← Answers

Revenue is flat but conversion is falling. What does that mean?

Short answer

Flat revenue with a falling conversion rate means traffic has grown enough to offset the fall, so you are paying for more visitors to stay in the same place. It is more dangerous than a visible revenue drop because every headline number looks acceptable while unit economics deteriorate: cost per acquisition rises, margin per visitor falls, and the trend compounds silently until traffic growth stops. The usual causes are a shift in traffic mix toward lower-intent channels, a checkout or site regression, or a price change that has not been absorbed.

What is actually happening

Revenue is traffic multiplied by conversion rate multiplied by order value. If conversion falls and revenue does not, something else in that product rose to cover it — almost always traffic.

Traffic is rarely free. Whether you paid for it in advertising or in the time spent earning it, more visitors for the same revenue means each unit of revenue now costs more to produce.

Why it is worse than a visible drop

A revenue drop gets attention. This does not, because nothing on the summary looks wrong.

It also compounds. The gap between traffic growth and revenue growth widens each month, and the day traffic growth stalls — a campaign ends, a ranking slips, a season turns — the underlying conversion decline arrives all at once as a sudden revenue fall that appears to have no cause.

The three usual causes

  • Traffic mix moved. New volume arrived from a channel that browses more and buys less. Site-wide conversion falls while each individual channel is unchanged — a composition effect, not a decline. Check conversion by channel before concluding anything is broken.
  • Something regressed. A checkout step, a payment method, page speed, or a mobile layout. Look for a step change on a specific date rather than a gradual slope.
  • Price changed, or a competitor’s did. Conversion falls but order value rises, and revenue holds. This is a deliberate trade rather than a fault — but it should be a decision, not a discovery.

How to tell which

  1. Split conversion rate by channel for both periods. If every channel is flat and only the total fell, it is mix — nothing is broken.
  2. Plot conversion daily rather than monthly. A cliff means a regression on that date; a slope means mix or competition.
  3. Check order value. If it rose as conversion fell, you are looking at a price change.
  4. Split by device. A fall confined to mobile is almost always a layout or speed regression.