Why data-driven attribution?

There are many types of attribution model. Which is precisely why it matters to proceed carefully when choosing one and pick the model that suits you best. Only by testing a model will you find out whether it works as you need it to. The numbers always show the truth.

What is DDA (data-driven attribution)?

DDA stands for the data-driven attribution model. It is a suitable model precisely for accounts with a sufficient number of conversions. It uses machine learning technology to determine how much credit to assign to each click on the user's path, so as to reward the keywords that appeared earlier on the click path. In practice it means it helps determine how much credit to assign to each click on an ad promoting a product the customer is most likely to be interested in.

A conversion need not follow immediately. But by using an attribution model you can find out that people who click an ad become real customers sooner or later.

How does it work?

The data-driven attribution model differs markedly from the others. By drawing on the conversion data in the account it shows the genuine contribution of each click on a search ad. It compares the conversion paths of customers who reached a conversion with those who did not. From those comparisons it determines which touchpoints matter most. The results take into account the number of ad interactions, the order they appeared in and the creative. The algorithms calculate which clicks matter on each conversion path.

If your account has reached at least 3,000 ad interactions and at least 300 conversions in 30 days, you meet the conditions for using data-driven attribution. Moving to this model has several advantages:

• it assesses every step on the conversion path,
• it works automatically with bidding,
• it works even on very short conversion paths,
• it is quick and simple to implement.

If you have decided to try the model, we have prepared a video guide on setting it up easily in your Google Ads account. The choice of model is entirely yours. There are cases where another type of model suits better. But if you have enough of the data needed, do try data-driven attribution.

Rules-based attribution models

Data-driven attribution draws its information from the account's history. If you do not have enough site traffic, you probably will not be able to use it. That does not mean you cannot try another model, though.

Choosing a model relates to the goals you have set for your Google Ads account. Some strategies are more conservative, others oriented towards growth or efficiency. For a better understanding, here is a clear table of the specifications.

The type of model you choose also shapes the relationship between you, the Google Ads account and the customer journey. Attribution helps you see keyword performance from the right angle. Keyword performance can differ from model to model.

A more aggressive model (first click, say) shifts the figures so as to reward the keywords that appeared earlier on the click path, while a more conservative one (last click, say) rewards the keywords that appeared later. So when updating the model in use we recommend reconsidering your account's performance goals as well.

Whichever attribution model you choose, we trust it will be the right one! :) If you need advice on it, do not hesitate to get in touch.