How to measure a micro-creator campaign honestly
Decide what the campaign is for before it starts, set up tracking before the first post, and compare the result with what the same money does in your other channels. Small campaigns produce small numbers, so the aim is a sound decision, not a perfect figure.
- One campaign, one goal. Content, reach and sales are measured differently, and a campaign judged on all three will look like a failure on at least one.
- Set up codes, links and a baseline before anything goes live. Tracking added afterwards measures nothing.
- Codes and links catch only part of the sales effect. Treat them as a floor and check the period against your baseline.
- Compare creators with each other and the campaign with your other channels. Those two comparisons drive every useful decision.
- Earned media value is an estimate with generous assumptions. It is not revenue and should never be reported as if it were.
Measuring a campaign with five or ten small creators is harder than measuring a paid ad, and it helps to admit that at the start. People see a video on a phone, search for the brand two days later on a laptop, and buy without ever touching a tracked link. The numbers from any single creator are small enough to swing by chance. None of that means the channel cannot be measured. It means the measurement has to be designed before the campaign runs, matched to one goal, and read with some humility. This page sets out which metrics fit which goal, the tracking to put in place first, how to read the results a week and a month later, and how to turn them into the only decision that matters, which is what to do with the next budget. It also names the numbers that look impressive and tell you nothing.
Pick one goal, because it decides the metrics
The goal decides what you count. A campaign bought to produce ad creative should not be judged on the creator's likes, and a reach campaign should not be judged on last-click sales.
| Goal | Primary metric | Supporting metrics | Do not judge it on |
|---|---|---|---|
| Content for your own ads | Performance of the creator video against your current best ad | Cost per usable asset, hook hold rate in the ad platform | The creator's follower count or likes |
| Reach in one niche | Cost per thousand views, against your usual paid media cost | Saves, shares, comments that mention the product, profile visits | Coupon redemptions alone |
| Direct sales | Tracked orders plus uplift over baseline during the window | Code use, link clicks, new customer share | Views |
| Social proof and reviews | Number of usable pieces and where you deployed them | Conversion rate change on pages where they were added | Reach |
If you cannot choose, choose content. For a first campaign on a small budget it is the goal with the clearest test and the longest-lasting result, which is why our content packages are built around it.
Set up tracking before the first post
Everything in this list takes an afternoon. None of it can be done after the fact.
- Record a baseline: daily sessions, orders, new customers and branded search volume for the four weeks before the campaign.
- Give each creator a unique discount code. Codes survive the jump from phone to laptop, which links do not.
- Give each creator a unique tagged link as well, and check that the tags arrive in your analytics before launch.
- Add a short question at checkout or on the enquiry form asking how the customer heard about you, with creators listed as an option. It catches sales that codes and links miss.
- Agree in writing that each creator will send screenshots of their post analytics at seven days and at thirty days. Views, reach, saves, shares and audience location are only fully visible to them.
- Keep other marketing steady during the window if you can. A sale, an email blast or a new ad campaign in the same week makes the result unreadable.
- Book the posts into a short window, ideally one week, so that any uplift shows up against the baseline.
Write the reporting requirement into the brief and the agreement, not into a message afterwards. The one-page creator brief has a place for it.
What to look at after seven and thirty days
Most views on short video arrive in the first days, so a seven-day reading is fair for reach. Sales and search effects take longer, so read those at thirty days.
| Metric | How to work it out | What it tells you |
|---|---|---|
| Cost per thousand views | Fee divided by views, times 1,000 | Whether this reach was cheaper or dearer than your paid media |
| Engagement rate by views | Likes, comments, saves and shares divided by views | Whether the audience cared, which follower-based rates hide |
| Quality of comments | Read them. Count those that mention the product or ask where to buy | Whether attention was on the product or only on the creator |
| Cost per tracked order | Fee divided by orders from that creator's code and link | A floor for sales effect. The real figure is usually better |
| Uplift over baseline | Orders and branded searches in the window, less the baseline | The effect the codes missed. Only readable if nothing else changed |
| Audience location | From the creator's analytics screenshot | How much of the reach could buy from you at all |
The engagement rate calculator works out the rate for any post, and the statistics page gives sourced benchmarks so you can see whether a figure is ordinary or unusual for the tier.
A worked example with invented numbers
The figures below are made up to show the arithmetic. They are not results from a real campaign and they are not a forecast.
A tea brand pays five micro creators $400 each, $2,000 in total, to post in the same week. After seven days the five posts have 46,000 views between them. Cost per thousand views is $2,000 divided by 46, which is about $43. The brand's paid social usually costs about $15 per thousand, so on reach alone this was expensive.
At thirty days, codes and links account for 31 orders at an average of $38, which is $1,178 in tracked sales. The checkout question adds nine more customers who named a creator. Branded searches ran about a fifth above baseline for two weeks. One creator produced 19 of the 31 tracked orders. Two produced one each.
Read plainly: as a reach buy it lost to paid media, as a sales campaign it did not pay back in month one, and it identified one creator whose audience buys. The sound next step is not another five new names. It is to book the strong creator again, buy paid usage on her video, and test it as an ad, where the brand controls how many people see it.
With numbers this small, one creator having a good or bad week changes the totals. Avoid firm conclusions from a single post. Two or three results from the same creator tell you far more than one.
Numbers that mislead
- Follower count and potential reach. Adding up followers says how many people could have seen a post, not how many did. Use views and reach from the creator's own analytics.
- Impressions reported without a source. Ask for the screenshot.
- Earned media value. It converts views and engagement into a dollar figure using assumed ad rates. It is useful for comparing one campaign with another on the same assumptions, and misleading when presented as money earned. Our earned media value calculator shows its assumptions for that reason.
- Engagement rate by followers on a post that went to non-followers. Short video is mostly shown to people who do not follow the account, so divide by views.
- Last-click sales alone. It undercounts a channel that people watch on one device and act on later with another.
- Likes. They are the cheapest signal there is. Saves, shares and comments that mention the product carry more weight.
Sudden spikes of generic comments or views from countries you do not sell to are also worth questioning. The page on how to vet influencers covers the checks to run before you book, which is a cheaper time to find these problems.
Turn the results into a decision
A report that does not change what you do next was not worth compiling. Four decisions cover nearly every outcome.
| What you saw | What to do next |
|---|---|
| One or two creators clearly ahead of the rest | Re-book them, and buy paid usage on their best piece |
| Good content, weak organic reach | Stop paying for posts. Buy content and distribute it with your own ad budget |
| Reach was cheap, sales were absent | Check the landing page, the offer and audience location before blaming the creators |
| Nothing stood out across five or more creators in one niche | Change the niche or the message before spending more, or pause the channel |
Keep a simple sheet with one row per creator: fee, views, engagement by views, tracked orders, notes on comments, and whether you would book them again. After two campaigns that sheet is worth more than any tool. The influencer budget calculator helps size the next round.
What we report, and what we do not claim
When you buy from us, we commission the creators and you receive finished content at a fixed price. Where creators post to their own audiences, we collect their analytics and report views, engagement and audience location per creator, with the date each figure was captured. For content packages, the useful measurement happens in your ad account, and we tell you which hooks we would test first.
We do not guarantee sales, views or a return, and we do not report earned media value as revenue. If the honest reading is that the campaign did not work, the report says so. Send a brief and a plan with a fixed price comes back free, or see the pricing page first.
Questions
What is a good ROI for a micro-influencer campaign?
There is no dependable universal figure, and published averages mostly come from vendors with something to sell. Compare against your own channels: cost per thousand views against your paid media, and cost per order against your usual acquisition cost. A first campaign often does not pay back directly. Its value is finding which creators and messages are worth repeating.
How many creators do I need for the results to mean anything?
Five or more in one niche, posting in the same week, is a reasonable minimum to see a pattern. One or two creators give you an anecdote. Spreading ten creators across ten niches gives you ten anecdotes.
Are discount codes or tracked links better?
Use both. Codes survive when someone sees a video on their phone and buys later on another device, and links give you cleaner click data. A how did you hear about us question at checkout catches some of what both miss.
Can I get view and reach numbers without asking the creator?
Only partly. Public view counts are visible on some formats, but reach, saves, shares and audience location are in the creator's private analytics. Make screenshots at seven and thirty days a written condition of the job.
How long should I wait before judging a campaign?
Judge reach at about seven days and sales effects at about thirty. Content bought for ads should be judged after it has run in your ad account long enough to gather a fair number of impressions against your existing creative.
General information. The worked example uses invented figures for illustration and is not a forecast. We do not guarantee results from any campaign.
Get a shortlist.
Tell us the product and the budget. Names, rates and recent work come back at no cost.