Targeting

Course correct your planned campaigns to meet its objectives

What is Targeting?

As explained in the main page for Improve, Targeting is one of the most unique algorithms that breaks the black box of audience platforms and gives directions to the user on how to optimize the interests and first-party audiences for an ad set. It also suggests adding more interests or audiences that have higher potential to perform based on the uniqueness and efficiency of the audiences in relation to the segments that are already performing for the ad set. In this article, we will list down all the options that you can use to get a clear view of the performance of your ad set and hence your campaign.

How to get to Targeting?

When you arrive at the improve section, you can see the hierarchy of account, campaigns and ad sets. However, you will also notice that this hierarchy is divided between Investments and Audience Resonance. The purpose of this division is its functionality. Since investments in the campaigns are normally set at the campaign level, campaigns are a function of Investments, while Adset and Ads are a function of Audience Resonance. To know more about how Investments work, you can refer to our article.

To access Audience Resonance, you cannot directly go to Adsets. If you do, it will redirect you to Investments. The only way to get to Audience Resonance is to click or select Campaigns. If you click on the campaign name, you will be taken to a list of ad sets within the campaign. If you select more than one campaign, then you can click on the Ad set tab and you will see a list of ad sets within the selected campaigns.

Since targeting is set at an Ad set level, you can click on any of the ad set names and it will take you to the next screen that has three sub tabs called Targeting, Copy (Creative) and Visual (Creative). You can learn about the creative tabs in another article. For now, we will focus on Targeting.

Understanding Audience Resonance

Note: To maintain the security of the client data, the screen shots shown have dummy data in it. Some data points may not add up.

To the right of the tabs, you have the option to select campaigns and ad sets. This option helps you switch between campaigns and ad sets without having to navigate through the entire hierarchy.

Current Targeting

Once you select an ad set, consumr.ai pulls out the demography of the ad set and displays it in this card.

If you build a campaign using consumr.ai, we give you recommendations for age and gender. However, while optimizing, this section only displays the existing age and gender targeting. If you want to change it, you can do so by clicking on the edit button. However, you cannot change the country from this section.

Learn More

The 'Learn More' feature may not have gotten a lot of real estate in this UI, but it is an extremely powerful feature to have. As the name suggests, this feature allows this ad set to learn from other campaigns and audiences, even if they are not related to this ad set or campaign.

While this feature is a powerful one, it needs to be used with a lot of caution. Caution in selecting the right audience or campaigns. You may have a campaign that is performing well, but if the objective or the communication of that campaign is different from the one you are optimizing, it may lead you down the wrong path. So choose carefully.

As shown in the screenshot above, once you select a campaign, it would take some time for the platform to come back to you with the learnings. The results are then shown in the next section called Interests.

Interests

This section shows the interests that exist as a part of the ad set targeting. As you can see, it will give the name of the interests and then give you the reach score and the conversion score for each interest. A reach score indicates the potential of an interest to reach the audience and the magnitude of it. A conversion score indicates how likely the interest is to give you conversions. To the left of the table, if you see a green thumbs up, it means it's a good interest to have and that you should keep it. If you see a thumbs down colored in red, it means it's recommended to remove this interest from this ad set.

Please note: a red thumbs down does not mean that this interest will not work for another ad set or campaign. This recommendation is only confined to the ad set you're evaluating.

If you click on learn more and select an audience or a campaign to learn from, this interest list will change to the interests from the campaign you want to learn from. You will get the same recommendations for the campaign you want to learn from.

At the bottom, you'll see a small check box to show insignificant interests. It's normally unchecked, because it's not ideal to see insignificant interests while making a decision and also because an insignificant interest would not really hurt a campaign since it may not be delivering. Check it, and you will be able to see all the interests, even the ones that do not have any reach. It is recommended to shut off the noise and focus on what is being delivered.

Real Time Reach Estimates

As the name suggests, this card shows you real-time reach estimates based on the changes that you make. The recommendations throughout this platform will point at areas that may be working well or not. Based on these recommendations, you may choose to add or remove certain audiences. This card would instantly show you the impact of that change. This allows the user to decide if the change in reach is acceptable or not, based on the business objectives or the aggression levels that the campaign needs to perform at.

Audience (Recommendations)

Audience recommendations primarily refer to first party audience recommendations. consumr.ai’s algorithm can understand the current targeting segments of an ad set and find first party audiences that are already built as part of the account and pick the ones that have a higher uniqueness index with the current targeting. By default, it does not give any recommendations unless you click on the button “recommend". Once you do, it may take a few seconds to get these recommendations in real time.

As shown in the screenshot, it takes you around a minute to get these recommendations. On a good day, you may get it in 30 seconds as well.

As shown in the screenshot above, the recommendations are now ready. You can check the boxes to the left of these audience names and add them to the cart.

Note: The names of the audiences have been redacted.

Interest (Recommendations)

The same is the way of working with the interest recommendations. Click on the recommend button and in a few seconds, interest recommendations will pop up. You can select the ones you want to add to your targeting.

Change Review

All the additions and removals will appear in this card. Think of this as your ecommerce cart where you store all your items before you finalize your purchase.

When you are satisfied with your changes, click on the "Calculate impact" button to see the final estimation of the impact.

Potential Impact Calculation

This card is pretty self-explanatory. It shows you the impact of your change on your reach score and your conversion score. It also shows you the reach score before and after the change. Since audiences work on overlaps, in the above screenshot, what shows as after change in reach score as -5.28, shows in the impact as 98.27 percent. Also, what you need to note is that your reach score before and after change are indexed numbers and not percentages. The delta and impact are calculated in percentage. So, please do not be alarmed when you see numbers like these. The message below in red raises a caution that finalizing this change may impact your reach significantly. It's not too late. You can still scroll up and reverse your changes and recalculate the impact until you're comfortable with the final numbers. Once you are, click on accept changes and these changes will be pushed to the ad set through the APIs.