Why is your organic traffic dropping, and which losses matter?
The sessions line in your analytics has turned down and someone wants a recovery plan. Before you commission one, here is how to name the cause in Search Console and price the lost visits in conversions.
- Most organic declines trace to five causes. AI answers on the results page, core and spam updates, demand moving elsewhere, lost rankings and technical or tracking faults each leave a different signature in Search Console.
- AI summaries cut clicks, and the studies agree on direction. Pew measured clicks on 8% of visits with an AI summary against 15% without, and Ahrefs reports a 58% lower click-through rate for the top result.
- Diagnose from cheap to expensive. Confirm tracking, rule out technical faults, match dates to confirmed updates, then split by query type and page type.
- Price the drop in conversions, not visits. In a hypothetical B2B example, visits fall 22.8% and demo requests 7.1%, because the pages that lose visits are not the ones that convert.
Google’s head of Search wrote in August 2025 that total organic click volume from Search to websites was “relatively stable year-over-year”. Independent panels measure something else: SparkToro’s 2026 study of Similarweb data found that 68.01% of US Google searches from January to April 2026 ended without a click. If your organic traffic is dropping, the cause is almost always one of five, and each one can be told apart in Search Console in under an hour.
The five are AI answers on the results page, core and spam updates, demand moving to AI assistants and other surfaces, rankings lost to competitors, and technical or tracking faults. Naming the cause is the quick part. The first job after that is not to win every visit back. It is to find out whether the lost visits ever produced conversions from organic search (purchases for a store, demo requests for B2B). A drop in visits is a traffic event, and whether it is a business event depends on which pages lost them.
This article gives the signature of each cause, the click-through data on AI Overviews, a diagnosis order, and a calculation that turns lost visits into lost conversions by page type. One worked example runs throughout: a B2B software site that counts demo requests.
Why organic traffic drops: the five causes
Each cause moves impressions, click-through rate and average position in a different combination, so the Performance report identifies it before any tool or agency is involved. Compare the three months before the drop with the three months after, then read which of the three numbers moved.
AI Mode and AI Overviews do not get their own filter. Google reports their clicks and impressions inside the normal Web search type, with no separate breakdown, so the AI effect has to be inferred from the pattern: impressions hold, position holds, clicks fall. Also, Google’s own guidance on traffic drops lists seasonality and changing interest as a cause in its own right, so widen the date range to 16 months before deciding that anything broke.
| Cause | Signature in Search Console | Pages it hits | What to do |
|---|---|---|---|
| AI Overviews and AI Mode answer the query | Impressions and position steady, click-through rate down | Informational posts, definitions, how-to guides | Price the loss before acting. Aim to be cited on the queries that convert |
| Core or spam update | Position and impressions fall together from a date inside a confirmed rollout | Often whole sections, sometimes the whole site | Compare the pages that fell with the ones that replaced them. Update, merge or remove |
| Demand moved to AI assistants and other surfaces | Impressions fall while position holds, including on brand queries | Topic by topic, gradual | Check search interest over 16 months. Measure AI referral visits separately |
| Competitors took the ranking | Position falls on specific queries, no update date, other queries unaffected | Individual commercial and comparison pages | Recover these first. They usually carry conversions |
| Technical or tracking fault | One directory, template or device drops overnight, or Search Console clicks hold while analytics sessions fall | Whatever shares the template, tag or rule | Fix the fault. No content work is needed |
Other surfaces deserve a short mention: some informational demand has moved to video platforms, forums and AI assistants, and it shows up as falling impressions, not falling position. That is a reason to measure where demand went, not a reason to start a channel programme from an SEO report.
Causes overlap. A core update can land in the same month that AI Overviews expand on your queries. When two signatures appear together, split the data by page type first (section four), because the split tells you how much of each loss is worth diagnosing further.
What the click data on AI Overviews shows
Every independent study published in 2025 and 2026 finds fewer clicks when an AI summary is present, and the studies disagree only on the size. The three most useful ones measure different things, which is why their numbers differ.
Pew Research Center tracked the browsing of 900 US adults in March 2025. Users clicked a traditional result link on 8% of visits to a results page with an AI summary and on 15% of visits without one. They clicked a link inside the summary itself on 1% of visits, and ended the browsing session on 26% of pages with a summary against 16% without.
Ahrefs compared 300,000 keywords, half with an AI Overview and half without, between December 2023 and December 2025. Its February 2026 update puts the click-through rate of the top-ranking page 58% lower where an AI Overview appears.
The third study adds the nuance a diagnosis needs. Seer Interactive’s April 2026 update covers 53 brands, 5.47 million queries and 2.43 billion organic impressions. Organic click-through rate on queries with an AI Overview hit a floor of 1.31% in December 2025 and recovered to 2.36% in February 2026, still below the 3.82% on queries without one. Across 2025, brands cited inside the AI Overview averaged 2.07% organic click-through rate against 0.94% for brands that were not cited.
Three conclusions follow for anyone reading their own report:
- The loss is real and concentrated. It falls on queries that an answer can satisfy, which are mostly informational. See how zero-click searches change what a ranking is worth.
- Being cited roughly doubles the clicks that remain. That is an argument for citation work on a short list of queries, covered in how to optimize for Google’s AI Overviews, not on every post.
- AI Mode is not yet the main cause. SparkToro found that only 0.34% of searches went to AI Mode from January to April 2026. The summaries on the standard results page do the damage.
None of these studies measures conversions. They count clicks, and a click that an AI summary absorbs was, on average, the click least likely to buy. That is the gap the rest of this article closes.
How to diagnose an organic traffic drop
Work from the cheapest explanation to the most expensive: measurement first, then technical faults, then update dates, then the split by query and page type. Most teams run this in reverse. They start with a content audit and discover the broken tag three weeks later.
- Check that the drop exists. Compare Search Console clicks with organic sessions in analytics for the same weeks. If clicks hold and sessions fall, the cause is a consent banner, a tag change or channel grouping, and there is no SEO problem to solve.
- Rule out technical faults. Filter by directory, template and device. A fall confined to one of them points to indexing, robots rules, canonicals, redirects or a migration. Read the Page indexing report, then the Security issues report and the report for any manual action.
- Match the start date against confirmed updates. Use the Google Search Status Dashboard, not third-party volatility charts. A fall that begins inside a rollout window and shows falling position is an update effect.
- Split by query type. Separate brand from non-brand queries, then informational from commercial. Steady impressions with falling click-through rate on informational queries is the AI Overview pattern. Falling brand impressions is a demand problem that no content will fix.
- Split by page type and attach conversions. Group landing pages into informational posts, comparison and commercial pages, product or service pages, and brand or home. Put visits lost next to the conversion rate of each group. This is the step that decides the plan.
For step three, these are the ranking incidents on the Google Search Status Dashboard from December 2025 to the date of this update. Two core updates and four spam updates have been logged in 2026 so far, on top of the core update that ran through the last weeks of 2025.
A date match is evidence, not proof. If your fall began on 27 March 2026, the March core update is the first suspect. If it began in mid-January, between rollouts, look harder at steps one, two and four. Keyword-level work comes after this, and only for the pages that survive the split: see keyword research when AI answers part of the query.
Which lost visits cost you conversions?
The lost visits that matter are the ones that used to convert, and they are rarely where the largest drop is. Informational posts take most of the AI Overview losses and carry a small share of conversions. Commercial and product pages lose fewer visits and each of those visits is worth many times more.
The calculation is one line per page type: visits lost × conversion rate of those pages = conversions lost. Use each page type’s own organic conversion rate from before the drop, not the site average. The example below is a B2B software site with 100,000 organic visits a month that counts demo requests. Hypothetical example
| Page type | Visits lost a month | Organic conversion rate | Demo requests lost a month |
|---|---|---|---|
| Informational blog posts | 21,000 (60,000 to 39,000) | 0.1% | 21 |
| Comparison and commercial pages | 800 (8,000 to 7,200) | 2.0% | 16 |
| Product and service pages | 600 (12,000 to 11,400) | 2.5% | 15 |
| Brand and home | 400 (20,000 to 19,600) | 1.5% | 6 |
| Total | 22,800 (down 22.8%) | Blended 0.82% before | 58 of 820 (down 7.1%) |
Read the two totals together. Visits fell 22.8% and demo requests fell 7.1%. Blog posts account for 92.1% of the lost visits and 21 of the 58 lost demo requests. Commercial and product pages lost 1,400 visits between them, 6.1% of the total, and 31 demo requests, more than half of the damage. A recovery plan ranked by visits would spend almost everything on the blog. A plan ranked by demo requests starts with two page types that barely show on the sessions chart.
A drop in visits is a traffic event, and whether it is a business event depends on which pages lost them.
Common mistake: reading the sitewide sessions line and responding by publishing more informational content. That adds pages to the one category where AI summaries absorb the click, and it leaves the commercial pages that lost rankings untouched. Split the drop by page type before approving any content budget.
Assisted conversions are the first gap. Last-click reporting undercounts informational pages that start journeys, so before you write off a post, check what it assists and how often its readers return through brand search. The calculation also needs volume: with only a handful of conversions per page type per month, the rates are noise, so widen the window to six or twelve months.
What to do about the losses that matter
Recover the visits that converted, stop paying to recover the ones that did not, and convert more of the traffic that still arrives. In that order, because the first is urgent, the second frees budget and the third is the only one fully in your control.
Recover commercial and product pages first. For each page that lost position, open the results page and compare it with what now ranks above it: fresher data, clearer pricing, a comparison table, proof. Update the page, merge thin pages that compete for the same query, and remove ones that serve no query at all. This is ordinary SEO work, aimed at a short list.
Stop chasing informational clicks that an answer now satisfies. If a post lost half its clicks with position unchanged, another rewrite will not bring them back. Keep the post if it earns citations, links or assisted conversions. Otherwise leave it alone and move the hours.
Get cited where it matters. Seer’s data shows cited brands keep roughly twice the click-through rate of uncited ones. Pick the queries closest to a buying decision and work on those. Google states that no special markup or file is required, and generative engine optimization is still mostly SEO.
Convert more of what you still earn. The visitors who click through after reading an AI summary have already had the basic answer. They arrive later in the decision, and pages built for a first-touch reader waste them. Start with whether a low rate comes from traffic quality or the website, then compare your rates by page type with conversion rate benchmarks by industry.
- Commercial and product pages that lost position
- Outdated pages on queries that convert
- Thin pages competing for one query: merge them
- Informational posts with steady position and fewer clicks
- Keep them if they earn citations or assisted conversions
- Citations in AI answers on decision-stage queries
- Conversion rate on the organic visits you still earn
- Publishing more informational content to refill the sessions line
The last point is the one most plans leave out, and the arithmetic favours it. In the example above, winning back every lost blog visit returns 21 demo requests a month. Lifting the conversion rate of the commercial and product pages by a tenth, on the visits they still receive, returns 43. Hypothetical example
The full method for that second option is in the cluster guide on combining CRO and SEO to convert the organic traffic you already rank for.
If revenue comes from advertising on pageviews, every lost visit is lost revenue and the page-type split changes nothing. If Search Console shows a manual action or a security issue, resolve that before any prioritisation. And if a migration or technical fault caused the drop, fix it for every page type: there is no trade-off to make.
Where Pathmonk fits
Pathmonk works on both halves of the calculation: the organic visits a site earns and the share of them that convert. It does not change what Google’s summaries do to informational clicks, and nothing does.
On the acquisition side, Pathmonk’s SEO agent plans, drafts and publishes content using the site’s own first-party behaviour data, so the content plan follows what visitors do on the site.
On the conversion side, Pathmonk reads visitor intent in real time from first-party, cookieless, in-session behaviour such as scroll, navigation and dwell. It predicts the buying stage (awareness, consideration or decision) and shows stage-matched microexperiences. The page’s indexable content stays as it is, no developers are needed, and results are measured against a control group.
The published SEO results are on the SEO case study page. Across the customers shown there, the averages are a 35% organic traffic lift and a 63% organic impressions lift in the first three months. By this article’s own argument, traffic and impressions are the start of the calculation, not the result. The figure to look at is the one stated in money: Joto Imaging Supplies, a B2B e-commerce company, reports $60k extra organic revenue in month 3.
Key takeaways
- Organic declines in 2026 come from five causes: AI answers on the results page, core and spam updates, demand moving elsewhere, rankings lost to competitors, and technical or tracking faults.
- Each cause has its own signature in Search Console. Steady impressions with falling click-through rate points to AI Overviews. Falling position from a rollout date points to an update.
- Pew, Ahrefs and Seer Interactive all measure fewer clicks when an AI summary appears. SparkToro puts zero-click US Google searches at 68.01% for January to April 2026.
- Diagnose in order of cost: tracking, technical faults, update dates, query type, then page type with conversions attached.
- Visits lost multiplied by the conversion rate of those pages gives conversions lost. Run it per page type, never on the site average.
- The page types that lose the most visits are usually not the ones that carry the conversions. Rank the recovery plan by conversions.
- Recover commercial and product pages, hold informational posts that still assist, and stop publishing more informational content to refill the chart.
- Raising the conversion rate of the organic visits you still earn can return more than winning back the visits you lost.
FAQs on losing organic traffic
Why is my website losing organic traffic?
Most declines come from one of five causes: AI Overviews answering informational queries on the results page, a Google core or spam update, demand moving to AI assistants and other surfaces, competitors taking rankings, or a technical or tracking fault. Search Console tells them apart by which of impressions, click-through rate and position moved.
How do I know if AI Overviews caused my traffic drop?
Look for steady impressions and steady average position with a falling click-through rate, concentrated on informational queries. Google reports AI Overview and AI Mode traffic inside the normal Web search type with no separate filter, so the cause is inferred from that pattern, not read from a dedicated report.
How do I check whether a Google update hit my site?
Find the first day of the decline in Search Console and compare it with the rollout dates on the Google Search Status Dashboard. If the decline starts inside a confirmed rollout and average position fell with it, an update is the likely cause. In 2026 the dashboard lists core updates starting 27 March and 21 May.
Should I try to recover all the organic traffic I lost?
No. Multiply the visits lost on each page type by that page type’s conversion rate to see what the loss cost. Recover commercial and product pages first. Informational posts that kept their position and lost clicks to AI summaries rarely justify more investment unless they assist conversions or earn citations.
Does a drop in organic traffic always mean fewer conversions?
No. In the hypothetical example in this article, visits fell 22.8% and demo requests fell 7.1%, because most of the lost visits came from informational posts with a low conversion rate. The reverse also happens: a small drop on commercial pages can cost more than a large one on the blog.
What should I check first when organic traffic drops suddenly?
Check measurement before anything else. Compare Search Console clicks with organic sessions in analytics for the same period. If clicks are steady and sessions fell, a tag, consent banner or channel grouping changed. Then check whether the drop is confined to one directory, template or device, which points to a technical fault.
Is publishing more content the right response to falling organic traffic?
Usually not. Publishing more informational content adds pages to the category where AI summaries absorb the most clicks. The better sequence is to recover commercial pages that lost rankings, update or merge outdated pages, and raise the conversion rate of the organic visits the site still earns.
- Pew Research Center, Google users are less likely to click on links when an AI summary appears in the results (2025) (pewresearch.org)
- Ahrefs, AI Overviews reduce clicks, updated study (2026) (ahrefs.com)
- Seer Interactive, AI Overview impact on Google click-through rate, update (2026) (seerinteractive.com)
- SparkToro, In 2026, less than one third of Google searches still send a click (2026) (sparktoro.com)
- Google, AI in Search is driving more queries and higher quality clicks (2025) (blog.google)
- Google Search Status Dashboard, ranking incident history (read 7 October 2026) (status.search.google.com)
- Google Search Central, Debugging drops in Google Search traffic (read 2026) (developers.google.com)
- Google Search Central, AI features and your website (2025) (developers.google.com)