Mindcase for Company Intelligence

If you're an agent (or building one) reading this rather than a human, the full machine-readable schema for every endpoint below lives at mindcase.co/skills.md.

"Research this company" is really three separate lookups wearing one request: who they are, what's happened to them lately, and what they're currently spending money to promote. Done by hand that's three tabs and a summary you write yourself. Done by an agent with the right endpoints, it's one prompt and a brief.

What does "company intelligence" actually require pulling together?

Three things, and they come from three different places. Firmographics — headcount, industry, HQ — answer "who is this company." Recent news answers "what's changed for them lately." Ad spend answers "what are they currently pushing, and how hard." None of the three substitutes for the others: a company can look stable on paper while quietly ramping ad spend ahead of a launch, or show up in headlines with no corresponding change in headcount at all. An agent doing this well pulls all three and lets you see where they agree or diverge, rather than reporting just one.

The value is in reading them together, not any one in isolation. Steady headcount plus zero recent news plus a sudden jump in active ad count is its own kind of signal — a company testing a new channel quietly, before it's a story anyone writes. Fast headcount growth plus a wave of press plus flat ad spend reads differently — that's a company whose growth is inbound-driven, not something they're currently buying. Neither conclusion comes from a single source; both come from noticing where the three actually disagree with each other.

How do you find and confirm the right LinkedIn company page?

Start from whatever you actually have — a company name, or a list of them. LinkedIn Companies API resolves either into a confirmed company page plus the firmographic baseline: headcount, industry, HQ, founding year.

curl -X POST "https://api.mindcase.co/v1/data/linkedin/companies/run?wait=true" \
-H "Authorization: Bearer $MINDCASE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
  "params": {
    "companies": ["https://www.linkedin.com/company/anthropic"]
  }
}'

# $0.004 per company

You don't need a LinkedIn URL to start. The same endpoint takes a query plus locations and companySize filters instead, so "mid-size fintech companies in Austin" resolves to a confirmed company list before you've touched a LinkedIn URL at all. Each row comes back with 27 fields — enough to know who you're briefing on before you pull anything else.

How do you filter company news for signal instead of volume?

A company with any public profile generates a constant stream of mentions, and most of it isn't worth an agent's attention. The fix isn't pulling more news, it's pulling less of it, deliberately. Google News API takes a query plus a timeframe filter, so an agent asks for the last 7 days by default rather than the entire archive.

curl -X POST "https://api.mindcase.co/v1/data/google/news/run?wait=true" \
-H "Authorization: Bearer $MINDCASE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
  "params": {
    "query": "Anthropic",
    "timeframe": "7d",
    "maxResults": 10
  }
}'

# $0.002 per article

Asking for 10 recent articles instead of every mention ever indexed costs two cents and gets an agent exactly what a briefing needs: what happened recently, not the full history. Widen timeframe only when you actually need the backstory.

This matters more than it sounds like it should. A company with any public profile generates far more coverage than is useful in a single brief, and the instinct to "pull everything so nothing gets missed" runs directly into per-row pricing — every extra article is an extra $0.002 whether or not it tells you anything new. Filtering by a tight timeframe isn't a limitation you work around, it's the actual design: pay for the ten articles that matter, not the archive.

How do you check what a company is actually spending on ads right now?

Headcount and headlines tell you about the past. Ad spend tells you what a company is betting on right now. LinkedIn Ads API pulls every currently active ad tied to a company page, with impressions, targeting location, and how long each one has been running.

curl -X POST "https://api.mindcase.co/v1/data/linkedin/ads-library/run?wait=true" \
-H "Authorization: Bearer $MINDCASE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
  "params": {
    "urls": "https://www.linkedin.com/company/anthropic",
    "maxResults": 25
  }
}'

# $0.002 per ad

A sudden jump in active ad count, or a shift in targetingLocation toward a new region, is a real signal a news search won't surface on its own — nobody issues a press release for an ad budget increase.

What can't you get from Mindcase?

Two cases.

You want historical ad spend. The Ads API returns what's currently active plus how long each ad has run — it doesn't reconstruct a company's ad history from before you started pulling. Track it over time yourself by re-running the same query on a schedule.

Financial data — revenue, funding, valuation — isn't part of any of the three. linkedin/companies is firmographic data only (headcount, industry, HQ), not a funding database — the same real gap covered in our Crunchbase alternative post if that's specifically what you need.

Which endpoint should you use for which job?

A full three-source brief on one company — firmographics, 10 recent articles, 25 active ads — costs under 15 cents.

FAQ