Understand why reporting metrics may not add up
It's possible that your metrics may not add up when comparing totals to breakdowns or when comparing metrics at different levels in the campaign structure (e.g. account, campaign, ad set, ad).
Terms to know:
We recommend that you familiarise yourself with the following terms before reading on.
- Unique metrics: Metrics that refer to the number of unique people rather than the number of actions. Most unique metrics also contain the word unique (e.g. unique link clicks), which makes them easy to identify.
- Sampled data: Sampling is a widely used approach to data analysis. Rather than analysing an entire set of data, sampling looks at a portion of the data. This portion is considered to be a good representation of the larger population and provides similar results.
Now, let's take a look at some examples of where metrics may not add up and why this might be happening.
In Facebook's campaign structure, your ad set or ad metrics may not add up to the numbers that you may see at higher levels (e.g. ad to ad set, ad set to campaign). This may be because your reports aren't counting deleted ad sets or ads.
As the default view in Ads Manager shows only ad sets or ads that haven't been deleted, check for any that may have contributed to higher-level numbers.
Unique metrics are de-duplicated metrics. As these metrics are de-duplicated, you may see different numbers based on the level at which you're viewing data.
For example, because of de-duplication, campaign-level reach may not match the sum of your ad-set-level reach, as reach refers to unique people reached. If ad set one reached person A and B, and ad set two reached person B and C, each ad set would have a reach of two and the campaign would have a reach of three.
It's possible that your metrics may not be adding up because you're looking at unique metrics that may rely on sampled data. Learn more about sampled data and how it can affect the metrics you see.
Sampling may cause differences between the sum of your breakdowns and your summary row total
When you select Breakdown above the metrics table in Ads Manager, you may notice a difference between the sum of those metrics and the total. This difference may be caused by sampling.
To see the total, look at the first line on the table:
In the example, it looks like the total reach is 29,700. Now, let's look at the breakdown.
To see the metrics that are broken out, look at the columns below the total:
When we add up the metrics in the breakdown for the sum, we see:
| Age | Reach |
|---|---|
| 13-17 | 9,723 |
| 18-24 | 18,314 |
| 25-34 | 1,531 |
| Total = 29,568 |
As you can see, there may be a difference between the total you see in the summary row and the sum of your breakdown. In order to report the correct unique metrics, Facebook has to calculate the unique value for each breakdown from that of the summary row. This may cause a difference in totals because the breakdowns and total metrics are all based on different sample sets.
Sampling may cause differences in summary row totals you see at different levels:
You may also see a difference in totals for unique metrics when you view results at different levels. Let's use the reach example again to visualise this.
When we view total reach at the campaign level, the number we get is 29,780:
When we view total reach at the ad set level, the number we get is 29,700:
The reason you may see a difference between the total at the campaign level and the total at the ad set level is because there are different sample sets at different levels.
Note: Differences in ads metrics may be caused by de-duplication, sampling or both.
* Nguồn: Facebook