Use the radar from your own tools.

Everyone on the team can pull today's demand into Claude, a script or a spreadsheet.

1. In Claude (or any AI assistant)

Add a custom connector with this URL. The team key is in Vercel, so anyone on the team's Vercel account has it: project demand_radar → Settings → General → Project ID (starts with prj_). If an environment variable TEAM_KEY is set in that project, use that instead.

https://demandradar-ivory.vercel.app/mcp?key=TEAM_KEY

Then ask things like:

• "What courses are rising in the Netherlands on the Demand Radar?"
• "Which things do people want in Germany that don't exist yet?"
• "What is coming up in the next 60 days, and what did searches do last year?"
• "Check demand for home batteries in Gouda."

Tools: demand_radar (today's radar per category and country), check_demand (one idea, optionally one city), opportunity_map (ranked niches for a theme), live_signals (today's spikes and calendar).

2. As data (no key needed)

GET /api/demand                              everything (JSON)
GET /api/demand?view=courses&cc=NL&limit=20  one category, one country
GET /api/demand?view=missing&format=text      readable text

Views: rise (the forecast), now, products, courses, books, services, apps, trips, money, unexpected, missing, coming. Countries: NL, DE, US.

3. Report campaign results

Tell the radar what really sold, so it ranks that demand higher. From Claude: the tool report_outcome. As data:

POST /api/outcome?key=TEAM_KEY
{"cc":"NL","term":"thuisbatterij","network":"google ads","clicks":120,"conversions":3,"spend":45,"revenue":600,"date":"2026-10-13"}

Results stay private: they only change the ranking, they are never shown on the site.

What the numbers mean

• Searches: Google searches in the latest month Google has released (usually one to two months back).
• Today: whether the topic is still rising, from daily data up to today.
• Per click: what advertisers pay in Google Ads.
• The radar is rebuilt every night at about 04:15 UTC and grades yesterday's picks.