Don’t let your North Star metric lead you astray. In e-commerce, stores tend to prioritize conversion rate as a barometer of overall performance. While a data-driven approach is critical in improving a shop, over-indexing on conversion rate could actually reduce total revenues and profits.
A singular focus on any metric creates perverse incentives and promotes gaming that metric. On paper, conversion rate optimization (CRO) only concerns the percentage of prospects that become actual customers. Conversion rates don’t encapsulate how valuable each sale is to the business or whether customers are satisfied with their purchases. CRO with no regard for profits pushes both your developers and your AI algorithms to recommend cheaper products so long as customers are more likely to buy them.
Exclusively selling cheaper products or ones that are more likely to yield conversions has unintended consequences:
- You leave money on the table. Suppose you let AI do product sorting on your webpage according to predicted conversion rate. Customers, including ones with large budgets, are less likely to discover and purchase your more expensive items.
- You risk reducing customer satisfaction. High-budget customers might get a worse experience with a cheaper product than they would have otherwise. Although you might initially succeed in driving sales, that means very little when unsatisfied buyers make many returns.
- You create UX friction and might cause churn. For example, say your store sells a product that comes in small, medium, and large SKUs. Poorly calibrated AI would pick a bad default option for users, making it a nuisance for a user to adjust later. A smart system would intuit the appropriate sizing option for the customer to make for a smooth user experience.
The nuanced solution is to factor in multiple metrics when optimizing your store. It’s important that your AI tools have a similarly nuanced objective function. With DataMilk, your store can incorporate smart components that intelligently assess a customer’s price range and recommend appropriate products to optimize both conversions and your bottom line.