How to Choose the Right Attribution Model 🧩

Selecting the appropriate model depends on the expected user behaviour and the type of action you want to measure.

For instance, if your focus is on engagement conversions or quick interactions on your website, the 1-day click, 0-day view attribution will be your best choice. This model allows you to capture only the most interested users, those who take immediate action after seeing your ad, avoiding artificially inflating the results.

On the other hand, if you are looking for non-commercial leads such as newsletter subscriptions or free content downloads, even for post-sale support, a 1-day click, 1-day view model would be more suitable. This model accounts for both direct interactions and views that, although they don’t result in immediate clicks, generate short-term interest and eventually convert.

Attribution models with short windows allow you to accurately measure the real value that engagement actions bring to your website. If we used broader attribution windows, the results could be inflated, and campaigns could lose weight in real conversions, especially if you are using multiple paid media channels and various campaigns with different objectives.

For most sales or commercial lead campaigns, the 7-day click, 1-day view attribution model is the standard. This model offers a broad window to measure the immediate impact of your ads without losing sight of those users who need a few days to convert after the initial click.

In the context of high-value products or B2B conversions that can take up to a month, the 30-day click, 1-day view model would be the most appropriate. This model is ideal for capturing considered purchase decisions that may require approval from multiple parties.

On the other hand, the 30-day click, 30-day view model is suitable for complex B2B purchase events and high-consideration products/services, regardless of whether the goal is sales or commercial leads.

Remember that you can adjust the attribution model afterwards if you find that user behaviour is not as expected.

What attribution model do you usually use in your campaigns? Have you considered adjusting your strategy based on these models? I look forward to reading your comments.