it.
Is twenty the right number?
It is a useful target, not a rule. The real question is whether your sample is wide enough that one outlier cannot dominate the conclusion. Ten is meaningfully better than three.
Will good creators accept a small guarantee?
Some will, if the commission is genuinely generous, the reporting is transparent, and you are explicit that it is a first step. Many will not, and that is a legitimate position rather than a negotiating stance.
Should I tell creators it is a test?
Yes. It is more honest, and it makes the "strong performers get bigger deals" message credible rather than sounding like a brush-off.
What if the whole cohort underperforms?
Then you have learned something about your offer, price or landing page for a fraction of what a single large campaign would have cost. That is a successful test with a disappointing result.
Can I do this with gifting instead of fees?
You can, and the response rate and quality will both fall sharply. See gifting vs paid collaborations.
Make Influence
Find creators with real audience data, run collaborations in one place, and see clicks and sales per creator while the campaign is live.
Book a demoCreate accountMake Influence
Apply to campaigns from brands that are actively looking, follow your own clicks and sales, and get paid without chasing invoices.
Create creator profileMore creator guidesMake Influence
Briefs, agreed terms, tracking links and results sit together — so brands and creators see the same numbers.
See how it worksBrowse the Academy
