When we think about the cost of poor customer support, we usually think about churn. An unhappy customer leaves, maybe writes a bad review. That's painful, but at least it's visible. What most e-commerce brands completely miss are the invisible costs — the revenue that never materialises because support was too slow.
The pre-purchase drop-off
A significant portion of customer queries arrive before a purchase is made. Shoppers have questions about sizing, compatibility, delivery times, or return policies. If those questions go unanswered for hours, the customer doesn't wait — they buy somewhere else. Studies suggest that 53% of online shoppers abandon a purchase if they can't quickly find an answer to their question.
The return that didn't have to happen
Many returns are avoidable. A customer who buys the wrong size because nobody answered their sizing question is a return waiting to happen. Faster, more accurate pre-purchase support directly reduces return rates — which is one of the most impactful levers in e-commerce unit economics.
Putting a number on it
- →Average e-commerce return rate: 20–30% of all orders
- →Cost of processing a single return: €8–€25 depending on product category
- →Revenue lost to pre-purchase abandonment due to unanswered queries: estimated 5–15% of potential sales
- →Lift in repeat purchase rate from high CSAT: up to 17%
For a store doing €1M per year in revenue, a 5% improvement in pre-purchase conversion from faster support is worth €50,000. The math on investing in AI support infrastructure becomes obvious very quickly.
What good looks like
The brands winning on support aren't necessarily the ones with the largest teams. They're the ones who have removed the bottlenecks — using AI to handle the repeatable, data-driven queries so their human agents can focus on the conversations that actually require judgment and empathy.