The short version: forecasting SEO traffic means combining real search volume, a realistic click-through-rate curve, and an honest look at your domain’s current authority against the sites already ranking, then multiplying it out over a 6 to 12 month timeline rather than guessing. Most forecasts fail not because the maths is hard but because people plug in keyword tool numbers without checking who they’d be competing against. Get that competitive check right and you can predict traffic within 20 to 30 percent of reality, which is close enough to make a real budget decision.
Why most content plans get built backwards
I’ve sat in more content strategy meetings than I can count, and the pattern is always the same. Someone pulls a keyword list from Ahrefs or Semrush, sorts by search volume, and builds a content calendar around the biggest numbers. Nobody checks whether the site has any real chance of ranking for those terms in the next year, or the next three years. Then six months later someone asks why the blog isn’t “working” and the answer is that it was never going to work, because the forecast step got skipped entirely.
A forecast isn’t a nice-to-have add-on you do after you’ve picked your topics. It’s the thing that tells you which topics are worth writing in the first place. Get this right before you commit a writer’s time, a freelancer’s invoice, or your own three hours on a Tuesday afternoon, and you’ll stop wasting effort on content that was doomed from the keyword research stage.
The four numbers you need
Forget the fancy dashboards for a minute. A usable forecast needs four inputs, and you can get all four for free or close to it.
- Monthly search volume for the target keyword and its close variants, pulled from Google Keyword Planner or a paid tool, cross-checked against Google Search Console data from any similar page you already have.
- A realistic click-through-rate curve by ranking position, because position 1 does not get anywhere near 100 percent of clicks anymore.
- Your site’s current authority relative to the pages already ranking on page one, measured by domain rating or domain authority plus the actual backlink count of the top three results.
- A realistic timeline for how long it takes a page like yours, on a domain like yours, to climb into a position where the CTR curve pays off.
Miss any one of these and the forecast is fiction. Most of what I see online skips the third one entirely, which is the one that matters most.
Step one: pull real search volume, not tool volume
Keyword tools estimate volume, they don’t measure it. Google Keyword Planner rounds and buckets numbers, especially if you’re not running active ad spend, and third-party tools model volume from clickstream data that’s often 18 to 24 months old. For a keyword like “how to forecast SEO traffic,” a tool might tell you 90 searches a month, or it might tell you 320, depending on which tool you use and which month you check.
The fix is to treat the tool number as a rough band, not a fact, and to cross-check it against Google Search Console impressions for any page on your site that already touches the topic loosely. If you’ve got a page ranking position 40 for a related term and it’s picking up impressions, that’s real data, not a model. I weight that above anything a third-party tool gives me.
Step two: build a proper click-through-rate curve
This is where most forecasts fall apart. People assume position 1 gets 30 percent of clicks and position 3 gets 10 percent, using averages that are years out of date and that ignore how much real estate Google now gives to AI overviews, featured snippets, People Also Ask boxes, and ads before the first organic result even appears.
From my own Search Console data across client accounts in 2025 and into 2026, here’s roughly what I’m seeing for commercial and informational queries that have an AI overview above them:
- Position 1: 8 to 15 percent CTR (down from the 25 to 30 percent that older SEO guides still quote)
- Position 2 to 3: 4 to 8 percent CTR
- Position 4 to 6: 1.5 to 3 percent CTR
- Position 7 to 10: under 1 percent CTR
If your query triggers a full AI overview with a summarised answer, cut all of those numbers again by roughly a third to a half. This is the single biggest reason forecasts built on old CTR tables are now wrong. If you’re still using a 2019 CTR curve to forecast 2026 traffic, you’re overestimating your results by two to three times.
Step three: check who you’re up against
This is the step everyone skips, and it’s the one that decides whether your forecast is real or wishful thinking. Pull the top 10 results for your target keyword and check three things for each: domain rating, referring domains to that specific page, and how recently it was updated.
Here’s the uncomfortable bit that most guides on this topic won’t say plainly: if you’re a domain rating 15 site and the top three results are DR 60 plus with hundreds of referring domains each, no amount of “great content” is going to get you to page one in 12 months. It might not get you there in three years. Volume and intent match mean nothing if you can’t win the competitive fight, and pretending otherwise is how content teams burn a year producing articles that sit on page four forever.
I learned this one the expensive way. Early on rebuilding my own business, I built a content plan around a set of keywords with decent volume, 1,200 to 2,500 searches a month each, all closely matched to my services. I forecast based on the volume and a rough CTR estimate and told myself we’d see meaningful traffic within six months. What I hadn’t checked was that every single page ranking in the top five was from a site with ten times my authority and years of backlinks I simply didn’t have yet. Six months in, I had articles sitting at position 24, 31, and 40. Traffic was close to nothing. The content wasn’t bad. The forecast was wrong because I’d skipped the competitive check and just multiplied volume by an optimistic CTR.
What I do now, and what I’d tell anyone building a plan, is grade every target keyword before it goes near a content calendar: green if your domain rating is within 15 points of the weakest page-one result, amber if you’re 15 to 30 points behind but the top result is old or thin, red if you’re more than 30 points behind competitors with fresh, well-linked content. Red keywords get parked, not written, until your domain authority catches up.
Step four: put a real timeline on it
Even a well-matched, winnable keyword rarely ranks fast. For a page on an established domain with decent internal linking, I typically see meaningful movement (page two to page one) somewhere between month 4 and month 9, assuming the page gets at least a handful of quality backlinks or internal links pointing to it. For a newer domain with limited authority, stretch that to 9 to 18 months, and build your forecast around a curve, not a single end number. A forecast that says “5,000 visits a month by month 12” without showing the ramp from month 1 to month 12 is a guess dressed up as a plan.
A simple ramp I use looks like this for a realistically winnable keyword at 1,000 searches a month:
- Months 1 to 3: page indexed, ranking 40 to 80, under 5 clicks a month
- Months 4 to 6: ranking 15 to 30, 10 to 40 clicks a month
- Months 7 to 9: ranking 6 to 12, 30 to 80 clicks a month
- Months 10 to 12: ranking 3 to 8, 60 to 150 clicks a month
That’s a single keyword, single page. Multiply across a realistic content plan of 20 to 40 pages, weighted by how many are green, amber, and red in your competitive grading, and you’ve got a forecast you can defend to a client, a boss, or yourself.
Where this connects to the rest of your marketing plan
A traffic forecast isn’t just an SEO exercise, it’s a business decision, and it should sit alongside the other channels you’re weighing up. If you’re deciding between building organic content and, say, guest posting to build authority faster, your forecast should show you honestly which route gets you winnable rankings sooner. Sometimes the answer is that six months of guest posting to lift your domain rating is a better use of budget than six months of blog content that’s forecast to sit at position 35 the whole time.
It’s worth looking at how bigger brands think about this too. When I looked at the Boohoo marketing strategy for a separate piece, one thing stood out: they don’t chase every keyword with volume, they pick the fights they can win and go hard on those. That’s the same discipline a forecast forces on a smaller business, just at a much smaller scale and with far less room for wasted spend.
It’s also worth studying what happens when companies skip this kind of grounded thinking entirely. A lot of the examples in famous brand failures come down to the same root cause as a bad content forecast: believing the opportunity was bigger and easier than the competitive reality allowed.
A quick worked example
Say you run a small accountancy practice and you’re deciding whether to write a page targeting “tax return help for freelancers,” which a keyword tool shows at 2,400 searches a month. Before committing a writer to it, here’s the five-minute check:
- Search Console cross-check: do you already have any page picking up impressions for related terms? If yes, note the actual numbers, not the tool estimate.
- SERP check: pull the top 10, note their domain ratings and roughly how many referring domains the ranking page has.
- Grade it: if you’re DR 25 and the top five are all DR 45 plus government or major accountancy brand sites, this one’s red. Park it.
- Look for the gap: check the same SERP for a long-tail variant like “self assessment help for freelance graphic designers,” maybe 90 searches a month, but with weaker, thinner competition. That’s green, and it’s winnable in four to six months rather than never.
- Forecast the green one using the ramp above, and build your content plan around a stack of these winnable long-tails rather than one unwinnable head term.
This is also exactly the kind of niche-first thinking I used when researching where restaurant businesses have room to grow: the biggest, most obvious markets aren’t automatically the best ones to fight for, and the same logic applies keyword by keyword, not just city by city.
The tools that help, and the ones that lie to you
Ahrefs and Semrush are fine for the SERP and backlink side of this, and Google Search Console is non-negotiable for your own real data. Where I’d be careful is with any tool that gives you a single “predicted traffic” number without showing its CTR assumptions or how it’s judging competitive strength. If you can’t see the working, don’t trust the output. The same caution applies to AI-generated content forecasts and strategy tools more broadly. A lot of businesses are now bolting AI into their SEO process without checking what it’s basing its numbers on, and that’s the same failure pattern I’ve written about in why AI implementations fail: the tool looks confident, the underlying data is thin, and nobody checks before committing budget.
The other quiet truth worth saying out loud: forecasting traffic is not the same as forecasting revenue, and a lot of content plans get built to hit a traffic number that means nothing commercially. I’ve had clients thrilled about a page hitting 400 visits a month that generated zero enquiries, and quietly worried about a page getting 40 visits a month that generated three. If you’re going to forecast, forecast the traffic and then sanity check it against a realistic conversion rate for that page type, because a beautifully accurate traffic forecast attached to the wrong keyword intent is still a waste of a content plan.
Putting it together before you write a word
The whole process, done, takes half a day for a content plan of 20 to 30 keywords, not weeks. Pull volume, cross-check with Search Console, grade competitively, build the CTR-adjusted ramp, and only then hand topics to a writer. It feels slower at the start. It saves months of writing content that was never going to rank, which is the far more expensive kind of slow.
Frequently asked questions
How accurate can an SEO traffic forecast be?
Done, with real competitive grading and an up-to-date CTR curve, you can usually land within 20 to 30 percent of actual traffic over a 12-month window. Forecasts built purely on keyword tool volume with no competitive check are often out by 200 to 300 percent, almost always on the optimistic side.
What’s the biggest mistake people make when forecasting SEO traffic?
Skipping the competitive check. Search volume tells you demand exists, it says nothing about whether your specific domain can win that demand within a reasonable timeframe. A high-volume keyword with page-one competitors who have ten times your domain authority is not a keyword you should be forecasting traffic for at all.
How long should I wait before judging whether a forecast was right?
Give it a minimum of 6 months for any newly published page, and 9 to 12 months if your domain has limited existing authority. Checking rankings at month 2 or 3 and calling the content a failure is one of the most common ways businesses abandon content that would have worked given proper time.
Do AI overviews and featured snippets change how I should forecast?
Yes, significantly. Any query that triggers a full AI overview or answer box should have its expected click-through rate cut by roughly a third to a half compared with older CTR tables, because a meaningful share of searchers now get their answer without clicking through to any website.
Useful references
Related reading: What Is Content Led SEO and How to Build a Strategy Around It and Why SEO Traffic Forecasts Are Usually Wrong.
This builds on my main SEO guide, my main guide on the topic.
Related: write for our digital marketing blog
If you would rather not build this yourself, here is write for us about SEO.
