Use nano KOLs while you are still finding the message that works, and macro KOLs once you have found it. Nano creators are affordable enough to test many angles inside real communities. Macro creators are expensive enough that they should only ever amplify a message that has already proven itself.
Most briefs we receive ask the question the other way round: we have this budget, which creators can we afford? That is how a brand ends up paying a famous face to broadcast an idea nobody has tested. The tier matters less than the order.
What is the difference between nano, micro, macro and mega KOLs?
The labels describe audience size, and the ranges below are common working definitions. Platforms and agencies draw the lines slightly differently, so treat them as a guide, not a rule.
| Tier | Typical following | Good at | Weak at |
|---|---|---|---|
| Nano | 1,000–10,000 | Trust inside a tight community, honest comments, cheap to test many angles | Reach. One post rarely moves a national number |
| Micro | 10,000–100,000 | A niche with some scale, consistent content, clear audience signals | Still limited reach, quality varies widely |
| Macro | 100,000–1,000,000 | Fast reach across broad audiences, cultural visibility | Cost per post, weaker trust, easy to waste on an untested message |
| Mega | 1,000,000+ | Mass awareness, a moment everyone sees | Very expensive, least personal, hardest to measure |
The useful question is not “which tier is best”. It is “which job does this phase need done”.
Why does the order matter more than the tier?
Our framework splits distribution into two phases. In the problem-solving phase we use organics and nano KOL. At scale we move to TVC, OOH and macro KOL. The logic is simple. Early on, you do not yet know which message, which trigger or which community will respond. Learning is the job, and learning needs many small, cheap experiments. Once one message keeps winning, the job changes to reach, and reach is what macro creators and mass media are built for.
Reverse the order and two things go wrong. You spend the biggest line of the budget before you know what to say. And you learn almost nothing, because one expensive post is one data point.
What should nano KOLs do in the problem-solving phase?
Give them room to test, not a script to read. In this phase, a nano programme should:
- Test several angles of the same product, each with its own hook.
- Place those angles in different communities, because a trigger that works for one group can fall flat in the next.
- Speak the community’s own language and visual codes, not the brand’s corporate voice.
- Report the signals that show real interest: saves, shares, the quality of comments, clicks and cost per result by angle.
When we worked with Cakap on user acquisition, the brief was to grow registrations while lowering the cost per install. Instead of one generic ad, we built native-language creative with raw visual codes for niche communities, each with a hyper-specific trigger. Registrations doubled and weekly purchases went up. The win came from matching the message to the community, which is exactly what small creators inside those communities do best. Read the Cakap case.
When is a brand ready to move to macro KOLs?
Scale when the evidence says so, not when the calendar does. We treat the move as a gate that has to be passed, not a date in the plan. Check these signals first:
- One message has won more than once, in more than one community.
- Cost per result for that message is stable, not a lucky spike.
- People are saving and sharing the content without being asked.
- The landing page, stock and customer service can handle ten times the traffic.
If any of these is missing, a macro push will mostly buy reach for a message you are not sure of. We wrote more about this way of working in Gates, not targets.
Can reach be earned instead of bought?
Sometimes the most powerful voice is one you never paid. Three of our cases make the point.
- ABC had budget for exactly one billboard. We designed it to be photographed and shared, and public figures, including Aldi Taher, reposted it. The result was multi-million organic views and a 300% sales lift on several key accounts. The billboard case.
- Frisian Flag grew its employer brand following six times, from 9,000 to 60,000, over three years with zero media spend or giveaways, by publishing content people found genuinely useful. The Frisian Flag case.
- When AQUA needed to answer doubts about its roots, the most credible voice was a loyal veteran employee, Bu Yanti. Sentiment turned 82% positive, and the film went from social to TV. The AQUA case.
None of these relied on a big creator fee. Each relied on a message people wanted to pass on. Creators of any size amplify that kind of message. They cannot rescue the other kind.
How do you measure each phase?
Measure the job each phase is doing, and decide on a fixed rhythm.
| Phase | What to measure | The decision it drives |
|---|---|---|
| Problem-solving, nano | Saves, shares, comment quality, clicks and cost per result, broken down by angle and community | Which message and which community to back |
| Transition, micro | Whether the winning message holds its cost per result at a larger audience | Whether the message is ready to scale |
| Scale, macro and mass media | Reach, frequency, branded search, sales by channel | Whether to keep, cut or rotate the spend |
Read the numbers weekly and let them change what runs next week. That is the closed loop we describe in Closed-loop marketing.
A simple sequence for your next launch
- List three to five angles for the product, each with a clear hook.
- Brief a group of nano creators across several communities, so every angle is tested in more than one place.
- Read results weekly. Cut the angles that do not move, and brief new variations of the ones that do.
- Move the winning message to a few micro creators and check that it holds.
- Pass the gate: stable cost per result, organic sharing, operations ready.
- Scale with macro creators and, where it fits, TVC and OOH, using the message that has already won.
- Keep a small nano layer running, so you keep learning while you scale.
FAQ
Is a nano KOL cheaper than a macro KOL?
Per post, almost always. Per result, not always. Compare cost per outcome, such as a registration or a sale, not cost per post.
How many nano KOLs do we need to learn anything?
Plan around the number of angles you want to test, not a follower total. Each angle should run in more than one community, so one lucky post does not decide your campaign.
Should we stop nano KOLs once the macro phase starts?
No. Keep a smaller always-on group. It keeps the content fresh and tells you early when a message starts to wear out.
Do TVC and OOH still matter if creators work?
Yes, at scale. Once a message has proven itself, mass media builds reach and memory faster than creators alone.
Planning a launch and not sure where creators fit? Send us the brief. Three lines is enough.
