Introduction
For years, SEO forecasting was almost boring. You pulled last year's organic traffic, drew a line that tilted up and to the right, added a growth assumption, and called it a plan. If you kept publishing and building links, traffic went up. The math rarely surprised anyone.
That straight line has snapped. A growing share of search demand is now answered before anyone clicks — inside AI Overviews, AI Mode, and chatbot answers that summarize your page without sending a visitor. According to Neil Patel's recent work on forecasting in an AI era, the old habit of projecting a single traffic number forward is now one of the fastest ways to set a budget you cannot hit. The uncomfortable truth for small-business owners is that your visibility can hold steady — or even improve — while your organic sessions fall.
That does not make forecasting useless. It makes honest forecasting more valuable than ever. A good forecast is a planning tool, not a promise, and in a volatile market the right move is to widen your confidence band rather than fake precision. Below is a repeatable, five-step method we use to build defensible traffic projections when a rising share of demand never becomes a click.
Key Takeaways
- AI Overviews and zero-click answers have broken the linear “traffic goes up if we keep doing SEO” assumption — visibility can hold while sessions fall.
- Start with a clean historical baseline pulled from Google Search Console (demand and clicks) and GA4 (what visitors did after arriving).
- Segment your queries: separate AI-exposed informational terms from high-intent, click-worthy terms, because they behave completely differently.
- Apply realistic click-through decay to AI-eligible queries instead of assuming a flat CTR that no longer exists.
- Forecast a range — conservative, expected, optimistic — and tie it to leads and revenue, not vanity sessions.
Why Do Old SEO Traffic Forecasts Break in the AI Era?
The classic forecast assumed a stable relationship between rankings and clicks: rank higher, earn a predictable slice of the search volume. That relationship has been quietly rewritten. AI Overviews now appear on a large share of searches — roughly 55% of all SERPs as of early 2026 per BrightEdge's 2026 measurement — and when they appear, they change how often people click at all.
The click data is where the old model falls apart. Seer Interactive's analysis of 3,119 queries found organic click-through rate on AI Overview queries fell from 1.76% to 0.61% — roughly a 61% drop — with paid CTR sliding from 19.7% to 6.34% over a comparable window. Independent work from the Pew Research Center reported similar behavior: pages accompanied by an AI summary saw about 8% of visits result in a click, versus about 15% without one, and 26% of users ended their session entirely after seeing an AI Overview compared with 16% when none appeared.
Zero-click behavior compounds the effect. The widely cited SparkToro and Datos clickstream study puts U.S. zero-click searches at 58.5% — meaning for every 1,000 Google searches, only about 360 send a click to the open web. If your forecast still multiplies “search volume × position CTR” using 2022-era click rates, it is projecting a world that no longer exists. This is the same dynamic we unpacked in our breakdown of the traffic drop AI Overviews are causing: the rankings can be intact while the clicks quietly leave.
How Do You Build a Clean Historical Baseline?
Every honest forecast starts with a trustworthy record of the past. Two data sources do the heavy lifting, and they answer different questions. Google Search Console (GSC) shows search-side demand — impressions, clicks, average position, and the specific queries bringing people to you. GA4 shows what happened after the click — engagement, conversions, and revenue events. You need both: GSC tells you what the search market offered, GA4 tells you what you actually captured.
To build the baseline, we recommend pulling at least 12 to 16 months of GSC data so you can see seasonality rather than mistaking a slow February for a decline. Export clicks and impressions by query and by page, then reconcile against GA4 organic sessions for the same period. Small gaps are normal — GSC and GA4 count slightly differently — but a large, growing gap between GSC clicks and GA4 organic sessions is itself a signal worth investigating.
A critical adjustment for 2026: your recent months are already contaminated by AI-Overview suppression, so a naive “last 3 months, annualized” baseline bakes today's decline into next year and can overstate the drop. We prefer to establish a pre-AI-Overview reference period, then measure the delta AI features introduced, so the forecast reflects the trend rather than a single distorted quarter. Google has also begun surfacing AI-search data inside Search Console, which helps you see which of your impressions are landing inside AI experiences rather than standard blue links.
How Do You Segment AI-Exposed Queries From the Rest?
Treating all queries the same is the single biggest forecasting mistake right now. AI Overviews do not hit every search equally. Per Ahrefs data reported by Omnibound, 99.2% of keywords that trigger AI Overviews are informational in intent — the “what is,” “how does,” and “why” questions. Transactional and high-intent queries, the ones where someone is ready to book or buy, are far less likely to be fully answered by a summary.
That distinction is the backbone of a modern forecast. We sort baseline queries into three buckets:
| Query segment | Example | AI-Overview exposure | Forecasting treatment |
|---|---|---|---|
| AI-absorbed informational | “what is a heat pump” | High | Apply steep CTR decay; expect click loss even if rankings hold |
| Hybrid / research | “best heat pump for cold climates” | Medium | Apply moderate decay; some clicks survive for comparison depth |
| High-intent / transactional | “heat pump installation quote Fort Wayne” | Low | Protect these; they still convert and drive revenue |
Doing this segmentation well is where intent gap analysis in Search Console pays off — it helps you find the queries where you rank but no longer capture the click, so you can price that loss into the forecast instead of being surprised by it. The goal is not to mourn the informational clicks; it is to know which part of your traffic is genuinely at risk and which part is durable.
How Do You Apply Realistic CTR Decay Instead of Flat CTR?
Once queries are segmented, replace the flat-CTR assumption with position-and-exposure-aware click rates. Baseline click curves have shifted downward across the board: ClickRank's 2026 benchmark compilation, citing First Page Sage, puts position-1 organic CTR at about 27.6% (down from 31.7% in 2022), position 2 near 15.8%, and position 3 near 11.0%. When an AI Overview is present, the same benchmark reports additional suppression in the range of roughly −34.5% to −58% depending on position.
Here is how the decay logic works in practice, using clearly hypothetical numbers to illustrate the method — these are not measured figures for any specific business:
- Suppose a query has 1,000 monthly searches and you rank #1. A pre-AI model might assume ~28% CTR, projecting ~280 clicks.
- If that query now triggers an AI Overview, apply the suppression band. A conservative estimate might discount the CTR by ~50%, projecting closer to ~140 clicks.
- For an AI-absorbed informational query, the realistic figure could be lower still.
The discipline is to apply different decay to each segment, not one blanket haircut. High-intent transactional queries keep something close to their historical CTR; AI-absorbed informational queries get the steepest cut. If a specific decay rate is not something you can measure from your own GSC data, say so and use a defensible range rather than inventing a precise percentage. Fabricated precision is the enemy of a forecast people can actually trust.
How Do You Build a Forecast Range Instead of a Single Number?
A single projected number implies a confidence you do not have. The stronger approach — and the one increasingly standard among analysts — is scenario forecasting: build three versions with named assumptions.
- Conservative: AI-Overview coverage keeps expanding, CTR decay lands at the harsh end of the band, and you hold rankings without gaining. This is your floor.
- Expected: Coverage grows modestly, decay lands mid-band, and your planned content and technical work earns incremental positions.
- Optimistic: You win featured positions and citations inside AI experiences, decay lands at the gentler end, and high-intent capture improves.
Each scenario should carry its assumptions in writing — the AI-Overview coverage rate, the CTR decay applied to each segment, and the ranking changes you expect from planned work. When the market moves, you adjust an assumption rather than scrapping the whole model. Widen the band when volatility is high: the CTR data itself has been unstable, with Seer's tracked AI-Overview CTR reportedly dropping to a floor and then partially rebounding into early 2026, so a wider confidence range is honest, not lazy.
Two guardrails keep this credible. First, never present the optimistic case as the plan — budget against the conservative-to-expected range. Second, revisit the forecast quarterly, because the underlying click rates are shifting faster than they used to. A forecast that is never revised is a forecast that is quietly wrong.
How Do You Tie a Traffic Forecast to Leads and Revenue?
A traffic number by itself is a vanity metric. The forecast that earns budget converts sessions into leads, and leads into revenue. This is also where AI-era nuance matters most, because the visitors who still click may be worth more per head than before.
Layer conversion assumptions onto each scenario using your own GA4 data where possible, and industry benchmarks where you lack history. The same ClickRank compilation, citing Semrush's 2026 data, puts baseline organic conversion around 2.7–3.0%, with local-services businesses often higher at roughly 4.0–6.0%. There is also early evidence that AI-referred visitors can convert at a stronger rate than traditional organic — Omnibound cites Semrush data suggesting AI-search visitors convert at meaningfully higher rates than standard organic — which means a smaller number of clicks does not automatically mean proportionally fewer customers.
To connect the dots without over-claiming, run the traffic range through a conversion rate to get a lead range, then through your close rate and average deal value to get a revenue range. Because AI muddies the picture of which channel deserves credit, pair this with a measurement discipline like blended customer acquisition cost and a structured five-layer GEO measurement framework so you are tracking outcomes, not just rankings. And because a growing share of your visibility now lives inside AI answers, the forecast should sit alongside a serious answer engine optimization plan — being cited in the summary is increasingly how the durable, high-intent clicks get earned in the first place.
What This Looks Like for a Fort Wayne Service Business
Bring this down to a real Northeast Indiana example. Picture a Fort Wayne HVAC company planning next year's marketing budget. In the old model, they would look at last year's organic traffic and assume steady growth. In 2026, that assumption would quietly overspend.
Segment their queries and the picture sharpens. Informational terms like “why is my furnace short cycling” or “what size AC do I need” are exactly the searches AI Overviews now answer in place — high impressions, shrinking clicks. Those get steep decay in the forecast. But high-intent, booking-stage queries — “emergency furnace repair Fort Wayne,” “AC installation quote Allen County,” “heat pump financing DeKalb County” — still send clicks to businesses ready to serve them, and they still convert. Local-services conversion rates tend to run on the higher end, so a smaller volume of the right clicks can carry the revenue plan.
The honest budget for this business is a range, not a single traffic number: a conservative floor that assumes AI keeps absorbing the “near me” research-stage queries and Maps surfaces take a bigger share, an expected case where their booking-stage pages hold and improve, and an optimistic case where they earn citations inside local AI answers. A local forecast should concentrate demand modeling on the Fort Wayne and Allen/DeKalb County terms that actually drive booked jobs, rather than the surrounding-town variations that carry little standalone volume. Budget against the conservative-to-expected band, protect the transactional queries, and treat the informational click loss as expected rather than a crisis.
Plan Your 2026 Traffic With a Forecast You Can Defend
If you are setting a marketing budget on the assumption that organic traffic keeps climbing, the AI-era math deserves a second look before you commit the spend. A defensible forecast separates the demand you can still capture from the demand AI now answers for free — and it hands you a range you can plan against instead of a number you will miss.
Get a Forecast Built on Your Real Data
At Button Block, we build these forecasts for small and mid-size businesses across Fort Wayne and Northeast Indiana, grounding them in your real GSC and GA4 history rather than industry averages. If you want an honest projection tied to leads and revenue — not vanity sessions — let us map what next year actually looks like for your business.
Frequently Asked Questions
- What is SEO forecasting in the AI era?
- SEO forecasting is the practice of projecting your future organic search traffic, and increasingly the leads and revenue it drives, based on historical data and market assumptions. In the AI era it must account for AI Overviews and zero-click answers that satisfy searches without a click, so a modern forecast segments queries by AI exposure and projects a range rather than a single number.
- Why did my SEO rankings hold but my traffic still fall?
- Because rankings and clicks are no longer tightly linked. When an AI Overview answers a query directly, users often do not click any result — Seer Interactive measured organic CTR on AI Overview queries falling from 1.76% to roughly 0.61%, and Pew Research found more sessions ending without a click. You can rank in the same position and still lose traffic as AI absorbs the click.
- How much historical data do I need to forecast SEO traffic?
- We recommend at least 12 to 16 months of Google Search Console and GA4 data so you can account for seasonality rather than mistaking a slow month for a trend. Because recent months are already affected by AI-Overview suppression, it also helps to identify a pre-AI reference period and measure the change AI features introduced, so you forecast the underlying trend rather than one distorted quarter.
- Should I forecast a single traffic number or a range?
- Use a range. Building conservative, expected, and optimistic scenarios with written assumptions is more honest than a single projection, because click-through rates on AI-exposed queries have been volatile. A range lets you budget against a defensible floor, adjust individual assumptions as the market shifts, and avoid promising precision that AI-era search cannot support.
- Which searches are most affected by AI Overviews?
- Informational queries are hit hardest. Ahrefs found that about 99.2% of keywords triggering AI Overviews are informational in intent — the "what," "how," and "why" questions. High-intent transactional queries, such as requesting a quote or booking a local service, are far less likely to be fully answered by an AI summary and still tend to send clicks that convert.
- How do I connect a traffic forecast to revenue without over-claiming?
- Run your traffic range through a realistic conversion rate to get a lead range, then apply your close rate and average deal value to estimate revenue. Use your own GA4 conversion data where you have it and clearly labeled industry benchmarks where you do not, and pair the forecast with measurement discipline like blended customer acquisition cost so you are attributing outcomes rather than guessing.
- How should a Fort Wayne small business forecast SEO traffic for 2026?
- Start with the same method, but weight it toward local intent. Segment your Google Search Console queries so that AI-absorbed research terms ("why is my furnace short cycling") are separated from high-intent, booking-stage searches ("emergency furnace repair Fort Wayne," "AC installation quote Allen County") that still send clicks and convert. Forecast a range rather than a single number, apply steep click decay to the informational terms, protect the transactional ones, and budget against the conservative-to-expected band. Because local-services conversion rates tend to run higher, a smaller volume of the right Fort Wayne and Allen/DeKalb County clicks can still carry the revenue plan.
Sources & Further Reading
- Neil Patel: neilpatel.com/blog/seo-forecasting — SEO Forecasting: How to Predict Your Traffic in an AI Era
- WordsAtScale: wordsatscale.com/ai-overviews-ctr-statistics-2026 — AI Overviews CTR Statistics 2026: Every Published Number on Clicks, Zero-Click & Citation Lift
- Omnibound: omnibound.ai/blog/zero-click-search-statistics — Zero-Click Search Statistics (2026): 52+ Data Points on Traffic Loss, AI Acceleration, and What Still Gets Clicks
- Omnibound: omnibound.ai/blog/ai-seo-statistics — AI SEO Statistics (2026): 57+ Data Points on Zero-Click Search, Rankings, and Organic Traffic Decline
- ClickRank: clickrank.ai/seo-benchmarks — SEO Benchmarks 2026: CTR, Traffic & AI Standards