About us
Profound Demo for B2B Marketers
How Profound measures brand visibility in AI answer engines like ChatGPT and Perplexity. Demo covers setup, features, and optimization workflows.
Your CEO is forwarding screenshots asking why your brand doesn’t show up in ChatGPT. You need answers and a plan.
Hana Hajebian, Manager of Solutions Consulting at Profound, walked us through how their platform tracks brand visibility across AI answer engines and how they help marketers improve their results.
TL;DR – Profound Demo Highlights
- Tracks brand mentions across ChatGPT, Perplexity, Gemini, Google AI Overviews
- Sends 17,000+ prompts in a week to measure visibility at scale
- Uses headless browsers (not APIs) to match real user experience
- Provides real user prompt data that answer engines don’t release
- Maps citation sources so you know which content influences AI responses
- Tests by persona and geography to find visibility gaps
- Builds optimization workflows to fix what’s broken
Setup: Minimal. Give them topics (or let them pull from your site). No integrations needed.
Who runs it: Your existing demand gen or performance team.
Which AI Tools Does Profound Track?
Why this matters:
Why this matters: Your buyers don’t only use ChatGPT. Its market share is falling (69% in January of 2025 to 45% in 2026), software engineers use GitHub Copilot, Cursor, or Claude Code. If you’re only tracking AI Overviews, you’re missing more than half the conversation.
The Answer:
ChatGPT, Gemini, Google AI Overviews, Perplexity, and other major LLMs. They add new ones as they emerge. Claude and DeepSeek weren’t on the list shown during the demo but the platform isn’t locked to a fixed set.
API or In-Browser Crawling?
Why this matters:
There are two schools of thought on whether GEO tools should optimize for what users see in the tool or for how agents use vectors to mine your content.
The Answer:
Headless browsers or what users see in the LLM tool. Not APIs and vectoring.
Profound tested both. API responses didn’t match what real users see. Headless browsers simulate someone actually using ChatGPT through their laptop, which they feel means more accurate response and citation data.
What Do You Need to Get Started?
Topics. That’s it.
Profound pulls data from the open internet. You give them topics relevant to your business, or they scrape your website and generate the list themselves.
No CRM integrations. No API keys. No waiting on IT to green light access to other systems.
How Does Prompt Research Work?
Hana’s advice: Lean on Profound.
This is new territory. Most teams don’t know how people actually prompt AI tools. Profound generates semantically relevant variations based on your topics, then validates them against real user query data.
Process:
- Profound creates prompt variations from your topics
- They backtest against real user queries
- You see which low-visibility, high-volume topics need optimization
They also provide real user prompt volume data, something AI answer engines don’t release. This shows you what people are actually asking, not what you guess they’re asking.
However, if you fancy yourself a prompt guru, you can take total control of the platform and use your own list of prompts.
Can You Target Geographies?
Why this matters:
Visibility varies by region. A brand that performs well in US queries might not show up in EMEA. This matters, especially if you’re targeting regions you don’t sit in yourself.
The Answer:
Yes. Headless browsers positioned globally.
Want to know how your brand shows up to someone in the UK vs the US? Profound sends prompts from a UK IP to simulate that exact experience. You can track regional gaps and optimize by market.
How Do Competitor Comparisons Work?
Profound tracks who else appears alongside your brand:
- Which competitors get mentioned more often
- Where you rank in the response (1st, 2nd, 3rd)
- Brand density (your mentions vs total mentions)
- Performance by topic
ChatGPT used to recommend 6-7 brands per response. Now it’s 3-4. Rank position matters more than it used to.
What Makes Profound Different?
In our opinion, the two things that stood out to us were:
1. Real user prompt volume data
Profound tracks what people actually ask in ChatGPT, Perplexity, Gemini, Claude. Answer engines don’t release this. Profound collects it anyway. Hana didn’t tell us how, but she did say it is “complex and expensive.”
2. Citation tracking and optimization workflows
Profound doesn’t just tell you if you’re mentioned. It shows you why. The platform maps which content (your site, third-party articles, social) gets cited in AI responses. Then it builds workflows to optimize underperforming content and generate new strategic briefs.
Cool Feature Alert: Persona Testing
Create buyer personas and simulate how different users experience your brand in AI responses.
Example: What does a Chief Sustainability Officer see vs what a marketer sees when they ask the same question?
Profound pre-contextualizes prompts with persona details to mirror how OpenAI and Gemini already personalize responses.
What B2B Marketers Should Know
Core strengths:
- Minimal setup: topics in, insights out
- Real user query data: see what prospects actually ask
- Geographic and persona testing: find visibility gaps by region and buyer type
- Citation transparency: know which content drives AI mentions
- Optimization workflows: not just measurement, action
Profound seems like a great fit for B2B teams that need to prove AI visibility without hiring specialists or building complex integrations.
Resources
- Profound: tryprofound.com
- Contact sales: tryprofound.com/contact-sales
Content teams: tryprofound.com/solutions/content-teams
