How much of our organic revenue is new demand versus branded demand capture?
Short answer
Set a branded regex in Search Console, join those queries to CRM revenue through the landing page, and report the two buckets separately. Then apply one adjustment most reports skip: a share of your "branded" revenue came from buyers who discovered you inside an AI answer and searched your name afterward. That's new demand your AI-search work created, sitting in the wrong column.
Why the split matters more than the total
Branded capture measures how well your brand compounds. Non-brand measures how well your content acquires. Blend them and you'll fund the wrong side, because branded revenue grows on its own momentum and makes flat acquisition look healthy.
The pattern to watch: branded revenue rising while non-brand stays flat means you're harvesting past work rather than building new demand, and it usually precedes a growth stall by two or three quarters.
See Is our non-brand visibility falling while branded traffic hides the loss?.
Set the regex once, use it everywhere
In Search Console, create a query filter using a custom regex covering your brand plus its variants:
(?i)(yourbrand|your brand|yourbrnad|your-brand|yourbrand\.com|[unique product name]|[founder name])
Include common misspellings, spaced and hyphenated versions, your domain, product names unique to you, and founder or executive names. Save it so every future report uses the same definition, because a regex that drifts makes quarter-over-quarter comparison meaningless.
The edge cases worth deciding once and documenting:
| Query type | Bucket | Why |
|---|---|---|
| "yourbrand vs competitor" | Branded | Buyer already knew your name |
| "yourbrand alternative" | Branded | Same, even though intent is to leave |
| "best CRM for agencies" | Non-brand | Buyer found you without knowing you |
| "competitor alternative" | Non-brand | Competitor created this demand, not you |
| Founder or exec name | Branded | Name only enters vocabulary after exposure |
Join queries to revenue through the landing page
GSC gives you query and landing page. GA4 gives you the session. Your form passes the landing page as a hidden field. Your CRM inherits it on the contact and then the opportunity.
If that hidden field doesn't exist, fix it before running any of this, because the whole chain breaks at that link. See How do I map organic landing pages to the customers they eventually produce?.
The AI-search adjustment that changes the answer
Here's what the regex can't see. A buyer discovers you inside ChatGPT or Perplexity, never clicks, then searches your brand a week later. Your regex catches that as branded capture, but the demand was created by AI-search exposure that got no credit.
Our own state of AI search research, covering more than 22,000 answers, found that once a brand is mentioned in an AI answer for a prompt, next-observation mention probability is 83.2% on ChatGPT, 83.3% on Perplexity, and 84.2% on Google AI Mode. AI-search presence compounds over weeks, and the branded search at the end of that journey is the visible tail of an invisible first touch. Pew Research found that when an AI summary appears, users click a source link in only about 1% of visits, so most of that first touch leaves no record at all.
The adjustment: add a "how did you hear about us" question at signup with AI surfaces as explicit options. If 20 to 30% of your branded-search customers name an AI surface as first touch, move that share out of branded capture and into new demand. In current B2B SaaS data this typically runs 10 to 25% and climbs quarterly, so re-measure rather than setting it once.
Run the analysis
1. Build and save the branded regex in Search Console.
2. Export 12 months of query, page, and click data.
3. Join to CRM revenue by landing page. Group closed-won by query bucket.
4. Apply the AI-search adjustment. Deduct the survey-attributed share from branded and add it to new demand.
5. Report three numbers. True branded capture, non-brand new demand, and AI-influenced new demand.
Automate the monthly split
Assembling this by hand each quarter takes 4 to 6 hours, and the regex tends to drift between rebuilds, which silently manufactures the trend you're trying to measure.
Running it as an agent freezes the definition. GSC Top Keywords for Site applies your saved branded pattern identically every month, and workflow-memory lets the agent compare against its own prior runs, so a change in the split can be attributed to behaviour rather than to method.
The monthly agent:
Start (schedule, 1st of month) → GSC Top Keywords for Site (branded regex, 24 months) → GA4 AI Traffic Overview and AI Landing Pages → HubSpot Get CRM Objects (closed-won last 90 days, with survey answers) → Citation Share → Get Visibility Events → workflow-memory (prior runs) → Code (bucket revenue into brand, non-brand, and AI-influenced, then check whether the regex definition changed since the last run) → Prompt LLM (report the three numbers, and flag any movement caused by a definition change rather than by performance) → Excel export → Send Email to CMO and CFO.
That definition-drift check is small and it protects the whole series. A branded regex that grows by two terms a year will shrink your non-brand number steadily, and without the check you'll spend a quarter investigating an acquisition problem that your own reporting created.
FAQ
Related answers
- How do I prove SEO is generating revenue and not just traffic?
- How do I tell branded organic conversions apart from demand created by paid campaigns?
- Is our non-brand visibility falling while branded traffic hides the loss?
- How do I know whether content is creating demand or just capturing demand that already existed?
Want the branded-vs-non-branded split on live data every month? Start a free Analyze AI trial and connect GA4, GSC, and your CRM in one afternoon.