Measuring
Facebook Ad
results
Learn Facebook Ads in five steps
Step four - measurement
When you first start running Facebook Ad campaigns, your ads and account will enter the ‘learning phase’. This is indicated by ads showing a ‘learning’ (or ‘learning limited’) message within the reporting section.
This is a process Facebook undertakes to optimise ads and find the ideal people to target within audiences. Facebook’s AI is discovering which ad creatives (e.g. images, videos, text copy etc) resonate best with each audience. Generally, you should allow three to seven days before making changes to an ad. If you make changes too early, your ads will re-enter the learning phase and it may take longer to achieve results.
Video
Different people have different opinions regarding the learning phase, but 20 to 50 conversions is usually required for Facebook to understand the best targeting options and to then exit the learning phase. For example, it may take 50 purchases for Facebook to understand which users (i.e. their interests, demographics, behaviors etc) are most likely to convert in future.
The learning phase may have longer or shorter time periods depending on the conversion objective (e.g. 50 clicks to the website is much easier to achieve than 50 purchases). If ads are continually displaying the ‘learning limited’ message, this is not a huge issue as they can still continue to provide results.
Resource
Facebook ads can follow a particular lifecycle of slow takeoff (e.g. learning phase), quick improvement, good results and then a quick drop off in performance.
Sometimes it can take up to six months of testing before you are able to find which ad approaches work well. If, after sufficient testing time, ads continue to underperform then turn off the ads or ad sets. Don’t stay with an audience simply because they used to perform well in the beginning.
Poor ad performance and drop off can be caused by:
- Small audience size fatigue (i.e. seeing an ad too many times and becoming fatigued)
- Daily budget (larger budgets will reach more people faster but the audience will fatigue faster if small)
- Market activity (e.g. bigger advertisers start promoting during busy retail periods – increased competition means higher costs).
During highly competitive periods, you can focus more on warm and hot audiences of existing users and customers instead of cold audience targeting. This can provide a cheaper method of continually running ads.
Some eCommerce brands will spend heavily on ads during non-competitive periods to gather data and customer lists who can then be retargeted for lower costs during busier event periods (e.g. Christmas, Black Friday sales etc).
Metrics and reporting:
You can measure the performance of your ads through a variety of metrics and reports provided directly within Ads Manager or through third-party tools.
You can also view more comprehensive reports within the Facebook reports section:
Google Data Studio also provides the opportunity to visually display analytics and data from Facebook Ads performance. You can use templates from Supermetrics, or other providers, and simply need to connect your Facebook account to Data Studio to start viewing reports.
You can also use Google Analytics to provide insight into how ads are performing and how users behave once on your website:
If users visit your website and quickly leave, this may indicate a disconnect between what the ad offered/promoted and what the user actually received once on the website.
In some cases, such as promoting a blog article, a high bounce rate may be acceptable as users will read the article then leave. Ideally, you want to place internal links throughout your article to encourage users to visit other pages of your site.
Video
Frequency and ad relevance:
When running ads, you want to keep an eye on the number of times users see the same ad (known as frequency). Frequency scores which are too high can be dependent on individual businesses/industries and audience types. Some people find results start to drop off after a frequency score of 3 whilst others can still achieve strong results with a score of 12.
Generally, a frequency score of around three or above is getting too high for cold audiences and results may start to drop off or become too expensive (i.e. people have seen the ad before so ignore it next time, or it takes three impressions before it’s clicked). A warm audience (e.g. past website visitors) can have a higher frequency of around six to ten. Changing creatives can help to reduce frequency and keep ads fresh for users.
Ad relevance diagnostics can help you determine whether the ads you ran were relevant to the audience you reached. If your ads aren’t meeting your goals, you can use ad relevance diagnostics to understand whether adjustments to your creative assets, website experience or audience targeting could improve performance.
The diagnostics assess each ad’s past performance in the ad auction over the date range you’ve selected. Diagnostics include:
- Quality Ranking: your ad’s perceived quality compared to ads competing for the same audience.
- Engagement Rate Ranking: your ad’s expected engagement rate compared to ads competing for the same audience.
- Conversion Rate Ranking: your ad’s expected conversion rate compared to ads with the same optimization goal competing for the same audience.
Whilst Facebook provides a range of analytics and data, the main metrics to focus on are conversions and cost per conversion – if those are working well, all other details are minor.
Ad performance and conversions:
You should ensure you are actively monitoring link clicks vs number of conversions. If you are experiencing a high number of clicks but few conversions, this means the ads are performing but the landing page may not be inspiring/persuasive enough to have users purchase or sign up. Facebook ads have achieved their purpose of directing potential customers to the site, but if there are no actual conversions then the issue may lie with the website and not the ads themselves.
If your ads aren’t performing well, they may be the result of the below mistakes:
1. Wrong campaign objective selected
2. Too focused with narrow interest and demographic targeting
3. Audience size too small (under 100,000) for top-of-funnel prospecting campaigns
4. Products/services that don’t stand out enough (no selling features or customer benefit)
5. Not refreshing ad creatives often enough leading to ad fatigue
This concludes step four of learning Facebook Ads (measurement). In step five (improving campaign results), we’ll cover how to leverage other channels to improve the overall performance of your Facebook Ads strategy. We’ll also cover how the Facebook Power 5 approach, automated rules and other strategies can help improve your campaign performance.
Learn digital marketing in five steps
