CRO benchmarks

Benchmarks are the most misused numbers in ecommerce. This page is mostly about why, and what to use instead.

​

Why external benchmarks mislead

Published ecommerce conversion averages typically land somewhere in the 1–3% range, and that figure is close to useless for any individual store. Conversion rate varies enormously by:

• Category — a £15 consumable and a £2,000 sofa are not comparable
• Price point — higher prices convert lower and should
• Traffic source — branded search converts several times better than cold paid social
• Device — mobile converts lower than desktop, consistently
• New versus returning — returning visitors convert far higher
• Market and season

A store at 1.2% might be excellent (high-value considered purchase, cold traffic) or poor (low-price impulse category, branded traffic). The benchmark cannot tell you which.

​

Use your own baseline

The only reliable comparison is your store against itself. Record, before making changes: Then measure change against those, over comparable periods, ideally year over year.

​

Segment or learn nothing

A blended sitewide conversion rate conceals almost everything actionable. The useful cuts:

• Mobile vs desktop
• Paid vs organic vs direct vs email
• New vs returning
• By landing page template
• By category

A sitewide rate that is flat can hide mobile falling while desktop rises.

​

Rate of change beats absolute level

The useful question is not “is 1.8% good?” but “is it better than last quarter, on comparable traffic, and why?” Improving from 1.4% to 1.8% is a 29% revenue increase on the same traffic. Whether 1.8% beats an industry average is irrelevant to that.

​

On guarantees

Ecom Autopilot advertises 15% more conversions or money back. That is measured against your own baseline, which is another reason to record one properly before starting. Confirm current terms at purchase — see plans overview.

​

Related

• CRO overview
• A/B testing