Is our non-brand visibility falling while branded traffic hides the loss?
Short answer
Split organic clicks into branded and non-brand with a saved regex and chart the two series separately. Branded traffic grows from PR, paid, product, and word of mouth, so it can rise enough to conceal a substantial non-brand decline in the blended total. Non-brand is your acquisition engine, and when it falls while branded carries the chart, you're harvesting existing awareness and no longer creating new demand.
Why a flat total can hide a serious problem
Branded and non-brand are different businesses sharing a chart.
Branded clicks come from people who already know you, driven by PR, paid campaigns, product usage, and reputation. Non-brand clicks come from people discovering you for the first time.
Suppose branded grew 40% because you raised a round and got covered, while non-brand fell 25% because a competitor and an AI Overview between them took your category queries. The blended total shows a mild increase, and nothing in your reporting flags the acquisition engine failing.
That pattern is common in companies that have just had a good quarter of publicity, which is precisely when nobody is looking for a problem.
Set the split up once and keep it stable
Create a branded regex in Search Console covering your brand, common misspellings, spaced and hyphenated forms, your domain, unique product names, and founder or executive names. Save it and reuse it in every report.
The important discipline is that the regex must not drift. A pattern that quietly grows over time will shrink your non-brand series artificially and manufacture the exact decline you're looking for. If you do extend it, recompute prior periods under the new pattern and say so.
Then chart three series monthly: branded clicks, non-brand clicks, and non-brand share of total.
Read the four patterns
| Branded | Non-brand | What it means |
|---|---|---|
| Up | Up | Healthy. Awareness and acquisition both working. |
| Up | Down | The masked decline. Harvesting awareness, not creating it. |
| Down | Up | Acquisition working, brand or PR softening. |
| Down | Down | Broad problem. Check technical causes first. |
Row two is this article. The tell is non-brand share of total falling steadily while the absolute total looks fine.
Watch the share metric rather than the absolutes, because share moves before the totals do and gives you a quarter's warning.
What actually causes the non-brand half to fall
Four causes, in rough order of frequency right now.
AI answers absorbing informational queries. Non-brand informational is where the click loss concentrates. Pew Research found that when an AI summary appears, users click a source link in only about 1% of visits.
Competitive displacement on category terms.
Content mix drift toward branded-adjacent topics. Teams write more about their own product over time, which ranks for branded-ish queries and never reaches new buyers.
Category demand contraction. Check before treating it as a performance problem, using How do I tell seasonality from a real SEO decline?.
The complication that inflates your branded series
Some of your branded growth is not brand strength. It's non-brand discovery that arrived through an AI answer and returned as a branded search two weeks later.
Our own state of AI search research, covering more than 22,000 answers, found Google AI Mode returns a citation-rich answer in 97.4% of matched B2B queries and Perplexity in 93.2%, and that once a brand is mentioned for a prompt, presence tends to persist.
So a share of what your regex counts as branded is really new demand your AI-search visibility created, arriving through a branded doorway. Sizing it matters because it changes the diagnosis: if the AI-influenced portion is growing, your acquisition engine is working through a channel your non-brand series can't see. Method in How much of our organic revenue is new demand versus branded demand capture?.
Report it so the problem can't hide
Three changes to your standard reporting, and the masking stops being possible.
Never report a blended organic total as a headline. Two lines minimum.
Make non-brand share of total a tracked metric with a target. It's the earliest warning available.
Split non-brand further by intent. A non-brand decline concentrated in informational queries with commercial holding is a very different situation from the reverse, and only the second is urgent.
Build the split with the AI layer included
The branded regex half is Search Console. The part that needs a second source is whether your AI visibility is feeding the branded series.
GSC Top Keywords for Site with the saved regex gives you both series from one run. Share of Voice and Citation Share give the non-brand visibility that no longer produces clicks, which is the portion your Search Console split will read as pure loss.
A monthly agent:
Start (schedule, 1st of month) → GSC Top Keywords for Site (branded regex, 24 months) → Code (compute branded and non-brand series, non-brand share of total, and split non-brand by intent bucket) → share-of-voice recipe → Citation Share → Get Visibility Events → HubSpot Search Contacts (survey answers naming AI surfaces) → Prompt LLM (report both series, flag a falling non-brand share, and estimate what portion of branded growth is AI-influenced discovery) → Excel export → Send Email to CMO.
The line worth having in that output: non-brand clicks fell 22%, non-brand share fell from 61% to 48%, and roughly a fifth of branded growth is AI-influenced discovery. That reframes a good-looking quarter into an accurate one, which is the entire purpose of the split.
FAQ
Related answers
- How much of our organic revenue is new demand versus branded demand capture?
- How do I tell branded organic conversions apart from demand created by paid campaigns?
- How do I know whether content is creating demand or just capturing demand that already existed?
Want branded and non-brand tracked separately with the AI layer included? Start a free Analyze AI trial and connect Search Console.