# SEO Forecasting: A Model That Survives Falling Impressions **URL:** https://mettevo.com/blog/article/seo-forecasting **Published:** 2026-10-01 **Updated:** 2026-10-01 **Author:** Oleg Silin **Category:** Blog | Mettevo > Understand SEO forecasting, including traffic projections, ranking trends, assumptions, and key metrics, to make smarter SEO decisions. ![SEO Forecasting](https://stage.mettevo.com/wp-content/uploads/2026/10/Team_analyzing_SEO_data_dashboard_20261001095349.jpg) --- SEO forecasting is the process of predicting future organic clicks by multiplying projected impressions by an expected click-through rate (CTR). Most templates pull that CTR from one fixed, published curve. The curve moved. [Mettevo's](https://mettevo.com/) own Google Search Console data shows impressions falling 63% while CTR rose 85% on the same pages, February to August 2026. ## TL;DR - Standard forecasts multiply search volume or impressions by one fixed CTR-by-position curve borrowed from a published benchmark. That curve is not stable through 2026. - On Mettevo's own site, impressions fell 63% (390,704 to 146,119, February to August 2026). CTR rose from 7.78% to 14.37%, and average position improved from 26.75 to 23.41. - In the same nine-property portfolio, a regional accounting firm moved the other way. Impressions rose 29%, clicks rose 194% (18 to 53), CTR rose from 0.75% to 1.71%, and position stayed flat near 28. - The model below replaces the fixed curve with two measured trend factors: one for impressions, one for CTR. Both come from a property's own two comparable Search Console windows, worked through here with real numbers a reader can rebuild in a spreadsheet. - A dedicated error-bar section names the assumptions that break the forecast, including small-sample volatility and clicks that Search Console still cannot report separately. ## What Is SEO Forecasting? SEO forecasting estimates future organic clicks, sessions, or revenue for a keyword, [page](https://mettevo.com/seo/on-page-seo), or property before the work that would earn them happens. Every method reduces to two ingredients: how many times a result gets shown, and what share of those views becomes a click. [Google's Search Console Performance report documentation](https://support.google.com/webmasters/answer/7576553) defines the second number precisely: CTR is "the click count divided by the impression count." Average position, defined on the same page, is the average position of a site's topmost matching result for a given query, page, or date. The disagreement between forecasting methods is not over that formula. It is over where the CTR half of it comes from. Most published templates pull CTR from a static curve indexed to ranking position. Position 1 earns roughly this rate, position 5 earns roughly that one, and the number does not move again until someone updates the spreadsheet. That single assumption is what breaks first, and it is what the rest of this model replaces. ## Why the Standard SEO Forecasting Model Broke A static CTR curve assumes two relationships hold steady: position to clicks, and position to impressions. Portfolio data from February through August 2026 breaks both at once. Mettevo's own property held its ranking almost flat, improving from an average position of 26.75 to 23.41, a shift of just over three spots. Impressions over the same window fell 63%, from 390,704 to 146,119. CTR did not fall alongside them. It rose 85%, from 7.78% to 14.37%. A model that infers impressions from position, or CTR from position, cannot explain that combination from position data alone. [GrowthSRC's analysis](https://growthsrc.com/google-organic-ctr-study/) of roughly 74,000 keywords ranking in the top 10 found position-1 CTR falling from 28% in 2024 to 19% in 2025, a 32% drop. Position 2 fell from 20.83% to 12.60% over the same year, a 39% drop. Positions 6 through 10, tracked as one combined group, moved the opposite way: up 30.63% over the same year. GrowthSRC attributes the shift mainly to AI Overviews, Google's artificial intelligence (AI) summaries shown above search results, which began rolling out in November 2024 and scaled through 2025. Ahrefs ran a comparable check twice on the same 300,000-keyword panel using aggregated Search Console data. In April 2025, it measured AI Overviews reducing clicks to the top-ranking page by 34.5%, against a forecasted CTR without an AI Overview present. By December 2025, the same method put that figure at 58%, a gap of 23.5 percentage points in eight months. Whatever curve a template used in April was out of date by December, from the same source and method. ## Inputs a Forecast Actually Needs A forecast that survives a moving CTR curve needs inputs measured from the property itself, not looked up from someone else's curve. Pull these from Google Search Console (GSC) for two windows of equal length: a baseline period and a later comparison period. - Impressions and clicks for the baseline window - Impressions and clicks for the comparison window - Average position for both windows, kept as a check rather than a driver - The exact start and end date of each window; GSC's own reporting can lag a few days CTR is not an input on its own. It is calculated per window as clicks divided by impressions, matching Search Console's own definition exactly. Average position stays on the list for a different reason. If a forecast misses and position moved a lot, position is the first suspect. If position barely moved, as in both cases here, the trend factors below are doing the explaining instead. ## Building the Model Step by Step The model uses two measured trend factors instead of one fixed curve. 1. Pick a baseline window and a comparison window of equal length from the same property. 2. Record impressions and clicks for each window, then compute CTR for each: CTR = Clicks / Impressions. 3. Compute the Impression Trend Factor (ITF): ITF = Impressions(comparison) / Impressions(baseline). 4. Compute the CTR Trend Factor (CTF): CTF = CTR(comparison) / CTR(baseline). 5. Project one more window of the same length: - Forecast Impressions = Impressions(comparison) x ITF - Forecast CTR = CTR(comparison) x CTF - Forecast Clicks = Forecast Impressions x Forecast CTR Here is that sequence run on real numbers, from the regional accounting firm inside Mettevo's nine-property portfolio, pulled through Mettevo's Ahrefs-connected Google Search Console access on September 10, 2026. Baseline is April 2026, comparison is August 2026, a four-month window, forecasting forward to December 2026. **Metric** **April 2026** **August 2026** **Trend factor** **Dec 2026, full trend** **Dec 2026, damped trend** Impressions 2,397 3,099 ITF = 1.293 4,007 3,553 Clicks 18 53 N/A N/A N/A CTR 0.75% 1.71% CTF = 2.277 3.89% 2.80% Forecast clicks N/A N/A N/A 156 100 The full-trend column carries both factors forward at the same rate observed between April and August: 3,099 x 1.293 for impressions, 1.71% x 2.277 for CTR, together for 156 clicks. The damped column hedges against a four-month window being a blip rather than a stable rate. It halves each factor's percentage gain before projecting: ITF's 29.3% becomes 14.65%, and CTF's 127.7% becomes 63.85%, for 3,553 impressions, 2.80% CTR, and 100 clicks. Both columns are legitimate reads of the same two Search Console pulls. That is why this article reports 100 to 156 clicks as a range, the high end 56% above the low end, instead of a single number. ## What CTR Curve Should a Forecast Use? None, as a starting assumption. The published curves closest to the top of the results page show CTR falling: positions 1 and 2, per GrowthSRC above. The one combined figure available for positions 6 through 10 shows CTR rising. Mettevo's property, averaging position 23 to 27, and the accounting firm, near position 28, both showed CTR rising too, by a wider margin than the published positions-6-to-10 figure. A search for a published, verified CTR rate at any position past 10 turned up nothing, so there is no external curve to check either property against. Pew Research Center tracked 68,879 Google searches from a 900-adult panel in March 2025, adding a mechanism without needing a position-level curve at all. Pages with an AI summary present got a click on a traditional result in 8% of visits, versus 15% with no summary present. Only 1% of visits to a summary page clicked a source cited inside the summary itself. Fewer clicks land near the top once a summary answers the query first. The clicks that remain have to land somewhere, and positions below the summary, missing from most published curves, are a plausible place for that share to move. [Google's own developer documentation](https://developers.google.com/search/docs/appearance/ai-features) states there are "no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary." It adds that "the best practices for SEO remain relevant" for these features. That statement is about what to build. It says nothing about what a given position pays in clicks once it ranks, which is the number this model exists to measure rather than assume. ## What Is the Best SEO Forecasting Tool? There is no single best tool, because the four common approaches need different data and fail in different ways. **Approach** **Where the CTR comes from** **Rebuilds when the curve moves** **Needs** Static CTR-benchmark template A published industry curve, looked up by target position No, fixed at build time A target keyword and target position Statistical trend regression Historical traffic, projected forward with a stated margin of error Partially; the margin absorbs some drift, not its cause 6+ months of stable historical data Rank-tracker traffic estimate The vendor's own aggregated CTR curve, refreshed on the vendor's schedule Only as fast as the vendor republishes its curve A rank-tracking subscription, not property-specific data Two-trend model, this article Two of the property's own GSC windows, remeasured every run Yes, by construction At least two comparable GSC windows for the same property A static template is still the fastest way to size a [brand-new keyword](https://mettevo.com/seo/keywords-research) with zero ranking history. The two-trend model only works once a page or property has its own history to measure, which is the gap the next section names directly. ## How Accurate Is SEO Forecasting? Not accurate enough to report as one number, and the assumptions behind this model explain why. Small baselines swing hard. The accounting firm's April baseline was 18 clicks. If it had landed at 20 instead, a difference well within normal week-to-week noise for an 18-click month, the CTF would compute to 2.049 instead of 2.277. That is a 10% change in the multiplier, caused by two clicks. A forecast built on a monthly click count under a few dozen should be read as a range, not a point estimate. The model cannot predict whether a query will gain or lose an AI Overview mid-window, and Google's own documentation confirms there is no markup or optimization that controls it either way. A window that opens or closes an AI Overview mid-period moves CTR for a reason the trend factor never sees coming. It also cannot separate AI Overview impressions from ordinary organic ones. [Search Console's Generative AI performance report](https://support.google.com/webmasters/answer/16984139), rolled out to every property worldwide on August 31, 2026, shows impressions only, by page, country, device, and date, with no click or query column. A forecast built on total organic clicks cannot say what share came from a blue link versus an AI citation. Search Console and Google Analytics will not confirm each other either. [Google's own comparison documentation](https://developers.google.com/search/docs/monitor-debug/google-analytics-search-console) notes that "clicks and sessions are calculated differently... you'll likely see different numbers," citing time zone, bot-filtering, and attribution gaps between the tools. Validate a forecast against Search Console clicks, not Analytics sessions, or expect an unrelated gap. Finally, nine properties is Mettevo's own client portfolio, not a random sample of any industry. The set spans two flower e-commerce shops, a neurology clinic, a long-distance mover, a sports-betting platform, an AI interior-design product, our own agency site, and the accounting firm above. One property's impressions collapsed while another's impressions and clicks both climbed. That proves a single CTR curve cannot describe every site in one small portfolio. It does not prove which direction any other specific site will move next. ## FAQ ### How Do You Forecast SEO Traffic? Measure impressions and clicks for two comparable Search Console windows, then compute CTR for each as clicks divided by impressions. Derive an Impression Trend Factor (ITF) and a CTR Trend Factor (CTF) from the two windows. Multiply both against the more recent window to project the next window of equal length. Report the result as a range, not a single number. ### How Far Ahead Should an SEO Forecast Run? Match the forecast window to the length of the data behind it. A four-month baseline-to-comparison window, like the accounting firm's April-to-August data in this article, supports forecasting roughly four months forward before the trend factors extrapolate further than the data justifies. ### Does SEO Forecasting Work for a Page With No Ranking History? Not with the two-trend model described in this article, since it needs two real Search Console windows to compute a trend factor. A brand-new page with zero history has to start from a static CTR-benchmark estimate instead. It can switch to the two-trend model once enough weeks of its own Search Console data exist to define a baseline window. ### Can an SEO Forecast Account for Clicks AI Overviews Absorb? Not directly. Search Console's Generative AI performance report shows impressions for AI Overviews and AI Mode by page, country, device, and date, but it has no click or query column. There is no way to subtract AI-absorbed clicks from a total organic forecast using Search Console data alone, because Search Console itself does not report that split.