TikTok Auto Targeting

Not long ago, TikTok released a new targeting option in its ad set setup. We are talking about the Automatic Targeting feature, and of course, we got our hands dirty and tested it out in some of our client’s accounts.

So, what’s this about? It’s giving the algorithm the freedom to automatically show our ads to what it considers “relevant users”. It will optimize our targeting settings according to our advertising objective, our ads content, and it also takes into account past campaigns data, among other details – which TikTok does not specify.

You may think, ”Well… this is pretty much the same as the Broad targeting we used to have, isn’t it?”, and we won’t lie, it is pretty close to it, but with slightly different available options to define the audience, we want to target. Let’s jump into their differences!

Broad vs. Auto Targeting

As a reminder, in the targeting ad set options we have:

• Demographics
  • Location
  • Languages
  • Age
  • Gender
• Audience
  • Inclusions
  • Exclusions
• Interests & Behaviors
  • Many Interests & Behaviors options we can specify
• Device
  • OS Versions
  • Device Mode
  • Connection type
  • Carrier
  • Device Price

By Broad, we usually mean a big audience size, but this could have different combinations such as certain age ranges, limitations in the minimum OS versions, or even a gender restriction. However, the audiences are kept broad by choosing many of them or even not restricting the algorithm by choosing any.

I also mentioned in the above bullet points “location & languages”, those are the only options that we still have in both Custom Audience (where we are able to build a Broad audience) & Auto Targeting features.

Test structure

We prepared a test in different languages and both OS, Android & iOS, in order to get good quality and strong data to draw our conclusions. The structure is detailed in the chart below:

Results

Let’s deep dive into the juiciest part, the analysis and results. For a better understanding, we will isolate each country and OS.

France Android

• The test ran from the 21st Feb to 7th March 2022
• We were optimizing towards Tutorials Completed
• The Auto Targeting ad set got a cheaper CPI & Cost per Tutorial Completed, -30% & -34% respectively
• The traffic KPIs were the key for the above results, by having cheaper CPM & CPC + also higher CTR in the Auto Targeting ad set
• The CVR was pretty much the same, with the Clicks to Installs slightly favoring the Broad audience while the CVR from Installs to Tutorials Completed was a bit higher in the Auto Targeting one

France iOS

• The test ran from Feb 21st to March 7th, 2022 but as SKAd conversions take from 24 to 72 hours to be reported, we considered from the 24th Feb to the 7th March
  • Bear in mind that the Broad ad set had been active before we launched the Auto Targeting one, so from Feb 21st to Feb 23rd, some conversions from the previous spend were included in the Broad ad group, which wasn’t the case for the Auto Targeting as it was only activated on  Feb 21st
• We were optimizing towards Install event because we haven’t enough Tutorial Completed events
• In this case, we have similar behavior as in the France Android test regarding the traffic KPIs, where we see better numbers on CPM, CTR & CPC, however, the key is the CTR, which is +66% in the Auto Targeting ad group.
• If we take a closer look at the CPI, we see that it is a bit cheaper (-10%) in the Broad ad set, which can be influenced by the previous spend as we explained above. The CVR from Clicks to Installs explains the low CPI.
• Now, if we focus the sight in the Tutorials Completed, we see that the Auto Targeting, once again, is the winner, with a 21% CPTC cheaper due to a higher CVR from Installs to TC
• As we were optimizing towards installs and that is the unique metric that favors the Broad ad set, we could conclude that this one is the winner, however, all the rest of KPIs say that the prize should go to the Auto Targeting one. That is why, with this data, we can’t pick a real winner, and we will check the results in a few weeks.

Spain Android

• The test ran from Feb 16th to March 7th, 2022
• We were optimizing towards Tutorial Completed events

We have a clear winner, the Auto Targeting is clearly better than the Broad one in all KPIs

Germany Android

• The test ran from Feb 16th to March 7th, 2022
• We were optimizing towards Tutorial Completed events
• In this case, we pushed the Auto Targeting audience way more than the broad one to see its behavior with a higher spend
• Traffic KPIs: the Auto Targeting ad set won everywhere
• Conversion KPIs: We also have a big win on the Auto Targeting side. Looking at the upper funnel, it drove a cheaper CPI but the same CVR from Clicks to Installs, however in the mid-funnel event, Tutorials Completed, it shows a -56% CPTC due to a higher CVR from Installs to Tutorials Completed

Conclusions

We definitely recommend launching and pushing the Auto Targeting setup in your TikTok campaigns for Android apps. On the iOS side, we encourage you to test it and see what you get, as we still can’t pick a winner.

As we always say, monitor your audiences and check your results to make your own conclusions and takeaways, as it might differ from one app or industry to another.

For more interesting insights, check out our latest articles!