Your creator posts about you on a Tuesday. In the next second, before reading a word, her audience makes one decision: is this something she uses, or something she was paid to hold?
Everything the campaign produces depends on that answer, and nothing in your brief addressed it.
You thought you were buying reach. Reach is the thing a creator sells that is worth the least and is easiest to fake. What you were trying to buy is a borrowed act of judgement, in front of people who trust it, which is a much narrower product with a much shorter list of suppliers.
So ask one question before you look at any follower count. Does this person’s audience believe she turns work down?
If the answer is no, you are buying reach at a premium, and there are cheaper ways to buy reach.
The engagement rate you were shown is partly measuring intimacy
You will be told that smaller creators get much better engagement, and the direction of that is probably right. The numbers are a mess.
I went looking. One widely quoted 2026 analysis puts micro and nano creators at two to three times the engagement of macro accounts. An agency piece reports micro at 4 to 5% against macro at 1.5 to 2%. A festive-season report gives nano at 3.93%, micro at 1.79% and mega at 0.48%. Then another source reverses the order, putting micro above nano. Tier definitions move between all of them, sample sizes are mostly unpublished, and nearly every one comes from somebody selling creator campaigns.
So I am not going to hand you a multiple. What I will hand you is the structural reason to be careful with the whole metric.
Engagement rate is a ratio, and a ratio improves when the denominator shrinks. A creator with 4,000 followers has a high engagement rate partly because a good share of them are her cousins, her classmates and the aunties from her building, who will like anything she posts because they know her. That is intimacy. It is real, and it is not influence over a purchase.
The honest counter-argument deserves saying too, because the small-is-better crowd skips it. A celebrity post at 0.8% across three million people is twenty four thousand interactions. A nano creator at 6% across five thousand is three hundred. Engagement rate without the absolute number attached is a percentage of nothing.
Which is why follower tiers are the wrong axis entirely, and category fit is the right one.
Why a single post is nearly always wasted
Brands buy one post. It is the default, and it is close to the worst thing you can buy.
The frequency argument I have made before, about local sponsorship, so I will not relitigate it. There is a second reason specific to creators, and it matters more.
A one-off appearance borrows none of the creator’s history. Her audience has watched her for two years and can see that you appeared once, said your piece and vanished, which is the exact shape of a paid placement and gets read as one.
A creator who mentions you three times across two months is making a different claim, and the audience reads it correctly: she is still using the thing. That is the claim you wanted. You cannot buy it in a single post at any price.
So if your budget allows one post with a 60,000-follower account or four with a 15,000-follower account, the second is almost always better, and not because of engagement rates.
Category fit is the whole asset
A creator’s audience trusts her about a specific thing, not about everything.
The food creator whose audience trusts her on where to eat in Bandra has no borrowable authority about your skincare, even though the demographics line up perfectly on paper. The demographics are not the asset. The accumulated history of being right about one category is the asset, and it does not transfer.
Two tests that catch almost all of the mismatches.
Scroll back three months and count her unpaid posts about your category. If there are none, she has no standing in it, and your campaign is where her audience finds out she will take work in anything.
Then count the paid posts, full stop. A creator who posts a sponsored recommendation every week has already spent the trust, and she is selling you something she no longer has. Her rate card has not caught up with that yet, which is the trap.
I would rather have a creator with 9,000 followers who has been publishing about my category for three years and takes four paid posts a year.
Barter works, and sometimes it is an insult
Free product for a post is reasonable in exactly one situation: when the product is itself the story.
If you make something a creator in that category would genuinely want to try, is new, and is worth a post on its own merits, barter is a fair trade and both sides know it.
If you make soap, and the ask is a scripted post with three hashtags and a discount code, you are asking for professional work and offering Rs 400 of goods. That is not a partnership. The creators worth having will decline politely and remember, and the ones who accept are telling you what their endorsement costs, which is less than you hoped.
The test is whether you would be comfortable if she wrote exactly what she thought, including the part she did not like. If the honest version would hurt, you were not offering a trial. You were buying a scripted ad and pricing it as a gift.
The rule that makes the thing you bought visible
Worth knowing precisely, because it is misunderstood in both directions.
ASCI’s social media disclosure guidelines require a label wherever there is a material connection, and material connection explicitly includes free products, discounts or perks rather than only money. Disclosure is required even for an unbiased review if the connection exists. The permitted labels are a short list, including Ad, Sponsored, Collaboration, Partnership and Free gift, and the label has to be hard to miss rather than tucked in a bio, buried in hashtags or added at the end. Responsibility sits with both the advertiser and the influencer, and ASCI is explicit that advertisers must ensure compliance and can require edits or deletion.
That last part is the bit for you rather than for her. It belongs in your contract, not in your hopes.
But notice what the rule means for the thesis. The thing you are buying is credibility, and the law requires it to be labelled as purchased. Which means a brand that needs the recommendation to look unpaid is trying to buy something it cannot legally have.
The good news is that the label is not what kills the effect. A creator with real standing can say “this is a paid partnership and I have used it for six weeks and here is what I think” and be believed, because the audience is assessing her history rather than her funding. A creator without standing cannot be saved by hiding the tag.
Where I come out
Stop thinking about this as media and start thinking about it as hiring a reference.
Which means the brief changes. Fewer creators, chosen for standing in one category, booked across two or three months rather than once, paid properly in money rather than fobbed off with product, and left enough room to say something you did not write.
And one honest limit. If your product is genuinely undifferentiated, no amount of borrowed judgement fixes that, because what you are asking somebody to do is stake their credibility on a thing they have no reason to prefer. The creators with the standing worth renting are also the ones who can tell.
If you’d like a second opinion on a creator list before you commit the budget, or on whether your category has the kind of trust that transfers at all, drop me a line on email, WhatsApp, phone or LinkedIn and we can have a quick chat, and my notes on social media marketing cover the owned side of it. I’m taking on a small number of contracts at the moment, so I’ll tell you straight whether it’s something I could help with.