- Calculate the monthly and annual revenue AI Overviews are taking from your organic traffic.
- Identify the exact citation share you need just to stand still — a line most sites have never measured.
- See a 24-month projection as AI Overview coverage climbs, with a marker splitting realised loss from projected.
- Compare your result against four published benchmarks to see which input the number actually depends on.
Your result, above
Milestones
The 24-month projection, in numbers
Same straight-line ramp as the chart above, read out at six checkpoints. A planning scenario, not a forecast — the real path will not be a straight line.
| Month | AI coverage | Monthly at risk | Cumulative loss |
|---|
How we get there
| Monthly organic sessions | — |
|---|---|
| Exposed to an AI answer | — |
| Gross sessions at risk | — |
| Recovered via citation | − |
| Net sessions at risk | — |
Where your traffic actually goes
0 not exposed to an AI answer
0 exposed, click retained
0 exposed, recovered via citation
0 net lost to AI answers
The number that matters most
The citation share you need just to stand still
Recovery only offsets loss above this threshold. Below it, a rising citation share is still a losing trade against a growing AI Overview.
—%
—Stress test
What happens when one assumption moves
Every one of these is a plausible correction, not a disaster scenario. If your exposure only looks manageable in the default case, it is not manageable.
Methodology
How this calculator works
Published in full so you can check it, argue with it, or quote it. Every figure below is derived from your inputs — nothing is hard-coded except the sources used to set the defaults.
-
1
Work out who is even exposed
exposedSessions = sessions × informational share × AI Overview coverage. Only the slice of your traffic on question-shaped queries where an AI answer currently appears is at risk of anything at all. -
2
Apply the click-through hit
grossAtRisk = exposedSessions × CTR loss. Not every exposed session is lost — this is the share that would have clicked through and now does not, based on the CTR reduction you set. -
3
Subtract what citation wins back
recovered = grossAtRisk × citation share × recovery factor. The recovery factor is the weakest-evidenced number in this calculator — it extrapolates from a per-impression click-rate comparison, not a measured recovery of specific lost sessions. Treat this step as informed extrapolation, not a hard finding. -
4
Turn sessions into money
netSessionsAtRisk = grossAtRisk − recovered, then multiplied by your conversion rate and value per conversion for a monthly figure, and by 12 for the annual one. -
5
Solve for the break-even citation share
The citation share at which recovered sessions fully offset gross sessions at risk is
10,000 ÷ recovery factor(as a percentage). At a 60% recovery factor that is 166.7% — impossible — which is the honest way of saying citation alone cannot fully offset the loss at that recovery factor.
Benchmark sources
- Zero-click search rate — SparkToro / Similarweb clickstream analysis, June 2026: US zero-click Google searches reached 68.01% in January–April 2026, up from 60.45% in 2024.
- Click-through with an AI summary present — Pew Research Center, 22 July 2025: users clicked a web result 8% of the time with an AI summary present versus 15% without, and clicked the AI-cited source itself roughly 1% of the time.
- CTR impact on the top-ranking page — Ahrefs: presence of an AI Overview correlates with roughly 34.5% lower click-through for the top-ranking organic result, across a 300,000-keyword sample.
- Zero-click rate with vs without an AI Overview — Similarweb: roughly 83% zero-click when an AI Overview is present versus roughly 60% when absent.
- Citation lift — Semrush: brands cited in an AI Overview earn roughly 120% more organic clicks per impression than uncited brands for the same queries.
- AI Overview coverage growth — Ahrefs' SERP-feature tracking: the share of Google keywords triggering an AI Overview grew from roughly 1.5% in September 2024 to roughly 32% by September 2025.
The recovery factor is the weakest-evidenced input in this model. The Semrush 120% figure is a per-impression click-rate comparison between cited and uncited brands — it is not a controlled measurement of specific lost sessions recovered, and this tool is extrapolating from it. The 24-month projection is a straight-line ramp of a single variable with everything else held constant. It is a planning scenario for stress-testing a decision, not a forecast of what will happen.
Questions people ask about AI Overviews and lost traffic
How much traffic do AI Overviews actually take?
A lot, and it is measurable. Pew Research Center found that when a Google results page carried an AI summary, people clicked a web result 8% of the time versus 15% without one — roughly half — and clicked the cited source itself only about 1% of the time (22 July 2025). Ahrefs separately measured roughly 34.5% lower click-through for the top-ranking page once an AI Overview appears, across 300,000 keywords. Similarweb puts the zero-click rate at roughly 83% with an AI Overview present versus 60% without.
Is zero-click search getting worse?
Yes, on two independent measures. SparkToro and Similarweb's clickstream analysis puts US zero-click Google searches at 68.01% in January–April 2026, up from 60.45% in 2024 (June 2026). Separately, Ahrefs' SERP-feature tracking shows the share of Google keywords triggering an AI Overview growing from roughly 1.5% in September 2024 to roughly 32% a year later. Both curves point the same way: more searches end on the results page, and more results pages carry the feature that causes it.
Does being cited in an AI answer make up for the lost click?
Only partly, and honestly nobody has firmly measured how much. Semrush reports that brands cited in an AI Overview earn roughly 120% more organic clicks per impression than uncited brands for the same queries — but that is a per-impression click-rate comparison, not a controlled measurement of specific lost sessions recovered. This calculator's "recovery factor" extrapolates from that figure and is the weakest-evidenced input in the model. Treat citation as a partial offset, not a fix.
Which pages lose the most traffic to AI answers?
Informational and definitional pages — "what is X", "how does X work" — lose the most, because an AI answer can fully substitute a fact with a citation nobody needs to click through to read. Transactional, tool, comparison and local-intent pages lose the least, because an AI answer can summarise a decision but cannot complete a purchase, run a calculator, or book an appointment for the reader. If your traffic is mostly definitional, your exposure is structurally higher than a competitor selling a transaction.
What can I actually do about it?
Three things, in order. First, make pages easy to cite — direct answers near the top, clear structured data, unambiguous entities — because citation is the only lever this model shows recovering any of the loss. Second, shift new content investment toward transactional, comparison and tool-shaped pages, which AI answers displace least. Third, track your citation share over time the way you would track rankings; it decides whether this exposure is a manageable tax or a slow bleed.
Common follow-ups
Does anything I type here get sent anywhere?
No. The calculator is a single JavaScript file running in your browser. There is no server call, no analytics event carrying your numbers, no account, and no storage. If you copy the result link, your inputs are encoded in the URL — so treat that link the way you would treat the numbers themselves.
Does this tool measure my actual AI Overview citations for me?
No — coverage and citation share are inputs you estimate, not data this page crawls or fetches. Pull a sample of your ranking queries, check which trigger an AI answer and whether you are named or linked in it, and use that ratio. The Answer Engine Scorecard next door is a good next step for a structural read on how citable your pages are.
Why might my real drop not match Google Search Console?
Search Console reports impressions and clicks for queries you already rank on, but it does not tell you which of those results pages carried an AI Overview, or whether you were the cited source inside it. This model fills that gap with estimates; GSC is still the right place to get your baseline session count.
Is the 24-month projection a forecast of what will happen to me?
No. It is a straight-line ramp of one variable — AI Overview coverage — with every other input held constant, useful for stress-testing a decision such as "is this worth fixing now." Coverage will not actually move in a straight line, and your other inputs will move too. Re-run it with updated numbers periodically rather than treating one run as a prediction.
Next
A tool like this is an acquisition channel, not a giveaway.
This page exists to demonstrate a strategy: a small, genuinely useful tool answers a question people already type, earns links no blog post earns, and survives the shift to AI answers because the value is the interaction, not the fact. More of the toolkit and the full argument are next door.