Creator Marketing Playbooks

Should you tell a creator how their video did?

They asked how it went. Which of the things on your side is an answer, and which ones only look like one.

They asked how it did. The answer you have isn't a number.

Two weeks after the campaign closed, a creator asks how it did. It's a fair question, and two things in front of you look like answers: the ads dashboard, and the number under the post.

The first describes the ads you ran. The second is already in front of them, under their own post. What's left is one thing: your own read on whether the video did the job you wrote for it.

What are they asking for?

The same words can carry two questions, and they don't have the same answer.

One is did the thing I made do its job. The other is how am I doing. Only the first is a question about the video, and it's the only one anything on your side speaks to.

Separate them before you pull anything up. What you reach for first is a count, and a count is easy to read as an answer to the second question even when the first one was asked.

For the jacket: you sell a rain jacket that packs into its own pocket, and the sentence the campaign had to land is that it lives in the bag you already carry and you stop noticing it's there. Two people were hired. One posted on their own account. The other filmed a version you ran as ads, and it was never posted on their account.

Why doesn't what's in your records answer it?

Because those records describe what you ran, not what they posted.

The guide to the numbers only a creator can see settles this from the other direction, since it's about you asking them and the question it answers there is whether to ask at all. The part that carries over is what it says about whose results are whose: "you're the one paying for them and running them, so results for what you're running come back on your side, and those results describe the ads." The post is a separate object. "The post found its own audience with none of that money behind it, and nothing on your side says how that post was watched."

For the jacket: the paid media ran for a week and a half, in front of people you picked. The post went up once, on a Tuesday, in front of people who follow somebody who walks to work. Your dashboard is describing the week and a half. It holds nothing about the Tuesday.

What do they already have that you don't?

The count under the post is public, and it is under their own post. Behind a sign-in nobody but them has, there is more.

The creator's own account reports how the video was watched across its own length, and the same guide describes what that is for: "It's kept for the person who posted it, so they can work out what to change next time." They don't need you to send them a count they can see.

There's a second reason, and it's about what makes a count mean anything at all. A count on its own tells them very little. What tells them something is the count set beside what that account usually does, because the same figure can be an unusual week for one account and an ordinary one for another. So what makes a count about their post mean anything is what that account usually does, and that history is theirs. Send a figure over and they will hold it up against a baseline you were never holding.

What happens if you send the numbers anyway?

A figure arrives carrying something it can't back up.

We've already ruled this reading out in the other direction. On whether to book somebody again, our guide to going back to the same creator is blunt: "So a video that reads like it missed isn't a fact about the creator at all." And on what a result that reads well proves: "What landed was a sentence, in a content format, in a campaign. It isn't a reason attached to a person."

That rule was written to stop you from making a bad hiring decision. Send the raw count back and you hand the same misreading to somebody else, pointed at themselves. What would correct it is the check you ran against the sentence, and that stayed on your side.

This is not about keeping anything back. A figure you send will be read as something, and what it can mean on their side is set by a history you don't hold, so the reading you intend isn't the one that arrives.

Got a message like this open in another tab? Send us the campaign and we'll work out the reply with you. Book a demo →

So what is yours to send?

One thing, and you may already have written it down.

When you worked out whether the video did its job, you did it against the sentence you wrote before the shoot, and that judgment produces a split: "Then you can say which parts landed and which didn't, and that split is the version you take into the next brief." How you arrive at it belongs to that guide.

What matters for this question is where that split came from. It exists on your side only, and it's about the video rather than about them. It answers the question they actually asked — did the thing I made do its job — measured against the sentence they were briefed on. That is the thing a count could not do: a count means something only against a history you don't hold, and this means the same on both sides. They can read the same comments you did; what they can't do is the check you ran. So it's the piece to send.

In practice it's short. What the video had to land. What showed up. What didn't.

There's a form for this already. The same guide sets it out for a reply to a creator who sent something over unasked: "a reply that says what you'll do with it is worth more than one that says thanks." That was about material arriving from their side and this is about material going out from yours, and it's the same test either way. The one to skip is the message that says the campaign went well. It's a verdict with nothing named in it, so there's nothing in it they can take into anything.

For the jacket: the video had to leave people understanding that it lives in the bag you already carry and you stop noticing it's there. People asked in the comments whether it fits in a tote, so they'd taken on the bag part and were working out the rest. Nothing came back to not noticing it, so nothing showed that half landed — it's going in the next brief as the opening rather than the payoff.

What if there's no post?

Then the split is the only thing that reaches them.

What you bought decides this. What you settle before the shoot puts the choice between a post on their account and a file for your ad account first among the things to agree, and names it exactly: "Same creator, same brief. Different purchase." Buy the file and nothing is posted on their account, so there's no screen on their side reporting how it was watched.

For the person who posted, your split sits next to a screen they already have. For the person who filmed a file, your split is the only thing about the video that reaches them at all.

For the jacket: the second creator has no screen reporting how the version they shot did, and nothing on their side will supply one. The ads are the only place it ran, so your read is made there instead of under a post. What goes back is still the same three lines: what it had to land, what showed up, and what didn't.

When do you send it?

When you've got the split.

The split comes out of judging the video against the sentence, so it exists as soon as you've done that work, and there is nothing to send before it. If the question arrives before then, say so: tell them when you'll be doing that work, and that the answer comes with it.

Frequently asked questions

We ran it as ads from the creator's own handle. Do those results go back?

The handle they went out from isn't what delivered them — your money did. Running ads from the creator's handle is a separate permission, and it doesn't move the answer in either direction. The same guide has already settled this case for asking, and the handle doesn't change it on the way out.

The one thing that does change: they can see the ads going out under their name, so they may ask about those specifically. What you can say about them is what you'd say about any of your own paid media: it answers for the ads, and your account is where that answer stops.

Nobody wrote a sentence for this campaign. What do we send?

Not a split, because there isn't one. The guide already names this case: when there was no sentence, nothing missed. You can't tell them which parts landed when nothing was named in advance.

What's honest is short: say what you saw, say what you're naming before the next one, and don't dress either up as a result. What's missing here is the sentence, and that's the thing the next brief fixes.

See what this looks like for your brand

Running one now and want the reply ready before it closes? We'll work through the campaign with you. Book a demo →

Or email us: business@popow.ai

Ready to run this with creators?

See how Popow briefs, matches, and ships TikTok-native content for your campaign.

Book a demo →business@popow.ai