SEO Forecasting Tool

For each keyword, monthly clicks = search volume × CTR at the rank × SERP-feature suppression. The forecast computes that twice: once at your current rank (clicks now) and once at the rank the model credits you (clicks at target). The difference, summed across the set, is your added traffic. We show the do-nothing baseline (inertial traffic if rankings never change) right beside the forecast, so the incremental gain is transparent rather than assumed.

Added clicks then flow to revenue: clicks × conversion rate × value per conversion, plus a Google Ads-equivalent value (clicks × CPC) as the cost-avoided anchor. Branded queries, when included, are discounted because they'd convert with or without SEO.

The default is a conservative blended non-branded curve (position 1 ≈ 25%, falling to ~1.8% by position 10) so the tool never over-promises. You can switch to the Backlinko 2024, SISTRIX, or First Page Sage curves, or paste your own from your GSC CTR-by-position report. Positions between listed points are linearly interpolated, and a smooth decay handles page 2 and beyond.

We deliberately exclude branded CTR, which runs 35-60% at position 1 and reflects existing demand, not SEO work. Both Advanced Web Ranking and SEOmonitor exclude it for the same reason.

Each feature on a query removes a fraction of organic CTR, combined multiplicatively with a floor so clicks never collapse to zero. AI Overviews are position-aware on an Ahrefs basis: −58% for the top result, easing to about −19% at position 10. Other features decay down the page, for example a featured snippet at −35% at the top, a shopping pack at −45%, top ads at −22%.

If your upload includes a SERP-features column, each keyword carries its own features automatically. Otherwise the global toggles apply across the set. This single AI Overview toggle is often the most consequential input in the entire forecast.

No published tool has a validated keyword-difficulty-to-rank formula, so we don't pretend to. Instead, the model scores reachability from your effective authority (Domain Rating plus committed link work) against the SERP's required authority (KD, blended with competitor DR when you supply it), weighted so content counts for more on low-competition terms where links dominate less. That reachability score credits the rank you'll realistically hold and tempers how fast you get there.

The ramp raises traffic month over month on an ease-out curve, fast early then settling toward target. Each keyword gets its own ramp stretched by difficulty, and the chart horizon covers the full climb plus a plateau. Technical health is treated as a pass/fail gate: leave it unaddressed and reachability is capped.

The deterministic result is the median. Around it, a Monte Carlo simulation runs 2,000 trials, each multiplying the headline by independent factors for search-demand variance, CTR-curve uncertainty (the largest unknown), and how fully the modeled rank is achieved. The sorted trials yield the 10th, 50th, and 90th percentiles, reported as Worst (P10) / Expected (P50) / Best (P90).

Because every factor is centered on 1.0, the expected case tracks the deterministic headline. The spread is honest uncertainty, not a second layer of pessimism stacked on a model that already discounts for reachability.

A forecast is a confident range, not a precise prediction. Accuracy depends on data quality and how volatile your SERPs are. The honest move is to forecast a best / expected / worst band and aim to surpass the conservative case, rather than chase a single number that algorithm updates and AI Overviews can break.

At minimum, just a keyword and its search volume. That's it. Everything else is optional and refines the estimate: current position, keyword difficulty, SERP features, and CPC. So a raw Ahrefs Keyword Explorer or Semrush Keyword Magic export, which has no current rankings because you don't rank yet, works fine: those keywords are simply modeled as ones you're not yet ranking for. You can also skip the upload entirely and enter a combined volume and rank by hand.

Most credible SEO forecasts run 6 to 24 months. Shorter than that and the ramp hasn't matured; much longer and the SERP will have changed enough to outpace the assumptions. This tool sizes its horizon to your push rate, covering the full climb plus a plateau.

Significantly. When an AI Overview is present, the top organic result can lose around 58% of its clicks (per Ahrefs, 2026). The tool applies a position-aware AI Overview suppression so your forecast reflects an AI-shaped SERP, not the ten-blue-links world that no longer exists.

No. Branded searches would convert with or without SEO, and their high CTR inflates the result. Forecast only non-branded demand, the traffic SEO actually earns. The tool is non-branded by default and discounts branded clicks if you include them.

Reaching page 2 or 3 is often quick and mostly on-page. Cracking the top 10 is the hard part and depends on authority versus difficulty, typically over several months to a year. Traffic rises gradually as rankings climb, which is why the forecast models a ramp, not an overnight jump.

Yes, manually. The CSV upload is built for Ahrefs and Semrush exports, which carry the search-volume column the model needs (a raw GSC export doesn't). But GSC is still your best source for two inputs: enter your average position from it, and paste your real CTR-by-position into the Custom curve option. Nobody's curve fits your SERPs like your own. There's no live Search Console connection; the tool stays entirely in your browser.

For a quick gut check a spreadsheet is fine. This tool adds what most templates miss: SERP-feature and AI-Overview-adjusted CTR, reachability gating against keyword difficulty, seasonality, a month-by-month ramp, and Monte Carlo ranges, all without exposing your data, since it runs entirely in your browser.