The standard segmentation of creators is by follower count, into tiers with names that vary by whoever is presenting. It is a poor way to make decisions because follower count is not what differs between them in any way that matters.
What actually differs
Three things, and they do correlate with size loosely.
Relationship depth. Smaller audiences generally have a closer relationship with the creator — more direct interaction, more sense of knowing them, more weight given to what they say.
Audience breadth. Larger audiences are more heterogeneous, which means a lower proportion of them are relevant to any specific product.
Professionalisation. Larger creators generally have management, contracts, rate cards and a workflow. Smaller ones frequently do not, which means both flexibility and unreliability.
Those are the variables. Follower count is a rough proxy for all three and a poor one, and there are large creators with tight communities and small ones with disengaged followings.
The engagement rate inversion
The consistently observed pattern that drives most of the argument for smaller creators.
Engagement rate as a proportion of followers tends to decline as follower count rises. Smaller accounts see a higher proportion of their audience interacting.
The explanations are plausible: closer relationships, more replies from the creator, and ranking systems that surface content differently at different scales.
The caution is that engagement rate is a ratio, so it falls arithmetically as the denominator grows even if absolute engagement rises. A creator with a hundred times the followers and thirty times the engagement has a lower rate and reaches far more people.
Which of those matters depends entirely on whether you are buying reach or advocacy.
What each size is actually good for
The practical framing I now use.
Very small creators with tight communities are good for credibility in a specific niche, for detailed product explanation, and for reaching an audience that is hard to reach any other way. They are poor for reach and expensive to manage at volume.
Mid-sized creators are the workhorses. Enough reach to matter, enough relationship to persuade, generally professional enough to work with, and priced sensibly. Most of the programmes I have seen work well are built here.
Large creators are a reach buy, comparable to media, and should be assessed as such. The relationship effect is weaker, the audience is broad, the cost is high, and what you are buying is exposure plus a degree of association.
Treating a large creator as advocacy and paying a premium for the relationship effect is where budgets get wasted.
The management cost problem
The thing that makes small-creator programmes harder than they look on a spreadsheet.
Working with fifty small creators is fifty briefings, fifty contracts, fifty content reviews, fifty payments and fifty relationships.
The per-creator cost may be low and the operational cost is substantial and is frequently not budgeted.
Platforms exist to manage this and take a fee, which erodes the cost advantage that made small creators attractive.
The practical answer we arrived at is fewer creators, longer relationships, which reduces the per-campaign overhead by amortising the relationship across many pieces of content.
Reliability scales with professionalisation
A blunt point that is worth making because it affects planning.
Smaller creators are more likely to miss deadlines, post content that does not meet the brief, forget disclosure, or disappear mid-campaign.
This is not a character judgement. Most of them are doing it alongside other work, without support, and a brand campaign is one of many things in their week.
The mitigation is longer lead times, clearer briefs, and building in the assumption that a proportion will not deliver.
Larger creators deliver reliably and negotiate harder, which is the trade.
What I would actually do
Stop segmenting by follower count and start segmenting by audience fit and demonstrated advocacy.
Ask for audience demographics before any selection decision.
Look at whether the creator has ever said anything critical, which is the strongest available signal that their recommendations carry weight.
Build a small stable of mid-sized creators and work with them repeatedly rather than a large rotating cast.
And budget the operational cost honestly, because a programme that looks cheap per creator and requires a full-time person to run is not cheap.
Employees and customers as a category
The group that sits outside the usual tiers entirely and is frequently the most effective.
Employees with modest personal followings, talking about work they actually do, reach audiences that are small, relevant and unusually trusting.
Customers who post unprompted are the strongest form of this and cannot be bought, only encouraged and amplified with permission.
Neither shows up in any creator tier framework, both cost very little, and in the programmes I have seen they frequently outperform paid partnerships on any measure that matters.
Paying fairly
A point about rates, since the market is opaque.
There is no established rate card, and quoted prices for comparable audiences vary by an order of magnitude depending on who is asking and who is representing the creator.
What has worked for us is being transparent about the budget and asking what they would do for it, rather than negotiating down from a quote.
It also matters that smaller creators are frequently underpaid relative to the value delivered, and that paying properly for a long relationship costs less than a series of one-off campaigns with people who feel used.