How to forecast SEO revenue without lying to your board

Every agency proposal contains a forecast. Almost none of them survive contact with a finance team, because they are built on the same three shortcuts: total keyword volume, an average click-through curve, and an assumed conversion rate borrowed from a blog post.

A defensible forecast starts from the opposite direction — from the revenue the business needs, working backwards through the funnel it already has.

Start with the demand you can actually address. Strip out informational terms that will never convert, remove branded volume you already capture, and cap the addressable set at keywords where a page like yours plausibly ranks in the top five within the engagement period.

Then apply a click curve derived from your own Search Console data rather than an industry average. Your brand, your SERP features, your category. Two sites at position three can have wildly different CTRs depending on how many ads and AI summaries sit above them.

Finally, model three scenarios and show the assumptions behind each one. A conservative case that assumes only technical wins land. A base case that assumes the content programme performs at your historical average. An upside case that assumes link acquisition hits target. Publish the assumptions next to the numbers, and update them monthly as real data arrives.

The result is less impressive-looking than a hockey stick. It is also the only kind of forecast that earns a second budget cycle.