Conversion Rate Calculator

The arithmetic

Conversion rate is conversions divided by sessions: orders ÷ sessions × 100. Sessions rather than users is the usual convention in ecommerce, and the two produce different figures because people visit more than once before buying.

The denominator choice matters more than it sounds. Order-to-user conversion is always higher than order-to-session, and comparing one business's session rate against another's user rate is meaningless.

Both are legitimate. Stating which is in use is what makes the number readable.

Numbers on it

1,000 orders from 40,000 sessions is a 2.5% conversion rate, producing $58,000 of revenue and $31,900 of contribution.

Improving to 3.0% gives 1,200 orders, $69,600 of revenue and $38,280 of contribution, $6,380 more a month from the same traffic.

To generate that $6,380 through advertising instead would take 200 additional customers at a $27 CAC, $5,400 of spend, plus the working capital to fund the payback.

The conversion improvement is permanent and applies to every channel at once; the advertising is recurring and applies only to what it buys. That asymmetry is the entire case for CRO.

What it does not tell you

Conversion rate is heavily determined by traffic mix rather than by the site. Branded search converts several times better than cold display, so a rate that falls after a prospecting campaign launches usually means the campaign is working.

Aggregate rates across devices are also close to useless, since mobile and desktop typically differ by a factor of two and the blend simply reports the device split.

What follows from it

Segment before optimising: by device, by channel, by new against returning. A single site-wide rate hides which of those is the problem and will send effort to the wrong place.

Then judge changes on revenue per visitor rather than conversion rate alone, since a change that lifts conversion while lowering order value can leave the business worse off.

Why conversion work compounds and advertising does not

A conversion improvement applies to all traffic: paid, organic, email, direct, for as long as it stays in place. An advertising improvement applies only to the channel it was made in.

It also lowers the effective cost per acquisition on every paid channel simultaneously, which raises the bid that can profitably be paid and therefore the volume that can be bought. A 20% conversion improvement raises the affordable click price by 20%.

That second effect is the one usually missed. Conversion work does not merely add orders; it expands the set of auctions the business can profitably compete in, which is a growth lever rather than an efficiency one.

Statistical power is the constraint most small sites hit before anything else. At 1,000 monthly orders, detecting a 10% relative improvement reliably takes several weeks, and acting on a result before that point is acting on noise dressed as a finding.

Calculating the required sample before starting, and committing to it, is what separates a testing programme that compounds from one that produces a series of reversals nobody can explain.

Seasonality also has to be accounted for when comparing periods, since a rate that falls in January against December is usually reporting the shift from gift buying back to ordinary demand rather than any deterioration in the site.