Fixing the Sales Leaks

A 500,000+ session e-commerce dataset showed clear customer drop-off points — but the most useful finding wasn't where customers leave. It was what happens when they engage too much.

The store wanted to understand why interest doesn't translate to purchases. I traced the customer journey across product views, cart-adds, checkout starts, completions, and broke it down by channel, device, and the active A/B variant.

Two big drop-off points dominate: product page → cart-add, and cart → checkout. Most shoppers are interested but never convinced. Many start the buying process but never finish.

The less obvious pattern: customers who view many products or add many items to their cart actually convert lower than focused shoppers. The behavior reads as overwhelm, not engagement. More options, more friction.

A few channel and device findings worth noting:

• Email and Paid Search bring the highest-intent traffic. Organic and Direct convert poorly — cheap to acquire, expensive to convert.
• Mobile users convert better than desktop on both cart-add and completion. Mobile here isn't lower-intent; it's higher.
• Variant A in the A/B test didn't increase cart-adds, but did slightly increase completed purchases — likely more confidence at checkout.

The action: simplify product choice, lean into email-driven nurture, and treat the checkout flow as the highest-leverage area to fix.