Case: ID-FREE
Cookie-based vs. ID-free Targeting
Published: 2020-02-26
Abstract
To replace traditional cookie-based targeting with ID-free targeting options is possible without losing campaign performance. The ID-free methods performs on par or better than the cookie based targeting methods.
Background
As the cookie-based targeting options are diminishing we are looking into other ways of reaching the right consumers. During the start of 2020 we have conducted a live test of our new ID-free targeting solution, with real customers and real campaigns. This is a case study of our findings.
Method
We have measured two separate display campaigns with the same banner materials, where the only difference is the targeting settings. The campaigns were deployed through our DSP. We targeted iOS-devices with the ID-free targeting option. As over 40% of the devices in the Nordics run on iOS, they are a substantial blank space in targeted display advertising. As iOS devices don’t allow for 3rd party cookies we used Click Through Rate (CTR) to compare the campaign performance, even though View Through Rate (VTR) is our standard method of measurement. We consider VTR to be more representative for the campaigns impact on the targeted market segments.
Targeting
Campaign 1 had a target to reach a specific socio-demographic attribute and the banner materials message reflected this attribute. This attribute was targeted through 3rd party DSP-cookies for Campaign 1A, and through ID-free targeting of the same attribute for Campaign 1B.
Campaign 2 also targeted a specific socio-demographic attribute and also here this attribute was reflected in the banner materials for the campaign. Campaign 2A was targeted through 3rd party DSP-cookies and Campaign 2B was targeted through a combination of ID-free and 3rd party cookies.
Disclaimers
- As we compare cookie-based vs. ID-free/iOS targeting we need to understand that the use of different operating systems might create a slightly skewed view of the users based on underlying factors such as socio-economic factors and behaviors that are linked to this. For the purpose of this case study we will disregard those factors.
- Tracking of unique users through DSP is reliant on 3rd party cookies, and we have observed that the targeting of iOS devices generates a very low number of impressions per unique user, even at the same frequency cap as the cookie-based targeting option. This is most likely due to the fact that 3rd party cookies are cleared after a session or prevented from being placed. This, in part, puts the frequency capping out of play for the iOS-campaigns. In this case study we have disregarded this effect.
- Mobile devices often have higher CTR than desktop devices, this may be due to a higher level of “miss-clicks” but may also be due to a different type of user experience. For the purpose of this case study we have disregarded this as well.
Data
Below is the data collected for the case study.
1A – Cookies | January | February |
Impressions | 916009 | 702435 |
Clicks | 1220 | 859 |
Cookie CTR | 0,133% | 0,122% |
1B – ID-free | ||
Impressions | 228302 | 238570 |
Clicks | 353 | 362 |
ID-free CTR | 0,155% | 0,152% |
2A – Cookies | January | February |
Impressions | 326970 | 360297 |
Clicks | 422 | 359 |
Cookie CTR | 0,129% | 0,100% |
2B – ID-free + Cookies | ||
Impressions | 342163 | 432231 |
Clicks | 430 | 525 |
ID-free CTR | 0,126% | 0,121% |
Summary
Campaign 1
We can see that the ID-free campaign targeting outperforms the cookie-based targeting by an average of 20,5%. This doesn’t necessarily mean that ID-free is better than cookie-based targeting, especially if we take the disclaimers above into consideration. But it does shows that the CTR-performance is at least same or better for the ID-free targeting option.
Campaign 2
Here we observed the cookie-based targeting with cookie + ID-free targeting. This observation shows the campaign performing on par with each other in terms of ad performance, having almost equal CTR for January and the ID-free + cookie-based targeting outperforming in February by 21%.
These results have led us to start recommending the ID-free targeting option to our customers and use ID-free in combination with cookie-based targeting to reach the “hard-to-target” devices (iOS) with maintained campaign CTR performance.