A brand-lift study gives you insights on the effectiveness of your campaign in relation to advertising impact on brand equity. All studies will receive the following metrics through comparing the test (exposed) versus control groups:

• Estimated incremental number of people who remembered seeing your ad
• Cost per estimated incremental person who remembered seeing your ad, based on your budget
• Percentage point lift, which is the difference in the percent of people who recall seeing your ad in the exposed group versus those in the control group

Estimated incremental number of people, cost per estimated incremental person and percentage point lift are also available for other metrics such as message association, favourability and purchase intent. Not all metrics are available for all studies. Please work with your Facebook Account Team to discuss which metrics are best aligned with your marketing needs.

There are many factors that influence how much your campaign can move brand metrics. Some factors include:

• The layout of your ad and visual creative that you use (images, video, text and links)
• The audience or people who have been targeted to see your ad
• The number of people that your ad reaches and how many times they see your ad
• The perception of your brand or unconscious biases towards your brand among study participants

The success of your brand-lift study can be assessed by a positive percentage point lift in desired responses, which is measured by calculating the difference between the test (exposed) and control groups' responses. Generally speaking, the larger the difference between the response rates in test (exposed) and control, the better. Comparing the size of the difference with regional or vertical norms can also help you assess your performance. (Please refer to the Performance Summary section below for more information on benchmarking.)

For example, In the case of ad recall, a large percentage point difference between test (exposed) and control is a sign that your ads resonated with people.

After you receive your brand-lift study results, you can assess your campaign to see what strategies you may want to change to improve ad performance on your next campaign. For example,

• If you use different creative on a new campaign and those brand lift results come back with a higher percentage point lift on the metrics that you've chosen to test, you might consider using that creative on future campaigns.
• You might adjust your campaign targeting based on your audience's TV viewing habits. For example, if you saw a large percentage point lift among people who exhibit low TV viewership behaviour, you may want to reallocate your percentage of budget to target more low TV viewers in your next campaign.
• Benchmarks, or norms, can also help you understand whether you should continue with a campaign strategy, as compared to similar campaigns in your vertical or region.
• Additionally, you can benchmark past campaigns with the optimisation best practices provided in your brand-lift study results. For example, you may try selecting mobile feed as a placement option or try aiming for 1-2 impressions per person a week to improve your next campaign.
• You can continue to test strategies and compare studies to optimise campaigns.

Potential reasons for non-statistically significant results can include:

• Your creative didn't effectively communicate the intended message, brand or product. (Here are some tips on creative best practices and Facebook tips on engaging creative for mobile video ads.)
• Other campaigns running at the same time for the same brand or product affected the test (exposed) and control groups.
• Your poll question wasn't focused enough on the brand, product or message that you were advertising. For example, You may have asked an ad recall question for Jasper's Market, but the campaign was about Jasper's Market Home Delivery Service. For better results, the question should relate to Jasper's Market Home Delivery Service.
• The budget, bid, runtime or creative was changed during the study.
• The frequency of the campaign was too low for people to recall the ad.
• Your message didn't resonate with your target audience or your target audience already had a bias or affinity towards your brand.

The Performance Summary section provides norms compiled from other Facebook brand-lift studies similar to yours that share the same region or vertical.

Median results are compiled across all placements and are updated daily to reflect all lifetime results while a study is running. When a study ends, the norms in the report remain frozen for consistency.

This allows you to see how your brand-lift study compared to other studies in your region and vertical for the metrics that you've chosen.

The Lift results section provides a comprehensive overview of your brand lift results, including the question(s) used in the brand poll.

Three lift result numbers will be available for review:

• Estimated people who took the desired action (i.e. people who recall seeing an ad, people who remember a brand, etc.)
• Cost per incremental person to take that desired action
• Percentage point lift (the difference between the people in the exposed group and the people in the control group who took the desired action). In the example below, it's the percentage of people reached who recall seeing your ad as a result of the campaign.

The Study details section summarises the number of people polled and the equivalent desired response rate (in both percentage points and actual percentage) for the exposed and control groups.

The Campaign details section gives a short summary of key campaign metrics such as reach and impressions.

Within the Responses Breakdown section, you can see an overview of all exposed and control group responses and gain insights into which demographic group (age, gender, TV viewership, etc.) remembered your ad the most or least.

TV viewership

Facebook has partnered with Nielsen to compare your brand-lift study among three different TV viewership habits – light, moderate and heavy. The TV Viewership tab provides TV viewership data for the exposed and control groups, which is modelled based on nationally representative Nielsen panel data. This data allows you to determine how your ad is performing among people with certain TV viewing habits.

Video Views

The Video Views tab shows the impact of video view length on ad recall. This section shows which view lengths drive the most impact for an advertiser's campaign.

Within this section, you will see tips to help you optimise your campaign and get the most out of your brand-lift study using strategies such as including mobile feed as a placement option, aiming for 1-2 impressions per person a week and choosing auto-bidding.

By identifying particular age groups and genders with the highest lift, you could tailor upcoming campaigns to a specific demographic. For example, a campaign with a low ad recall lift in a younger demographic group might mean that the ad didn't resonate with that audience. For upcoming campaigns, you might consider using different creative catered towards a younger demographic.

By analysing the impact driven by different view lengths of your video ad, you get a better understanding of what view lengths offer the best value for driving desired responses. Here are some examples of conclusions that you could draw:

• a high amount of ad recall when the video ad was watched from 0-3 seconds might indicate that people are watching the first few seconds of your video. Therefore, you may want to incorporate more important aspects of your message or brand into the first three seconds of the video.
• Long video views generally have more impact per person, but limited scale. You may find that shorter video view lengths (under ten seconds) drive the majority of impact from campaigns.

  Here are some Facebook tips on engaging creative for mobile video ads.

Note: Video viewing performance results are modelled.

By comparing your audiences' TV viewership, you can see how your ad is performing among people with certain TV viewing habits and adjust your campaign targeting as needed. For example, if you saw a large percentage point lift with people who don't watch much TV, you may want to reallocate some budget to specifically target that type of person in your next campaign.

By comparing ad recall lifts and creatives, you may be able to see patterns for successful ad creative. For example, if you see little or no lift from your campaign, it could be a result of not sufficiently incorporating the advertised brand or product into the creative, or creative that didn't resonate with your target audience. On your next campaign, you could try running new creative that really incorporates your brand and is designed to appeal to your target audience specifically.

Here are some tips on creative best practices.