Sometimes, data quality is just a number. The client in this case, a well-known retailer in the Netherlands, had an overall data quality of 98%. Almost all tags functioned correctly. But the ones that didn’t had quite an impact on the business, especially on the conversion attribution.
Besides web analytics, many companies also use conversion attribution tools, to determine which campaigns (banners, ads, and emails) lead to a conversion. So did this particular client. They also were aware of their data quality, which was 98%. Not a bad score of course, but the remaining 2% was in a very specific area and had quite an impact. Because after implementing Qmon, we found that the campaign tags were not always set
30% of campaigns not measured
A campaign tag tells you where a visitor or an order came from. For instance, the retailer in question lists his products at an independent comparison website. If a customer decides to buy a product on this website, the order process takes place on the retailer’s own website. When a visitor lands at the retailer’s website, a tag has to be set so the conversion attribution tool can determine the origin of the order. If this tag fails, the tool assigns the order to the wrong campaign. Or to no campaign at all. Validation with Qmon turned out that a stunning 30% of all campaigns were not measured in this case.
The errors in the campaign had several implications. First of all, the comparison website has to be paid for generating traffic. But the number of orders they claimed to have generated, did not match the data of the conversion attribution tool nor the web analytics tool. This misinformation triggered a discussion about the number of orders to be paid. Furthermore, the retailer made budgeting and optimisation decisions based on wrong data. The products on the comparison website did in fact perform very well, but that did not show from the data of the conversion attribution tool. So investments might have gone to the wrong campaigns.
Why Qmon could find this issue
The failure of the campaign tag did only occur in specific cases. Manual, scenario and crawler tests did not detect it, since they only tested a very limited amount of traffic. Errors like these can only be found by monitoring 100% of the website traffic, which is exactly what Qmon does. Based on Qmon’s findings, the error could be solved. A very welcome improvement for the retailer himself, but also for the affiliate (the comparison website) and the conversion attribution tool:
– Affiliate: eliminate differences in the amount of orders, ruling out discussions over the payments.
– Conversion attribution tool: an error in the tool was found and solved, improving the tool itself.
– Retailer: having correct data for optimisation and budgeting decisions, and no more tiring discussions with ad vendors. Optimal decisions have direct impact on the bottom line results, in terms of the number of conversions and the cost per conversion