Difference in quality of users cast by 1-day attribution and 7-day attribution facebook business ads

Conclusion: Facebook has the option of 1-day or 7-day attribution when pitching AEOs, usually pitching 1-day comes with higher quality users (short-term gains).

Details of the change of Facebook ad attribution window period

This phenomenon is actually reflected in the small loans, games, social inside, usually run 1 day attribution recovery will be better (short-term), of course, there will be exceptions (such as the product conversion rate is particularly low, 1 day attribution conversion less lead to learning not out).

Reason: Usually, 1 day to produce conversion (such as payment, or lending) of the user, is the more urgent need for the product, this part of the user itself relative product quality is higher.

7 days to attribute the user to the product demand is not so urgent, including the subsequent payment or lending behavior are not necessarily active, recovery will be worse.

Exceptions:

1, branded products, users are not rushing to the demand, but rushed to the word of mouth, the brand over, although the short term is not necessarily immediate demand, but the future demand, these products 1 day and 7 days attribution may not have much difference.

2, this is a bit sophomoric feeling, but it does exist, such as microfinance (high-quality long-cycle microfinance) users are in a hurry, but the more likely that the user is a “poor” user, not necessarily a long-term can “cut” the user. Or similar to the BNPL, credit card and other types of products, may be more demand for long-term user value products, rather than just look at short-term data.

3, just-demand products, such as taxi, takeaway, in fact, not necessarily (can only say not necessarily, may be wrong) 1-day and 7-day attribution on the user how much difference between the impact of the demand is there, just not immediately in demand.

1-day attribution and 7-day attribution filter out only the more “urgent” users, which has an impact on the short-term data.

Does my product run 1-day attribution or 7-day attribution?

Does the product run for 1 day or 7 days… This is a bit of a randomization, I tend to 7 days, especially the conversion rate itself is not high products, 7 days more to put together the conversion can learn more stable.

Conversion rate is particularly good, it does not matter that the day conversion more, this time to run a day may be better.