How TikTok Campaigns Are Actually Measured

The reporting dashboard shows a lot of numbers, most of which are inputs to decisions rather than answers to questions. Knowing which is which saves a lot of misreading.

The two that decide the campaign

Cost per result, where the result is whatever event you designated. If you optimised for purchases, this is cost per purchase, and it is the number that determines whether the campaign is viable.

Purchase conversion rate across the whole path, not just the ad. This is where the campaign's weakness usually is, and it is the one dashboards report least clearly.

The rest, briefly

• Impressions and reach — delivery, not outcome. Useful for frequency, misleading as a headline.
• Video views — how long someone stayed. A good creative signal and a poor sales signal.
• CTR — how many clicked. Strong on cold audiences, weak as a target to optimise toward directly.
• Frequency — how often the same person saw it. The number that explains why performance degrades over a long campaign.
• Learning phase — the period during which delivery is unstable. Judging a campaign here is judging noise.

How the platform chooses what to show

Delivery optimises toward the event you designated. This is the single most important fact about the dashboard: if you ask for traffic, you will get traffic, efficiently, indefinitely, because traffic is what the system was instructed to find. Many campaigns that look like near-successes on engagement are simply traffic campaigns that were never asked to do anything else.

The measurement habit worth having

Check cost per result against the event you care about, and separately check the conversion rate all the way to the sale. If the first improves while the second falls, you are buying traffic to a page that is losing people, and no budget increase will fix it.