How accurate is your online retail data? In truth, 100-percent accuracy will always be hard to come by—or impossible. That doesn’t mean you can’t have confidence in your data and glean actionable insights from it.
One key way that you can increase your confidence in the quality and accuracy of your data is through product matching. This is defined as being able to identify one of your products on your website and find the same product on another website. Then, the two can be compared. Matching is important because it begins to build out a complete data set. For actionable insights, you need a full picture of the competitive landscape and that can begin with product matching.
However, finding a match isn’t always possible. Sometimes, that comparable product isn’t out there. Or, it isn’t located due to a lack of standard product identifiers in your industry or another reason. When this happens, your match rate—the number of matches divided by the number of items attempted to match—can drop.
Why Can Match Rates Drop?
It is completely normal for match rates to change over time. As noted above, one reason why is because there is a lack of standardized identifiers across similar products in your industry. If each brand uses different terminology, it can become a challenge to accurately compare and contrast products.
There are other reasons, too. Due diligence must be performed before matching can begin to ensure identifiers are accurate and present on all targeted websites. This initial data set must be comprehensive, or else like products can be missed in the matching stage. Product information that is often analyzed includes UPC, brand, model, product title, image, and more.
Another common hurdle is blocking. Many retailers block pricing and other product information on their sites until shoppers either add the item to their carts, create a user account, or complete another step. This can prevent website crawling from recording an accurate match. At worst, match rates can drop simply because a competitor site is inconsistent or even broken. Spelling discrepancies and errors, for example, can hinder matching.
Why Monitor for Drops in Match Rate?
Given that drops in match rate will happen, you may be asking what the value is in tracking these fluctuations over time.
At Wiser, we monitor match rate closely for several reasons. The simplest is because it ensures that the data extracted is accurate, comprehensive, and actionable. If products aren’t being matched, there is little value in analysis.
Beyond that, match rate drops can be illuminating. This has to do with the reasons for the drops. When a drop is first noticed, the objective then becomes answering why. If the reason is because of inconsistent product titles across websites, the data extraction process can be refined to increase the match rate.
Finally, match rate drops speak to a larger consideration with data extraction: quality over quantity. If the matching process is too loose, with too broad of parameters, that can provide a high number of matches but a decrease in accuracy. On the contrary, strategic matching can yield lower matches but a higher confidence in accuracy.
Do you want accurate data or a lot of data? Your match rate can help determine which you are getting.
Overall, this is why it is important to monitor match rate drops and not be too concerned with fluctuations. Instead, use those fluctuations to refine the matching process and have a dialog with your data provider. Use match rates as a conversation starter to improve transparency between your provider and you, as this insight can help you better understand the accuracy and value in your data.
If you’re interested in more information into the challenge of online data extraction, download our eBook Solving the Challenge of Online Data Collection in Retail today. Or, reach out to us and let us know your questions.