As part of improving analytics, setting up online conversions allowed us to increase efficiency by 22% over the analyzed period. However, this is only part of the result. In addition to setting up tracking for FTD, KYC, and other metrics, we also implemented predictive modeling .
Problem
There was a 24-day window between signup and FTD, which is a very long conversion window. Given that the FTD cost was over €1,000, waiting 24 days to understand the effectiveness of the campaign was impractical. This created the risk of significant costs without the ability to make quick adjustments.
Decision
We developed a prediction model during the KYC phase. The model estimated the probability that a user would complete an FTD in 24 days. The egypt mobile database prediction result was given as a value between 1 and 1000, which was passed to Google Ads in the format value . Based on this data, we used bidding strategies such as Target ROAS or Maximum Conversion Value to optimize campaigns.
Results
- Conversion window reduction: company-wide: from 24 to 18 days. Within Google Ads: from 23 to 18 days.
- FTD quantity growth: FTD price decreased by approximately 22%.
- 90-day testing: Comparison of the period before and after the model implementation showed significant improvement.
However, the testing did moderation and problems not take into account external factors, such as market characteristics, that could have influenced the results. Therefore, some of the positive changes could have been due not only to analytics, but also to market conditions.
Licensed countries on Google
The next problem is slovakia business directory moderation, which is difficult due to the specifics of the Forex industry, which involves speculative trading.