SEO ROI calculator

The arithmetic is deliberately simple enough to audit. Incremental sessions at maturity are your current organic sessions multiplied by the uplift. Those sessions become leads at your visit-to-lead rate, leads become customers at your close rate, and customers become revenue at your average deal value. Revenue becomes gross profit once your margin is applied, and gross profit is the only figure worth comparing to the cost of the programme.

The ramp is what separates a credible forecast from a sales deck. Rankings move on a lag: technical fixes take weeks to be recrawled, new content takes months to earn its position, and links take longer still. The model applies a linear ramp from month one to the maturity month you set, which is conservative — real curves are S-shaped, slower at the start and faster in the middle — and then holds the gain flat rather than assuming it keeps compounding.

Costs accumulate every month whether or not results have arrived, which is exactly why the payback month matters more than the annual ROI. A programme costing 2,500 a month that reaches maturity in month nine can look magnificent over three years and still be unaffordable in the year you actually have to fund it. If the payback lands beyond month eighteen, the honest move is to reduce scope, not to extend the horizon until the number looks good.

The uplift figure is where optimism concentrates. Traffic growth is not distributed evenly across a site: a handful of templates and a handful of intents usually carry the whole gain, and if those pages are already at position three, there is far less headroom than a site-wide percentage suggests. Before trusting a number here, check in Search Console how much impression volume sits at positions four to fifteen — that band is where realistic growth actually lives.

The second trap is the conversion rate. Organic traffic is not homogeneous: a page answering an informational question converts an order of magnitude worse than a comparison page, so growth concentrated in blog content produces far less revenue per session than the site average implies. If you can, split the model: run it once for commercial pages with their real rate, and once for editorial traffic with its own.

The third is margin. Agencies are rarely asked for it and clients rarely volunteer it, yet it changes the conclusion completely: at a 15 % margin, a programme needs nearly four times the revenue to justify the same fee as at 55 %. If you do not know your margin on incremental business, use the contribution margin — price minus the costs that vary with each additional sale — rather than the accounting margin.

Present three scenarios rather than one. A conservative case at half the uplift and a slower ramp, a central case, and an upside case. A single number invites the response "where does that come from"; a range with explicit assumptions invites a conversation about which assumption to challenge, which is a much better conversation to be having.

Pair the model with evidence that the uplift is achievable. That is exactly what the verified Search Console data on this directory is for: instead of promising a number, an agency can point at its own measured performance and at comparable programmes. If you are on the buying side, ask for the equivalent — anonymised exports, not screenshots of a rankings dashboard.

Finally, decide up front how you will judge the forecast. Agree the metric (gross profit from organic, not sessions), the reporting cadence, and what happens if month six looks like the conservative case. Programmes rarely fail because the forecast was wrong; they fail because nobody agreed in advance what a wrong forecast would trigger. Our guides on KPIs and reporting and on the questions to ask an agency cover how to write that into a contract.