SEO Forecasting: How to Build an Honest Traffic and Revenue Forecast, With Ranges Instead of Promises

Why a forecast to the decimal is a sales document

Every SEO forecast is a stack of guesses multiplied together. Search volume is an estimate. The position you will reach is a hope. The click rate at that position depends on what Google puts above you that month. Multiply four uncertain numbers and the error does not average out; it compounds. A single confident figure at the bottom hides all of that, which is usually the point.

That does not make forecasting useless. It makes it a different tool. A good forecast answers one question before money is spent: if things go roughly as expected, can search plausibly produce enough customers to pay for the work? Our SEO strategy guide runs the arithmetic backward, from a revenue goal to the leads it needs. This page runs it forward, from the demand that exists to what a business could reasonably capture.

The five inputs, where each comes from, and how wrong it can be

Every SEO ROI forecast uses the same chain: searches, then clicks, then leads, then customers, then money. The work is in choosing each number honestly and writing down where it came from.

Notice that the most reliable numbers sit at the bottom of the chain, inside the business, and the least reliable at the top. That is why a forecast is only as good as the conversion tracking behind it. Without real lead counts, the bottom half is guesswork too.

Search volume: three sources that never agree

Start with demand, and expect it to be fuzzy. Google Keyword Planner reports average monthly searches for a keyword and its close variants, averaged over 12 months by default, for the location and network settings you choose. Google calls its figures estimates and warns that searches change with seasonality and current events. Very low volume terms are not forecastable at all, which for a local business can mean half the list.

• Keyword Planner is the closest thing to Google's own count. Because each figure already includes close variants, listing "foundation repair" and "foundation repairs" as separate rows and adding them together counts the same people twice.
• Search Console impressions show real appearances for searches where your site already shows up. Google counts an impression whenever your result appears on the page of results the searcher is viewing, so a page sitting on page two understates demand, because few people click through to page two.
• Google Trends shows shape, not size. It divides each data point by total searches in that place and period and scales the result from 0 to 100, and small terms can read as zero. Use it for seasonality, never for volume.
• Third-party tools model volume from their own data and often disagree with Google and each other. Our keyword research tools guide explains why.

Our practice: take the lower of the credible figures as the base and note the higher one beside it. A forecast that turns out conservative builds trust; one that turns out generous ends the engagement. If you would rather have the research done, keyword research is exactly this job, and keyword mapping assigns each group of searches to one page.

Click curves: borrow the shape, measure your own slope

Every forecast needs a guess at how many searchers will click at a given position. The published click curves agree on the shape: clicks fall steeply after the first few results. We keep those figures, with their dates and samples, on our SEO statistics page rather than repeating them here, because the best-known curves predate AI Overviews and describe results pages that many searches no longer show.

Your own data is a better input. In Search Console, filter the Performance report to your non-branded queries and group them into position bands: 1 to 3, 4 to 6 and 7 to 10. Divide clicks by impressions for each band. That is your curve, on your searches, with your map pack, ads and AI Overviews already priced in. Our click-through rate guide shows where each number lives.

1. 1

   Separate brand from everything else

   Brand searches click at rates no service search will match. Leave them out of the curve, or the forecast will assume strangers behave like loyal customers.

2. 2

   Split questions from hiring searches

   A question Google answers on the page draws fewer clicks than a search for someone to hire. If your target list mixes both, give them separate rates.

3. 3

   Discount for AI Overviews where they appear

   Search the target terms yourself. Where an AI Overview sits above the results, lower the expected click rate; AI Overviews taking clicks shows how to measure the effect on your own queries.

4. 4

   Use a range, not a point

   Take the low and high click rates you see within each band over the last year. Those become the low and high cases.

A worked forecast for a hypothetical foundation repair company

Every number in this example is invented for illustration; replace each one with your own. Say a foundation repair company in Houston targets five groups of searches that together show about 2,400 searches a month in Keyword Planner. From its books, it closes 30% of qualified leads and the average job is $6,000, with a 40% gross margin. Those three figures stay fixed across the cases, because the company knows them. What changes is the part nobody knows: where the pages will rank, how often searchers click, and how well the pages convert.

Weight each case by its chance and the expected revenue at full strength is about $12,700 a month. The weights are judgment, not data, and they belong on the page where the owner can change them. An owner who thinks the competition is tougher can move ten points from High to Stalled and watch the answer fall.

Full strength takes time. Google's SEO Starter Guide says some changes show within hours and others take several months, and our page on how long SEO takes sets expectations by situation. Here we assume nothing in months one to three, a quarter of full strength in months four to six, half in seven to nine and three quarters in ten to twelve. That ramp means the first year delivers 37.5% of a full year: about $57,000 of expected revenue, ranging from roughly $15,600 in the low case to $186,600 in the high one.

From revenue to a yes or no: break-even and the low case

Revenue is not the number to compare with the invoice. Gross profit is. At a 40% margin, each $6,000 job is worth $2,400. Say the SEO program costs $1,500 a month, $18,000 for the year. Break-even in year one is 7.5 jobs.

The honest reading is that this is a second-year investment with a risky first year. That is a common shape for SEO, and it is worth saying out loud before signing. If the business needs calls this quarter, paid search can fill the ramp; our guide to SEO and PPC covers the split, and what Google Ads costs gives the click prices to compare.

If only the high case pays for the work, the forecast is telling you not to buy it.

Seven ways SEO forecasts get inflated

• Every keyword at position one. Nobody wins every search. A forecast should assume positions you can defend after looking at who ranks now.
• Close variants counted twice. Keyword Planner already folds close variants into each figure, so a list padded with plurals and spellings overstates demand.
• Twelve-month averages applied to every month. A pool builder's July is not his January. Use Trends or last year's Search Console data to shape the months.
• Brand traffic in the base. A lead rate measured on visitors who searched your name will flatter every page built for strangers.
• Old click curves. Rates measured before AI Overviews overstate clicks on question searches, as the studies on our statistics page show.
• All the hopeful numbers at once. Multiplying the best volume, best position, best click rate and best lead rate produces a case that almost never happens. Vary the unknowns together in honest cases instead.
• No ramp. A forecast that starts paying in month one is describing ads, not SEO.

If someone hands you a forecast, check for these seven before reading the bottom line. Our guide to hiring an SEO covers the other questions worth asking a vendor whose numbers look too clean.

Checking the forecast against what actually happened

A forecast is worth more after launch than before it, because it gives every quarter a yardstick. Each quarter, put the actuals beside the cases: impressions for the target pages, to test the volume; average position, to test the ranking guess; click-through rate by band; and leads and jobs from organic search, from GA4 and the job log. The SEO KPIs guide sets up the one-page report that holds those numbers, and our Google Search Console guide shows where each one lives.

Then revise. If impressions came in below the volume estimate, cut the base. If pages climbed faster than expected, move weight from Low to Likely. A forecast updated with evidence every quarter becomes the most useful document in the account, and one that is never revisited becomes a souvenir of the sales meeting.

If you want the forecast built from your own data, DataCram's free site audit is the starting point, and our SEO programs open with the keyword research and tracking a forecast depends on. Programs start at $575 a month and depend on complexity, setup and budget, as the SEO pricing page explains; SEO financing can spread the cost. The rest of the subject, from strategy to measurement, is in the SEO guide.