How much organic traffic could AI answers put at risk for your site?
This free AI traffic impact calculator turns three scenarios you set into a range of organic sessions, conversions and value that AI answers could put at risk. It has no built-in assumptions, because published estimates of click loss disagree widely. It runs in your browser and nothing you enter is sent anywhere.
Part 1
Use the tool
Everything runs in your browser. Nothing you paste is sent anywhere.
1. What are your numbers and assumptions?
Enter your sessions and three scenarios of your own. The tool multiplies them out and shows a range. It has no built-in assumptions. Each calculation uses one of your free tool runs.
Part 2
How it works, how to read it, and its limits
The method, what the result means, how it differs from other tools, and what it cannot do.
2. How is the number worked out?
Sessions at risk equal your informational sessions, times the share of those queries that show an AI answer, times the share of clicks lost when one shows.
| Step | Calculation |
|---|---|
| Informational sessions | Monthly organic sessions x share from informational queries |
| Sessions at risk | Informational sessions x share of queries with an AI answer x click loss |
| Conversions at risk | Sessions at risk x conversion rate (if you enter one) |
| Value at risk | Conversions at risk x value per conversion (if you enter one) |
You choose the assumptions. A range from low to high scenarios is more honest than one figure.
3. What do published studies say about click loss?
They disagree, because they measure different things, and the newest evidence shows a partial recovery, which is why the tool asks for your assumptions instead of supplying one.
| Study | What it measured | Finding | Caveats stated or visible |
|---|---|---|---|
| Amsive, April 2025 (opens in a new tab) | 700,000 keywords on 10 websites in 5 industries (finance, education, SaaS, healthcare, pets); 10,000 that triggered an AI Overview and already ranked | Average CTR change of -15.49% for keywords with an AI Overview: -19.98% for non-branded, +18.68% for branded | Date range not stated; 10 sites only; written by an agency that sells SEO |
| Ahrefs, updated February 2026 (opens in a new tab) | 300,000 keywords (150,000 with an AI Overview, 150,000 informational without); desktop CTR from Search Console; December 2023 against December 2025 | Position-one CTR 0.073 to 0.016, which Ahrefs says correlates with a 58% lower CTR than its forecast; -50.8% at position 2 down to -19.4% at position 10 | A correlation built on a forecast from the control group, not a direct measurement; desktop only; Ahrefs cites other studies from 47.5% to 65.2% |
| Seer Interactive, September 2025 (opens in a new tab) | 3,119 informational queries across 42 organizations, June 2024 to September 2025 | Organic CTR with an AI Overview 1.76% to 0.61% (-61%); without one 2.74% to 1.62% (-41%) | Authors cannot prove causation; informational queries only; CTR also fell without an overview |
| Seer Interactive, April 2026 update (opens in a new tab) | 53 brands and 5,471,127 queries, January 2025 to February 2026, all intents | Informational 2025 averages: brand cited 2.07%, not cited 0.94%, no AI Overview 3.35%; actual CTR in January and February 2026 beat the earlier projection; all-intent CTR with an overview 1.31% in December 2025, 2.36% in February 2026 | Described as directional; authors cannot claim causation; one account is 47% of one segment |
- The headline drops range from about 15 to 20 percent (Amsive, all non-branded keywords) to about 58 to 61 percent (Ahrefs and Seer, position one or informational queries). They are not the same measurement.
- Being cited matters in the Seer data: informational queries where the brand was cited kept a higher click-through than those where it was not. The authors say they cannot show that citation causes this.
- Click-through on queries with no AI Overview also moved, so part of any drop is not caused by the overview.
- The newest Seer data shows a partial recovery in early 2026, so an older figure may overstate today's loss.
- Google disputes the framing, arguing that the clicks that remain are higher quality.
Confirmed The figures above were read on the studies' own pages on 2 November 2026. Each is from a vendor or agency with a commercial interest, so treat them as directional, not as a benchmark for your site.
If you want published anchors for your scenarios, a low scenario near the Amsive non-branded figure and a high scenario near the Ahrefs figure are two defensible ends, but they measure different things. Better than any benchmark is your own data: compare Search Console clicks and click-through for the same query group before and after, as in the query sorter.
4. How is this different from other traffic-loss calculators?
It does not pick an assumption for you, and it keeps your numbers in your browser.
| Question | Typical calculators | This calculator |
|---|---|---|
| Click loss | A built-in figure from one study | Three scenarios you set, with the studies listed above for context |
| Result | One estimate of lost traffic | A low to high range |
| Your data | Entered into a vendor page | Stays in your browser |
A. How do you measure your own click loss?
Compare Search Console clicks and click-through for the same group of queries before and after AI answers appear for them.
- Export your queries and sort them into groups, for example with the query sorter.
- Pick two equal date ranges, one before and one after the change you care about.
- Compare impressions, clicks and click-through for each group.
- Check the Generative AI performance report in Search Console, which Google points to for traffic from its AI features.
- Use what you measure to set your low, middle and high scenarios.
Confirmed Google says third-party tools cannot see its internal metrics, so check any tool's claims against Google's own reports.
B. How do you choose the three scenarios?
Make the low scenario one you almost certainly exceed and the high one a stretch you can still defend.
- Base the share of queries with an AI answer on a sample of your own queries, checked by hand.
- Base the click loss on your before-and-after data where you have it.
- Keep the middle scenario between them and explain why.
- Update the numbers each quarter.
The example numbers behind the "Load example" button are made up to show the layout and are not a forecast.
5. Questions about AI traffic impact
Short answers on assumptions, ranges and measurement.
How much traffic will AI Overviews take from my site?
It depends on how many of your queries show an AI answer, which query types you rank for, and whether you are cited. Studies range from an average click-through drop of about 15 to 20 percent (Amsive) to about 58 to 61 percent (Ahrefs, Seer), with newer Seer data showing a partial recovery in early 2026. Use scenarios, then measure your own Search Console data.
What is a good click-loss assumption?
There is no single one. Use a low, middle and high scenario that you can defend, and update them with your own before-and-after data.
Why a range and not one number?
Because the inputs are uncertain. A range shows how much the answer depends on the assumptions.
Does this include conversions?
Yes, if you enter a conversion rate and a value. Otherwise it shows sessions only.
Is my data sent anywhere?
No. It is calculated in your browser.
6. What can this calculator not do?
It multiplies the assumptions you give it. It does not know your queries or forecast traffic.
- It is arithmetic on your assumptions, not a forecast.
- It treats all informational queries alike.
- It does not model being cited, or traffic that arrives from AI assistants.
Next: sort your queries → see the full audit → see all the free tools →