Facebook is still one of the giants of digital advertising, but as the years go by, the online environment becomes increasingly complex and it is more difficult to measure the results of campaigns.
To help marketers overcome this challenge, Facebook has launched Conversion Lift, an advanced tool to measure the real impact of Facebook and Instagram ads on sales. We tell you how it works and what it can contribute to your brand.
What is Conversion Lift?
Conversion Lift (called “Increase of conversions” in Spanish) is a solution aimed at measure the increase in conversions generated by an ad campaign, going beyond click-based measures.
According to the data handled by Facebook, 91% of users who could buy a product do not click on the ads. Also, as users use multiple devices throughout their customer journey, cookies only reflect part of the story; in fact, it is estimated that data based on them misses no less than 37% of conversions.
Therefore, Conversion Lift focuses on measuring the increase in conversions among users who have seen one of your ads, but they haven’t necessarily clicked on them. To offer greater precision, this system considers that there are three types of conversions:
- Users who have converted and have not seen any ads.
- Users who have converted and seen the ad, but would have become customers anyway.
- Users who have converted thanks to the ads. This is the increase in conversions that we intend to measure (also called “incremental conversions”).
Facebook recommends using Conversion Lift in combination with the classic attribution models, since the precision of this new option is higher but obtaining results may take more time.
How does Conversion Lift work?
Conversion Lift use a methodology similar to scientific studies, separating the users into the experimental group and the control group.
After identifying the audience we are targeting and the business objective we want to achieve, Facebook separates the audience into the experimental group and the control group randomly. Then it shows the ads only to the experimental group and measures the conversion results of both groups (either through Facebook pixels, an upload of data or events in the app). Finally, Facebook compares the conversions in the experimental group and the control group, calculates the increase in conversions in the people who have seen the ads and shows the results to the advertiser through the Ad Manager. For example, if 100 users have converted in the control group and 150 have converted in the experimental group, we would have a 50% increase in conversions.
Types of tests in Conversion Lift
The simple tests (or “single cell”) aim to measure the amount of incremental conversions achieved with a specific strategy and the acquisition cost of each of them. As we have just seen, the audience is divided into an experimental and a control group and the test is started.
Instead, pruebas multivariable (or “multi cell”) aim to compare different strategies with each other. In this case, therefore, the audience is randomly divided into groups for each variable and, within each of them, into an experimental group and a control group. In this way, we can know which strategy achieves more incremental conversions at a lower cost.
For example, we can use a multivariable Conversion Lift test to find out if we are more interested in launching ads with a strategy of optimization for conversion or optimization for clicks.
Types of reports in Conversion Lift
When viewing the results of your Conversion Lift tests, you can choose between basic and advanced reports.
The basic reports they include only the data observed during the experiment. They may be suitable if you think your audience is large enough that the results are representative and you don’t mind waiting a little longer to see the results.
Instead, advanced reports They include historical data from similar tests, which means they are available faster and, according to Facebook, more accurate. They are especially useful if you are working with relatively small audiences or if you need to have results in a short time.
Other reporting features to take a look at include:
- Increase in conversions by gender and age. Here you can see at a glance which groups are driving the most conversions, both in a graph and with a quick summary of the most profitable segment. Use this information to target specific audiences.
- Comparison between the increase in conversions and the conversions attributed by clicks. In this section of the reports you can see how many incremental conversions you have achieved and how many could be attributed to clicks with different attribution periods (1, 7 and 28 days). This way, you can easily rectify your attribution models.
Why use Conversion Lift?
Because it allows you to accurately measure the results of your ads on Facebook
Facebook and Google are the online advertising solutions to which marketers spend the most budget, so it is very important to be able to accurately measure what we are doing. With the Conversion Lift tool, you can answer questions like:
- How many conversions am I getting thanks to Facebook?
- Should I allocate more budget to this campaign?
- How should you balance your new customer acquisition budget with your remarketing budget?
- Which audience has the lowest cost for each conversion attributable to ads?
- What is the real ROAS of my investment in Facebook?
- Does my actual attribution model match the results of a controlled experiment?
Because it improves current systems
This system is a revolution from other ways to measure the impact of ads on sales, such as A / B testing and attribution based on the last click.
In the case of tests A/BThe problem is that they measure total conversions, but don’t distinguish between “base conversions” (which would have occurred anyway) and those that can be directly attributed to your ads. Thus, it is possible that if we compare two strategies with each other, one of them produces fewer conversions, but the conversions attributable to the ad itself are higher.
As for the systems based on the last click, the problem is that they do not cover the entire customer journey, which is becoming more and more complex. Thus, when comparing the real increase in conversions with those attributed to the last click, Facebook has found differences of up to 48% in the number of conversions and up to 79% in the cost per acquisition, which can produce large imbalances in the budgets.
Because it goes from CPA to CPIC
Until now, the star metric of online advertising was the CPA or cost per acquisition.
The CPA informs us of total sales from specific channels, both from current and new customers. Variables like period and attribution model can have a big impact on measured conversions, and can undervalue channels that are at the top of the funnel. Therefore, it is suitable for measuring paid media results on products with short life cycles or flash sales.
Instead, Conversion Lift proposes to move to CPIC, this is, the average cost of conversions that wouldn’t have happened if it weren’t for your ads. This metric is based on experimental design (which is considered a more rigorous methodology), includes the effect of long-term channels such as SEO, and can be used to correct errors in attribution models. Therefore, it offers us a more accurate way to measure net sales or conversions, optimize budget allocation, and test new formats and platforms.
Because it’s really cross-channel
The attribution models that we usually use in online marketing arose in an environment in which most of the activity of Internet users took place on computers. In this context, it made sense to measure user behavior through cookies, as we could keep track of the different points of contact of the user with the brand.
However, today we consume content through multiple devices, and mobile phones are gaining more and more prominence. In addition, users are increasingly rejecting cookies, and Chrome has already announced its intention to stop allowing them.
Instead, Conversion Lift allows you to measure the impact of a given channel on conversions, regardless of the devices a user is usingor. In addition, it allows taking into account both online conversions and those that occur in physical stores, since the advertiser can upload their own data to the system.