If organic traffic is down but leads are up, the most common explanation in 2026 is that AI Overviews and AI assistants are absorbing low-intent informational clicks while higher-intent visitors, many of whom already researched you elsewhere, still arrive and convert. Before you celebrate, confirm it: check that leads are genuinely organic, that lead quality held, and that branded and AI referral traffic explain the lift.
Key Takeaways
- Falling sessions with rising leads is a real pattern, but it can also be an attribution artefact. Verify before reporting it.
- Split organic into branded, non-branded informational, non-branded commercial and AI referral. The story lives in the split, not the total.
- Measure leads per 1,000 organic sessions and qualified pipeline, not just form fills.
- AI Overviews hit informational queries hardest; Ahrefs found a 58% lower position-one CTR on informational keywords with AI Overviews (Dec 2023 vs Dec 2025).
- Self-reported attribution ("How did you hear about us?") is now essential, because AI-assisted journeys rarely show up cleanly in analytics.
The dashboard that confuses everyone
Here is a pattern I've seen across edtech and B2B clients through 2025 and 2026: GA4 organic sessions down 20-35% year over year, Search Console clicks down by a similar amount, and demo requests or counselling-call bookings flat or up. The leadership question is always "so is SEO working or not?"
The honest answer is: the old metric (sessions) is measuring a smaller and smaller part of what organic search does. The job hasn't disappeared. Part of it has moved into the answer layer.
Why this happens
Three forces usually combine.
1. Informational clicks are the ones disappearing. Ahrefs' updated study of 300,000 keywords measured a 58% lower CTR for position-one pages on informational keywords with AI Overviews. Those "what is X" and "how does Y work" visits were always your lowest-converting traffic. Losing them hurts the traffic chart a lot and the lead chart a little.
2. People research inside AI, then arrive decided. A buyer asks ChatGPT, Gemini or Perplexity for options, sees your name, then searches your brand or types your URL. That shows up as branded organic or direct, and those visitors convert far better than a cold blog reader.
3. The remaining clicks are more deliberate. When someone clicks through past an AI Overview, they usually want something the summary couldn't give: pricing, a tool, a comparison, a sign-up. That lifts conversion rate on what's left.
First, rule out the fake version
Before you tell your CEO that "quality is up", eliminate the boring explanations:
| Check | What to look at | Red flag |
|---|---|---|
| Channel grouping change | GA4 Admin > channel groups; any custom rules edited | Leads moved from Direct/Referral into Organic |
| Form or tracking change | Tag Manager version history, key event definitions | New event double-counts, or thank-you page fires twice |
| Spam leads | Lead list in CRM, email domains, phone patterns | Surge of junk submissions |
| Paid/brand overlap | Brand campaigns paused or cut | Brand clicks moved from paid to organic |
| Offer change | New lead magnet, free trial, pricing change | Conversion rate up because the offer changed, not the traffic |
The paid-brand one catches a lot of teams. If someone paused brand search ads, branded clicks flow to organic listings and "organic leads" jump overnight. Nothing about SEO changed.
Split organic into four buckets
Totals hide everything. Build these four views:
- Branded organic. Search Console queries matching a brand regex (include misspellings). In GA4, you can't see queries, so use landing pages like the homepage, pricing and login as a rough proxy.
- Non-branded informational. Blog and guide pages. This is where AI Overview loss concentrates.
- Non-branded commercial. Category, comparison, "best X for Y", pricing-adjacent pages.
- AI referral. Sessions from chatgpt.com, perplexity.ai, gemini.google.com, copilot.microsoft.com and similar, captured with a GA4 custom channel group using a source regex. Keep the list updated; new assistants appear often.
In most "traffic down, leads up" accounts I look at, bucket 2 falls hard, bucket 1 grows, bucket 3 is flat, and bucket 4 is small but converts well. That combination is the healthy version of the pattern.
The metrics that should replace "sessions" on your dashboard
- Leads per 1,000 organic sessions. A cleaner efficiency metric than raw conversion rate.
- Qualified leads and pipeline from organic. Pull from CRM, not GA4. Form fills can rise while quality falls.
- Branded search impressions in Search Console, trended monthly. A proxy for demand you are creating.
- AI visibility. Share of tracked prompts where your brand is mentioned or cited, using a prompt panel or a tool.
- Generative AI impressions. Search Console's generative AI performance report (global rollout reported 31 August 2026) shows how often links to your pages appear in AI Overviews and AI Mode. It's impressions only, but trend it.
- Self-reported attribution. An open "How did you hear about us?" field on forms. Tag answers mentioning ChatGPT, Gemini, "AI" or "Google AI".
Read lead quality, not just lead count
Rising leads are only good news if they close. Pull these from the CRM for organic leads, comparing both periods:
- Sales-accepted rate
- Meeting or demo show rate
- Opportunity creation rate
- Average deal size or course fee
- Time to close
If quality held or improved, you have the healthy pattern. If leads rose but acceptance fell, you may have made forms easier, added a low-intent lead magnet, or attracted spam. At Masai School, where I worked on organic social growth, the lesson that transferred to search was the same: volume numbers look great in a deck, but the team downstream decides whether the growth was real.
How to explain it to leadership
Keep it to three sentences, with numbers:
"Organic sessions are down 28% because AI Overviews now answer many of our informational queries. Organic leads are up 12% because branded and commercial search, which convert much better, grew. We're shifting reporting from sessions to qualified pipeline and AI visibility, which is where search value now shows up."
Then show the four-bucket chart and the pipeline table. Don't lead with the traffic line; it anchors everyone on the wrong number.
What to do with the informational content
Don't delete it because traffic fell. Informational pages are what AI systems read when deciding which brands to mention. Instead:
- Put a short, direct answer near the top of each page, so it's easy to cite.
- Add original examples, data or frameworks an AI summary would have to credit.
- Link each guide to a relevant commercial page, tool or template, so the fewer visitors you get have a clear next step.
- Merge thin posts that compete with each other.
When the pattern is a warning sign
"Traffic down, leads up" can be the early stage of a slow decline. If your branded search is flat, you're not being mentioned in AI answers for your category, and leads are rising only because of a new offer, the lead line will follow traffic down within a few quarters. The pattern is healthy only when demand indicators (branded search, AI mentions, direct traffic) are growing too.
A simple monthly review routine
- Search Console: branded vs non-branded clicks and impressions.
- Search Console: generative AI report, top exposed pages.
- GA4: four-bucket sessions and key events.
- CRM: organic leads, acceptance rate, pipeline.
- Prompt panel: AI mention and citation rate for 30-50 tracked prompts.
- Form answers: count of AI-related "how did you hear" replies.
Twenty minutes a month, and you'll never be surprised by the traffic line again.
FAQ
Why is my organic traffic down but conversions up?
Usually because low-intent informational visits are being answered by AI Overviews and assistants, while higher-intent visitors still click through. Rule out tracking changes, paused brand ads and spam leads before you accept that explanation.
Is falling organic traffic bad if revenue is stable?
Not necessarily. Sessions are a weaker proxy for value in 2026. Watch pipeline, branded search demand and AI visibility; if those are stable or growing, the business impact is limited.
How do I track AI referral traffic in GA4?
Create a custom channel group with a rule matching sources like chatgpt.com, perplexity.ai, gemini.google.com and copilot.microsoft.com, placed above the default Referral channel. Review the list quarterly.
Why don't AI-influenced leads show as AI referral?
Many people read an AI answer, then search your brand or type your URL. That journey shows as branded organic or direct. Self-reported attribution on forms is the best way to capture it.
Should I stop writing informational content?
No. It feeds the answers AI systems give and builds topical authority. Make it more citable and connect it to conversion paths instead of judging it on sessions alone.
What's a good leads-per-1,000-sessions benchmark?
It varies too much by industry and offer to give one honest number. Benchmark against your own history and watch the trend after AI Overviews expanded in your market.
Could the lead increase be from paid brand campaigns being paused?
Yes, and it's common. If brand ads stop, those clicks move to your organic listing. Check paid brand spend and clicks for the same periods.
How do I show AI impact to executives?
Report qualified pipeline from organic, branded search trend, AI visibility share and generative AI impressions, with a short note explaining the traffic decline. Keep sessions as supporting context.
Let's read your dashboard together
If your numbers are telling two stories at once, I can help you work out which one is true and rebuild reporting around what matters. I bring 4+ years of marketing experience across organic growth, SEO and content for edtech and startups. Take a look at my work and reach out through the contact form at younusfardeen.in.