Free Growth Assessment
SEO & AEO

SEO Forecasting: How to Project Traffic and Leads Before You Invest

A forecast is a range with its assumptions attached. Here’s the chain from search volume to signed agreement, a worked example with invented numbers, and where the math lies.

By Rhett Collver Last updated: November 19, 2026 10 min read

SEO forecasting turns keyword demand, click-through rate and your own conversion rates into a range of traffic and leads. Treat it as a budget input with three scenarios, not a promise.

Two men in dark jackets at a concrete workbench, one holding brass drafting dividers over a blank sheet of drafting paper
Write the assumptions down before you draw the line.

You wouldn’t sign off on a new hire without a payback number. Yet the SEO line gets approved on a slide that says “organic growth” and nothing else.

A forecast fixes that. Not because it will be right. It won’t be.

It makes you write down what has to be true for the spend to pay, which means search volume, click rate, conversion, deal size and time, five assumptions that are each a risk you’re choosing to carry. Write them down.

Then you can argue about the assumptions instead of arguing about a feeling. That’s a better fight.

At C4 we say sales is math. SEO is a sales channel, so it gets the same math. Below you’ll find the chain, a worked example with invented numbers, and the places where the result lies to you.

Key Takeaways

The Short Version

An SEO forecast is a chain. Monthly searches for the keywords you’d target, times the click rate for the position you expect, times your visit-to-lead rate, times your close rate, times deal value. Run it three times, low, base and high, take a haircut for AI answers, and compare the result with the spend over 24 months. Google gives no fixed timeline for results, so assume a slow start and re-forecast every quarter against Search Console.

This sits inside our wider view of SEO, AEO and GEO for tech companies. The whole-marketing version of the payback question is in our guide to MSP marketing ROI. The search-only version, with the payback month worked through, is in our guide on how to calculate SEO ROI. We sell SEO and AEO, so you’ll also get the cases where the math says don’t buy.

What Is SEO Forecasting? A Range, Not a Promise

SEO forecasting is estimating the organic clicks, leads and revenue a set of pages could earn over a stated period. It starts from keyword demand and click-through rate, adds your own conversion rates, and ends as a low, base and high range. It’s an estimate you can check, not a ranking promise.

Google doesn’t promise much either. Google’s SEO Starter Guide says some changes take effect in a few hours and others could take several months, and it doesn’t even promise that a given site will be indexed. The guide was last updated December 10, 2025.

No date, then. A range, with its assumptions attached.

The Chain: What Goes Into an SEO Traffic and Lead Forecast

Every forecast is a chain of the same links. Each has a source, and each one lies in its own way.

The Six Links, and Where Each One Lies

LinkWhere the number comes fromWhere it lies
Monthly searchesGoogle Keyword Planner’s average monthly searches, or a third-party toolA 12-month average hides seasonality and folds in close variants
Click rate by positionYour own Search Console data first, a named study secondMoves with the search page layout and with AI answers
Expected positionYour current average position, plus a judgment about how far pages can moveThe shakiest link. Nobody can promise a position
Visit-to-lead rateYour analytics: organic sessions to forms, calls and demo requestsBranded visitors convert differently from strangers
Lead-to-deal rate and deal valueYour CRM: lead to opportunity, opportunity to close, first-year contract valueA handful of closed deals gives a noisy rate
TimeThe ramp: when clicks start and how fast they climbGoogle publishes no schedule

No row is optional. Skip the CRM rows and what you’ve built is a traffic forecast. Sessions don’t pay invoices.

Start With Your Own Search Console Data

If the site has history, pull it before you open any study. In the Performance report, Google’s Search Console help on performance data defines an impression as a link to your site that a user saw or could have seen, and position as an approximation of where that link appeared relative to the other results, averaged across impressions.

Average position isn’t a rank. Google’s own example is a property that shows up at positions 2, 4 and 6 for one query. It counts as position 2, the best of the three.

Pull a year of non-branded queries by page, because branded clicks come from people who already know you, and mixing them in flatters the click rate you’d get from strangers. Our guide to branded search shows how to split the two.

For each target page, write down impressions, clicks and average position. That gives you your own click rate by position, and it beats any industry curve, because it’s your titles, your search page layouts and your buyers.

No history, because the site is new or the pages don’t exist yet? Then every number in the chain is borrowed. Say so in the forecast. Out loud.

Click-Through Rate by Position: Borrow a Curve and Label It

When you don’t have your own curve, borrow one and name it. The SISTRIX study of Google click rates covers more than 80 million keywords. The page carries a July 14, 2020 publish date and a July 7, 2025 modified date, and it reports average click rates of 28.5% at position 1, 15.7% at position 2, 11% at position 3 and 2.5% at position 10. SISTRIX sells SEO software.

PositionAverage click rate (SISTRIX)
128.5%
215.7%
311%
102.5%

Two warnings. The page doesn’t say whether the data was refreshed after 2020, so I treat it as a baseline from before AI answers. And the same study says position 1 earns anywhere from 13.7% to 46.9% depending on search intent and page layout, with a Google Shopping box or ads at the low end and sitelinks at the high end.

One curve for every keyword averages away the part that matters.

A Worked Example: Three Scenarios for an Invented MSP

Every number below is invented, the MSP included. It shows the arithmetic. Nothing else. Put your own numbers in each row.

  • 12 non-branded keywords, 2,400 searches a month combined.
  • Click rate by scenario: 3% in the low case, 8% in the base case, 13% in the high case. I rounded down from the study above because the keywords are niche and commercial.
  • A 25% haircut on clicks for AI answers. The reasoning is two sections down.
  • 2% of visits become leads, 25% of leads become opportunities, 20% of opportunities close.
  • A first-year contract value of $48,000, against $6,000 a month of total SEO spend.
  • The ramp: no clicks for four months, then a straight line up to the scenario level at month 12. That’s my working assumption, not a Google rule.
CaseClick rateClicks / moLeads / moDeals, 24 moValue, 24 mo
Low3%541.10.9$43,000
Base8%1442.92.4$114,000
High13%2344.73.9$185,000

Clicks and leads are monthly run rates at month 12. Deals and value add up the whole 24 months, ramp included, and value counts first-year contract value only.

Spend over the same 24 months is $144,000. The base case returns about 79 cents on the dollar. Only the high case clears the spend. The low case isn’t close.

Year one alone is worse. It returns about $12,000, $31,000 and $51,000 against $72,000 of spend. If your board needs payback inside 12 months, SEO is the wrong channel for this business, and the forecast told you before you spent the money.

Not a verdict on SEO. It’s a map of which input has to move. More keywords, a larger deal, lower spend, or a longer wait you’re willing to fund.

I left out two things on purpose, renewals and margin, because an MSP agreement that renews for three years is worth more than its first year and gross margin is yours to supply. Add them and the base case may flip. Or it may not.

A person in a dark overcoat climbing a wide straight flight of dark stone steps into blue haze
The ramp is slow at the start. Plan for it.

Where an SEO Forecast Lies

  • Search volume is an average. Google Ads Help page on Keyword Planner forecasts says average monthly searches is, by default, averaged over 12 months and includes close variants. A seasonal keyword looks flat in that number.
  • Keyword Planner also has a Forecasts view, and it’s built for ads. The same page says forecasts take your bid, budget, seasonality and historical ad quality into account. Don’t read organic clicks off it.
  • Rank is a guess. Treat expected position as the input to argue about hardest, and widen the range around it.
  • Small numbers. With 2,400 searches in the whole set, the output is a fraction of a deal. Say “between one and four deals”, not 2.4. Precision to the decimal is theatre.
  • Two of your pages chasing one query split the clicks. Forecast by page, not by keyword list.
  • Your visit-to-lead rate was measured on the traffic you have today. Expect new traffic from broad guides to convert lower than traffic from searches that already name a service.

What AI Answers Do to the Forecast

Click rates built on older studies run high. The Ahrefs study of AI Overviews and clicks compared 300,000 keywords, half with an AI Overview and half informational keywords without one, using desktop click rates from aggregated Search Console data. In its February 4, 2026 update, using December 2025 data, an AI Overview correlated with a 58% lower average click rate for the top-ranking page. Its first run, in April 2025, had found 34.5%. Ahrefs sells SEO tools, calls the result a correlation, and the sample is informational keywords, so a commercial keyword may behave differently.

Pew Research Center followed the browsing of 900 U.S. adults in March 2025. People clicked a traditional result link in 8% of visits when an AI summary appeared and in 15% when it didn’t. They clicked a link inside the summary in 1% of visits. One month, one panel.

Here’s how I’d use both. A haircut, not a model. If half your target keywords show an AI Overview and each loses about half its clicks, take roughly a quarter off the forecast. That’s the 25% in the example. It’s my arithmetic, and it’s a placeholder until your own Search Console shows the drop.

Don’t forecast citations in ChatGPT or Perplexity. I found no study that ties a citation to a lead for an MSP. Track mentions as a separate scoreboard, the way AI search visibility tracking does, and keep them out of the lead math. Not yet.

Check the Forecast Every Quarter

A forecast nobody checks? A slide. Here is the sequence.

  1. Save the forecast with its date and every assumption. Don’t edit it afterward.
  2. Each quarter, pull actual clicks, leads and opportunities for the pages in scope.
  3. Compare them with the three scenarios. Which line is the real number tracking?
  4. Find the link in the chain that missed: impressions, click rate, visit-to-lead rate or close rate. One link usually explains the bulk of the gap.
  5. Re-forecast from the new baseline, and write down what changed and why.

Our guide to SEO KPIs that tie to pipeline lists what to track between checks, and a monthly SEO report shows how to present it.

The back half of the chain is a sales problem, because pipeline coverage, win rate and stage conversion decide whether a lead becomes revenue, and if the process leaks, an SEO forecast just forecasts more leaks. Our post on MSP sales process covers that side.

When a Forecast Is the Wrong Tool

No history. No pages yet. You’re guessing with extra steps, so publish a handful of pages for real buyer questions and read a quarter of Search Console before you build the model.

If the keyword set is too small to produce even one whole deal, forecast the channel instead, with leads from search, from referrals and from LinkedIn compared at cost per opportunity.

And if you can’t close the leads you already have, no forecast helps. Fix the follow-up first. If you want a second read on the chain behind your numbers, book a growth strategy call.

Before You Put an SEO Number in the Budget

How accurate is an SEO forecast?

Wide enough to be useful and no wider. A range of one to four deals is honest, and a single number to the decimal is theatre.

I found no study that measures how often SEO forecasts land inside their range, so I won’t quote an accuracy figure.

Weeks, months or a year: when does SEO show up in the numbers?

Weeks to months, by Google’s own account. The SEO Starter Guide says some changes take effect in a few hours, others could take several months, and that you’ll likely want to wait a few weeks before judging a change.

For the forecast I assume a slow first stretch and a straight-line climb. Your first two quarters of Search Console will correct it.

What click rate should I assume for position one?

28.5% is the SISTRIX average, and the same study’s range runs from 13.7% to 46.9% by search type.

Ahrefs reports about 58% less for the top result where an AI Overview shows, on informational keywords. Use your own Search Console number where you have it, and haircut the borrowed ones.

Can I forecast traffic from ChatGPT or Perplexity?

Not with anything I’d defend. No published study I found links AI citations to leads for MSPs.

Track mentions and citations as their own scoreboard, and leave them out of the revenue chain.

What do I need from my CRM?

Four rates. Visit to lead, lead to opportunity, opportunity to close, and first-year contract value. Pull the last 12 months.

Closed only a handful of deals? Label the rates rough and widen the range.

Should I forecast by keyword or by page?

By page. One page can earn clicks from dozens of related queries, and two pages can fight over one.

Use Search Console’s queries for each page as the starting list, then add the keywords you want it to win.

SEO Forecasting Search Console Pipeline Math SEO & AEO
Share
READY TO ACT?

Put a range on your SEO budget before you spend it

Bring us your Search Console export and your CRM close rates, and we will walk through the chain with you, show which input carries the risk, and say whether the math supports the spend.

Book a Growth Strategy Call