Mental Models for the AI EconomyAI Economics
How to Measure AI Visibility Without Lying to Yourself
What platforms like Profound measure, why one answer proves nothing, and the KPI stack that turns probabilistic mentions and citations into defensible evidence.
A dashboard says your brand has 42% AI visibility. The number looks precise. The system producing it is probabilistic. Before the percentage can guide a budget, you need to know how many prompts, engines, regions, personas, and repeated runs produced it.
What these platforms actually measure
- Mention rate: how often the brand appears in the sampled answers.
- Citation share: which domains and URLs receive attribution.
- Answer prominence: where the brand appears in list-style responses.
- Sentiment: a model’s classification of how the brand is framed.
- Prompt coverage: how well the tracked questions represent the market.
- Volatility: how those measures move across runs and dates.
One answer is an anecdote. Repeated answers become a signal.
The KPI stack I would put in front of an executive
- Coverage: tracked prompts divided by the agreed customer-question universe.
- Mention rate with sample size and an uncertainty band.
- Citation share by engine and source domain.
- Answer position for ranked responses only.
- Same-day run variance, so sampling noise is visible.
- Conversion evidence from AI referrals, where the platform exposes it.
AI visibility is a distribution, not a screenshot.
Where Profound fits
Profound’s public documentation describes daily tracking across major answer engines, citation and sentiment analysis, prompt research, and regional/persona variants. Its reference corpus and enterprise workflow make it useful for programs that have enough brand and prompt complexity to justify formal observability. Smaller teams can apply the same measurement discipline with narrower tools — but cannot skip repeated sampling. Profound documentation.
For the methodology problem, see “Don’t Measure Once”. Vendor scores are estimates from selected prompts, not census data about every possible answer.
Get the next essay the day it publishes.
A few a month, written from real builds. Unsubscribe in one click.
Performance marketing and search leader who builds — 15+ years in digital growth and product (AT&T/DirecTV, Microsoft, Employbridge, Site Impact), now running his own properties with AI. These essays are field notes from that loop: what AI actually changes about work, product, and the economics underneath. Based in South Florida.
- AI EconomicsSynthetic Customers: Test the Funnel Before AI Traffic Finds It
Browser agents can pressure-test a journey before launch. Here is where synthetic users find real friction — and where plausible simulations become expensive theater.
7 min read · interactive - AI EconomicsAgentic Checkout: Is Your Store Buyable by AI?
Being recommended is no longer the finish line. AI agents need reliable product state, policy, authority, and checkout rails to complete the transaction.
7 min read · interactive - AI EconomicsYour AI Can Finally Stop Talking and Start Doing
For two years, AI has been the world's most brilliant intern locked in a room with no laptop. Model Context Protocol just handed it a badge, a login, and actual work to do.
5 min read - AI EconomicsThe Next KPI Every AI Company Will Track: The LLM Token Expenditure Index
We've spent two years asking which AI is smartest. The better question — the one that will actually show up on executive dashboards — is which AI gets the job done using the fewest resources.
6 min read - AccessibilityYour AI Chat Box Is Probably Unusable With a Screen Reader
Streaming replies, focus jumps and silent errors lock out the users chat could help most. What breaks, the eight checks that fix it, and a five-minute test.
9 min read · interactive - Web & AI SearchReddit, Forums and the AI Recommendation Engine
Why community discussion feeds AI answers, what the Google and OpenAI Reddit deals changed, and a practitioner's playbook for earning mentions without astroturfing.
8 min read