Branded and unbranded prompts, never mixed
Proofsource sorts every tracked question into one of two kinds: brand questions that name you, and buyer questions that do not. It scores them apart and warns you when too many branded prompts are lifting your AI visibility. Marketers get an honest number for the questions that win new buyers in up to 10 AI engines, from ChatGPT and Perplexity to Gemini, Copilot and Grok.
| Kind | Questions | Share of set |
|---|---|---|
| Buyer questions | 18 | 72% |
| Brand questions | 7 | 28% |
Branded vs unbranded prompts, in short
The facts, ready to quote.
- Two kinds
- Brand questions name you; buyer questions do not
- One rule
- The same rule decides the kind for every question, everywhere
- Never mixed
- Brand and buyer questions are scored separately
- Inflation warning
- Shows your branded share and warns when it lifts visibility
- Per engine
- Brand-question results shown for each engine your plan covers
- Why it matters
- AI names a brand far more often when the question already names it
Branded vs unbranded prompts, explained
A branded prompt names the brand: "Is Driffle good?" or "Driffle vs Otter". An unbranded prompt, also called a buyer question or non-branded prompt, describes a need without naming anyone: "best AI meeting notes app for sales teams". The difference matters because AI engines almost always mention a brand when the question names it. If branded prompts sit in the same pool as buyer questions, your visibility score rises without any change in how often AI recommends you to new buyers.
Branded prompts are still useful. They show what AI says about you to people who already know your name, which affects trust and conversion. Unbranded prompts show whether AI puts you on the shortlist for people who do not know you yet. A good AI visibility tool measures both, but never adds them together.
Proofsource applies one rule to every question: does it name the brand? Brand questions and buyer questions are then kept apart in every count, chart and report. A prompt mix check shows what share of your tracked set is branded and warns you when that share is inflating your visibility.
How it works
Keeps questions that name your brand apart from buyer questions, and warns when branded prompts inflate your AI visibility.
Read each question
Every tracked question is checked for your brand name.
Label it brand or buyer
One rule labels it, so the same question gets the same label on every screen.
Score each set apart
Visibility for buyer questions and brand questions is measured separately.
Check the mix
If branded prompts make up a large share, you get a warning that they inflate visibility.
See it in the product
Real screenshots from the Proofsource app show a live brand; drawn screens carry sample data for an example brand.
| Question | Kind | Driffle named |
|---|---|---|
| best ai meeting notes app | Buyer | 2 of 4 engines |
| otter alternatives for sales teams | Buyer | 1 of 4 engines |
| is driffle good | Brand | 4 of 4 engines |
| driffle vs fireflies | Brand | 4 of 4 engines |
| meeting recorder for zoom | Buyer | 0 of 4 engines |
Brand questions are scored on their own line, never added to the buyer score.
| Kind | Questions | Named |
|---|---|---|
| Buyer questions | 18 | 31% |
| Brand questions | 7 | 96% |
Mixed together these would read 49%, which overstates how often AI recommends Driffle to new buyers.
Every capability, live today
3 shipped capabilities behind branded vs unbranded prompts, each in the product now.
Brand question rule
One rule decides whether a question names your brand, and buyer and brand questions are never mixed in any number.
Brand question set
Each brand has its own set of brand questions, tracked and shown apart from buyer questions.
Prompt mix warning
Shows the branded and unbranded share of your tracked set and warns when branded prompts inflate visibility.
Who it is for
Report a number you can defend
Show leadership visibility on buyer questions, not a figure padded by your own brand name.
Audit a client's old prompt set
Spot when a client's previous tool mixed branded prompts in, and reset expectations.
Watch what AI says about you
Use brand questions to check how engines describe you to people who already know your name.
Why it matters
A score you can defend
Buyer-question visibility shows how often AI recommends you to new people.
Both views kept
Brand questions still tell you how AI talks about you, on their own line.
An early warning
The mix check flags a skewed set before it misleads a report.
Same label everywhere
One rule means dashboards, exports and reports agree.
Google AI OverviewsGoogle AI Mode
Branded vs unbranded prompts: Proofsource and other tools
Several tools separate branded and non-branded prompts. Here is what their own pages say.
| Capability | Proofsource | Peec AI | Otterly.AI | SE Ranking | Semrush | Ahrefs | Scrunch |
|---|---|---|---|---|---|---|---|
| Branded vs non-branded labels | YesOne rule labels every question brand or buyer | YesAuto prompt classification: branded vs non-brandedSource | YesPrompt tags: branded, non-branded, funnel stageSource | YesLabels prompts as branded or non-brandedSource | YesGenerates branded and non-branded prompts from your domainSource | RelatedBranded keywords behind mentions, with search volumeSource | Not listed |
| Scores kept apart | YesBrand and buyer questions never share a denominator | Not listed | Not listed | YesBranded and non-branded prompts kept separateSource | Not listed | Not listed | Not listed |
| Warning when branded prompts inflate visibility | YesPrompt mix warning with the branded share | Not listed | Not listed | Not listed | Not listed | Not listed | Not listed |
| Tags on every prompt | YesBrand questions are a set of their own, tracked and shown apart | YesAdd own prompts and organize with tagsSource | YesBranded, non-branded, top and bottom of funnel tags; add via APISource | Not listed | Not listed | YesTag prompts by topic, funnel stage or campaignSource | YesControl how topics, tags, models and funnel stages attach to promptsSource |
| Prompts classified automatically | YesOne rule decides if a question names your brand | YesBranded vs non-branded and intent: informational, transactional, commercialSource | Not listed | Not listed | Not listed | Not listed | Not listed |
Where each of them shines
Automatic prompt classification by branded status and by intent. Compare in full
Flexible prompt tags for branded status and funnel stage, also settable through the API. Compare in full
A clear branded and non-branded split inside SE Visible, with prompt prioritization.
Generates both branded and non-branded prompts from your domain and location. Compare in full
Branded keywords behind each mention, with search volume. Compare in full
Scrunch lets you filter prompts by funnel stage and map tags to prompts. Compare in full
Questions about branded vs unbranded prompts
What is the difference between branded and unbranded prompts?
A branded prompt names the brand, such as "Is Driffle good?". An unbranded prompt describes a need without naming anyone, such as "best AI meeting notes app". Unbranded prompts show whether AI recommends you to new buyers.
Why should branded prompts be kept out of my visibility score?
AI engines almost always mention a brand when the question names it. Mixing those answers in makes your score look higher than the share of buyer answers that actually recommend you.
Does Proofsource still track branded prompts?
Yes. Brand questions are tracked and shown on their own, per engine, so you can see how AI describes you to people who already know your name.
How does Proofsource decide if a question is branded?
One rule checks whether the question names your brand. The same rule is used on every screen, report and export, so a question never changes kind between views.
What does the prompt mix warning do?
It shows the branded and unbranded share of your tracked set and warns you when branded prompts are large enough to inflate your visibility.
Does it work with every AI engine my plan covers?
Yes. The split applies to every engine your plan covers, and both sets are shown per engine.
How is this different from prompt tags in other tools?
Tools such as Peec AI, Otterly.AI and SE Ranking label branded prompts too, which is useful. Proofsource also keeps the two kinds out of each other's scores in every number and adds a warning when the mix is skewed.
Keep exploring
Try it on your own brand.
Proofsource is live: scan your brand and see branded vs unbranded prompts at work with your own answers, competitors and sources.
Google AI OverviewsGoogle AI Mode