Facebook Streamlines Ad Testing & Campaign Optimisation
The ol' business adage goes, "what gets measured gets improved".
No one's quite sure who coined it; Peter Drucker, Lord Kelvin, Dumbledore?
But the sentiment is especially true within digital marketing, with online testing and optimisation being the lifeblood of effective activity.
However, within social channels, testing is somewhat convoluted and ad hoc, therefore, to make this a bit easier, Facebook has released a new "Experiments" section within Ads Manager. Experiments centralise all Facebook ad tests and results, making it simpler to tune up campaigns.
Rightly or wrongly, marketing spend is likely to get clipped over at least the next two quarters; therefore, organisations will need to squeeze a lot out of a little.
Hence, for those running paid campaigns, testing and optimisation become paramount.
Experiments combine multiple types of tests, formerly labelled as "Test and Learn" and "Split Testing", allowing businesses to A/B test campaigns and measure conversion and brand lift in one place.
Testing All Day Err Day
While many of these testing features existed previously, the majority are lightly used, in part due to convoluted and un-user-friendly nature of Facebook Business Manager and Ads Manager.
Facebook has effectively spent the past year trying to unbundle and streamline both, unlocking buried and useful features and minimising visibility for less material ones.
Within Experiments you can choose from several templates that help you measure your advertising and get recommendations on improving performance in a step-by-step manner.
Depending on the test that you're interested in creating, you can:
- Compare one or more ad campaigns to see which one performs best.
- Measure how your ads affect brand awareness or recall.
- Measure the incremental lift of your advertising or compare the incremental performance of multiple ad campaigns with a holdout.
- Determine the impact of new advertising strategies against your current strategies.
Experiments are designed to help marketers make better strategic decisions with more powerful testing tools. It's easy to get the right test up and running, fast. Just select the test, add information and Experiments will set up the test.
Facebook Ad Test Types
From the main Experiments user interface, Facebook leads in with four test types:
1. A/B Test
A/B tests allow marketers to compare strategies for creatives, audiences and other variables to determine what works and optimise the next campaign.
A/B tests can help marketers answer questions like: Which creative performs better? Which call to action performs better with this audience?
Automating campaigns makes this process more efficient for advertisers, but A/B tests are useful for comparing manual and automated optimisations and accessing additional insights about performance.
2. Holdout Test
When measuring ad performance, how can marketers confirm that the conversions took place because people saw an ad?
Holdout tests make it easier to understand the true impact of advertising by comparing conversion rates of people who had the chance to see the ads with those of people in a holdout group.
The holdout group is intentionally withheld from seeing the ads, which the test uses to calculate lift in conversion events like purchases.
3. Brand Survey
Brand surveys are key to understanding if brand campaigns are having the impact marketers are looking for on awareness, perception, recall and more.
Brand surveys poll the people who've seen an ad and poll a holdout group who hasn't. By aggregating and comparing results, the brand survey can better measure how ads impacted a marketer's brand objective, helping confirm they're investing in the right strategy.
4. Campaign Budget Optimisation (CBO) Test
The campaign budget optimisation test lets you use an existing ad campaign as a template to see how campaign budget optimisation affects your cost per result performance.
The test will copy your ad campaign and automatically create an A/B test – one with campaign budget optimisation turned on, and one with optimisation turned off, so you can easily compare performance between each strategy.
From awareness to performance campaigns, and everything in between. Experiments empower marketers to not only squeeze more results from reduced budgets but also importantly, provide a more explicit way to demonstrate business value generated for trigger happy CEOs and executive teams.
Check out Experiments now.