← Journal

How to read the numbers of a TikTok campaign

Views are the number everyone quotes and the one that tells you least. Here is the order we actually read a report in, and what we do when the shape looks wrong.

Chance Worldwide Published 6 min read Campaigns
An artist performing to a festival crowd
The only number that matters is the one that still moves after you stop paying.

Read a campaign report in this order: organic creations, then saves and completion, then the streaming curve, and only then views. Views measure how much was delivered. The first three measure whether anyone cared. A campaign is working when creations keep appearing after the paid posts stop.

Why views are the wrong first number

Views tell you that a video was served to people. That is a statement about distribution, not about the record. Any budget can buy views; nothing can buy the decision to use a sound in your own video.

It matters because views are also the easiest number to inflate, and the one most reports lead with. If a deck opens with a total view count in large type and buries everything else, that is usually deliberate.

The four numbers we read first

1. Organic creations

How many videos used the sound without being paid to. This is the closest thing to a truth serum in music marketing. A paid creator posting is a delivery; a stranger picking up the sound is a signal. Track it as a daily count, not a total — the shape tells you more than the sum.

2. Saves and shares

Saving a video or sending it to a friend is a costlier action than a like, so it is harder to fake and it correlates better with people going to find the track. We weigh saves far above likes and comments.

3. Completion

Whether people watched to the end. If completion is poor on videos featuring your sound, the sound is not the reason people stayed, and the campaign will not convert however many views it accumulates.

4. Where the sound sits in the video

Not a metric, a review. Watch twenty of the posts. If the track is inaudible under a voiceover, or the hook lands after the average viewer has scrolled, the numbers were never going to work. This takes fifteen minutes and explains more failed campaigns than any dashboard.

What a working campaign looks like against one that was bought

SignalCampaign that is workingCampaign that was bought
Creations after paid posts stop Keep appearing, sometimes accelerating Fall to near zero within days
View curve Uneven, with spikes when a post catches Smooth and flat, delivered on schedule
Comments About the song — lyrics, "what's this sound" Generic, emoji-only, or about the creator only
Streaming Rises during, holds part of the gain after Flat throughout, or a bump that vanishes
Geography Concentrates where the artist has a base Scattered across unrelated markets

None of these is proof on its own. Three of them together usually is.

Seven days for shape, twenty-eight for outcome

We read a campaign twice. At seven days, the question is whether the sound has any pull of its own — are creations appearing that we did not pay for. This is early enough to change the creative brief, swap the section of the song being used, or stop spending.

At twenty-eight days, the question is what survived. Compare the streaming baseline in the fortnight before the campaign with the fortnight after it ended. The gap between those two numbers is the campaign's actual result. Everything measured while you were still paying is a rental.

When the shape is wrong, change one thing

The instinct when a campaign underperforms is to add budget. That usually buys a larger version of the same result. The variables actually worth changing, roughly in order of how often they turn out to be the problem:

  1. The section of the song being used. The chorus is not automatically the hook. A pre-chorus, an ad-lib or a producer tag is often what people can build a video around.
  2. The brief given to creators. "Use this sound" produces filler. A specific action, joke or format produces videos other people want to copy — and copying is the whole mechanism.
  3. The creators themselves. Audience fit beats follower count almost every time. A creator with a smaller, tighter audience in the right country is worth more than a large scattered one.
  4. The budget. Last, and only once the three above are right.

What a report should contain

If you are commissioning a campaign, ask for these before it starts, so nobody is choosing the flattering metrics afterwards:

A partner who will commit to that list in advance is telling you something. So is one who will not.

If you take one thing from this

Judge a campaign on what still moves once the money stops. If creations, saves and streams all fall back to the pre-campaign line within two weeks of the last paid post, you bought attention rather than earning it — and the same budget spent on a different section of the song may well behave differently.

Common questions

How many views does a campaign need to move streams?

There is no threshold, because views are not what moves streams. What moves streams is people using the sound themselves and finishing the video. A campaign of 2 million views spread across creators whose audiences never touch the sound will do less than 200,000 views that produce a few thousand organic creations.

How long should you wait before judging a campaign?

Read the shape at seven days and the outcome at twenty-eight. The first week tells you whether the sound has any pull of its own; the fourth week tells you whether anything survived once the paid posts stopped.

What does it mean if views are high but streams are flat?

Usually one of three things: the sound is not audible or memorable in the videos that ran, the creators reached an audience that does not stream, or the traffic was not genuine. Check organic creations first — if almost nobody used the sound without being paid to, the campaign did not connect.

Working on something?

Tell us what you are working on. We answer everybody, and quickly.

Send us a message
All articles → Music & Influence →