How to Measure SEO ROI When the Sale Closes Months After the Search

The arithmetic is simple. SEO ROI is the revenue you can credibly attribute to organic search, minus the total cost of the SEO program, divided by that cost. The difficulty is in the two inputs: deciding what counts as organic revenue, and counting every cost honestly.

Define attribution before you calculate. Many teams count only last-click organic revenue, which understates search that starts a long journey. Others count any deal with an organic touch, which overstates it. Pick a rule, write it down, and report the alternative as a range so readers see the sensitivity.

On the cost side include strategy and consulting, content creation and expert review, developer time for technical fixes, tools, link-earning or digital PR, and the internal hours people spent. Leaving out internal time flatters the result and surprises finance later.

On the revenue side use closed-won revenue or gross margin from deals where organic search is recorded in the CRM, by your attribution rule. For lead-based businesses without a sale in the CRM, use qualified opportunities multiplied by your historical close rate and average deal value, and label it an estimate.

Group leads into cohorts by the month they arrived and follow each cohort to its outcome. Compare cohorts at the same age, for example 90 days after first touch, so a recent cohort is not judged against an older one that has had time to close. Show the forecast for immature cohorts separately from realised revenue.

Google's starter guide notes that some changes take effect in hours and others take several months. That lag is built into SEO economics: the investment comes first and returns arrive later, so a rolling view is the fair one.

It cannot prove causation. A reader may find you through search, return through a referral and sign after a sales call, and the credit rules you pick only describe a path. For a causal answer you need an experiment, such as holding back content or pages in comparable sections, which is rarely practical for organic search.

Report the number with its assumptions, the range between the strictest and loosest attribution rules, and the qualitative evidence from sales conversations. A defensible range beats a precise figure nobody believes.