My co-founder and I run a small apparel brand. We design and produce small batches of basics — tees, sweats, a few outerwear pieces — and sell them through our website and one storefront in Brooklyn. We have done influencer marketing in some form for two years. We have worked with 27 different creators across that time, ranging from 1,500 followers to 180,000. We have spent roughly $34,000 on influencer activity in those two years.
Here is the quiet truth: of those 27 creators, 6 produced real measurable revenue, 4 produced something we could not measure but might have helped, and 17 produced nothing. That is a 22% hit rate. If I had known that going in I would have planned differently. I want to write the article I needed to read 24 months ago.
What the influencer marketing industry tells you
Industry articles say things like 'micro-influencers deliver 7x the engagement of macro-influencers' or 'influencer marketing has a 5.2x ROI on average'. These numbers are technically true and meaningfully misleading. They are averages of campaigns run by brands with budgets, infrastructure, and audience data that small brands do not have.
What the industry never says is that the median ROI for a small brand doing influencer marketing for the first time is approximately zero. The average is dragged up by a small number of huge wins. Your campaign is much more likely to be in the long tail of zero than in the head of huge wins.
Why so many creators produce nothing
After 27 partnerships, the pattern is clear. Creators produce nothing when one of three things is true. First, their audience is not actually their audience. They have followers, but the followers do not act on their recommendations. This is the inflated-account problem and it is extremely common in the 10K to 80K range.
Second, the creator's audience is the wrong fit for your product. A creator with great engagement among 24-year-old fitness enthusiasts is not going to sell our cashmere crewneck to people who buy basics for the office. The fit has to be exact, and most creators will tell you their audience fits your product even when it does not.
Third, the creator is going through the motions. They have done so many brand deals that their audience has tuned out the sponsored content. The post happens, the engagement is real, the sales are zero because their audience has trained itself to scroll past anything that looks paid.
The six creators who actually worked
Looking at our six successes, they had three things in common. All three matter. Miss one and the campaign falls apart.
One: they had a real and demonstrable audience, not a follower count. Their engagement rate was 5%+ across their last 20 posts. Their comments were specific and conversational, not single emojis. The accounts engaging looked like real people. We checked this before working with them, every time.
Two: their audience was an exact fit for our product. Not 'fashion girlies' but 'people who buy minimal basics they will wear for years.' The narrower the fit, the better the result. The creator with 1,500 followers who wore our basics every other day to her literal day job at an architecture firm outsold creators with 80K followers and a broader 'style' audience.
Three: they were creating because they liked our product, not because they liked the contract. The creators who reached out to us first, or who had organically posted about us before we paid them, produced 4 to 5 times the results of creators we cold-pitched. The pre-existing interest was the most predictive variable in the whole game.
The math on influencer marketing only works if you treat the partnership as an extension of an interest the creator already has. Paying someone to pretend to care about your brand gets you scrolled past.
What the four 'maybe worked' creators taught us
Four creators produced engagement we could not directly tie to sales but seemed to help in less measurable ways. They produced UGC we used for months. They drove email signups that converted later. They built a kind of cultural credibility in a specific micro-community.
I am not sure how much we should value those outcomes. We did one analysis where we estimated the value of UGC we would have had to produce ourselves, plus the email list growth, plus the assumed lift on brand recall. The number came out to roughly break-even on those four creators. Not a hit. Not a loss. Tuition.
The 17 who produced nothing
These were the most expensive lessons. Average spend per creator in this bucket was around $1,400 in product plus fee. They had followers in the 8K to 50K range. They had media kits. They had social proof from other brands. They had nice DMs.
Three of them I now suspect had bot-inflated followers. Five of them had real audiences that did not actually fit our product. Nine of them were burned out on brand deals and their audiences had tuned out. The total spend on this bucket was about $24,000. The total attributable revenue was about $1,800.
What we do now
- We only work with creators who have organically posted about us first. That single rule cut our miss rate from 78% to about 15%.
- We pay smaller upfront fees and offer larger affiliate cuts. The creators who believe in the product accept this happily. The ones who do not, do not. The negotiation itself filters for the right partners.
- We require a live screen-share of analytics before sending any product. If they will not share their last 5 posts' real data, we do not work with them.
- We trial new creators with gifted product only for the first round. If something organic happens, we move to a paid partnership.
- We track every order from a creator code for 90 days, not 7. The long-tail conversions are real and often bigger than the immediate spike.
Honest cost-per-acquisition math
Across all 27 creators and $34K in spend, we generated roughly $52K in directly attributable revenue. Looks profitable. But the gross margin on our products is 38%, so the $52K in revenue is $19,760 in gross profit. We spent $34K to generate $19,760 in gross profit. Net loss of about $14,240 across two years of influencer marketing.
If we exclude the 17 misses, the 10 partnerships that worked or might have worked produced about $48K in revenue on $10K in spend. That subset was extremely profitable. The problem is that we did not know in advance which 10 those would be.
The lesson is that influencer marketing for small brands is not a marketing channel. It is a portfolio. Most of the bets fail. A few pay for everything. You need the discipline to spread small, kill fast, and double down only on what has proven itself.