Use case
Competitive and market intelligence
Teams tracking companies, products and market movements.
The problem, and the approach
The problem
- Building a picture of a company by search means many queries and a manual merge, repeated whenever anything changes.
- The valuable finding is usually a relationship — who is connected to whom — and relationships are exactly what a ranked list does not express.
- Coverage is dominated by press releases, which look like many sources and are one.
The approach
- dossier returns mentions, top sources by authority and co-mentioned entities in one call. The co-mention list surfaces relationships you did not know to ask about.
- path finds the shortest chain linking two entities or passages, which is the connect-the-dots operation no amount of searching reconstructs.
- merged_duplicates on a corroboration check distinguishes a press release that travelled from independent reporting.
The parameters that matter
| Parameter | Value | Why |
|---|---|---|
dossier?entity= | the company | Mentions, sources and co-mentions in one call. |
path?from=&to= | two entities | The shortest documented connection between them. |
content_type[] | news, article | Excludes the company’s own product pages when you want outside coverage. |
exclude_site | the company’s own domain | Separates what others say from what they say about themselves. |
curl -H "x-api-key: $UNLOB_API_KEY" "https://api.unlob.com/dossier?entity=Nebius&limit=10"
curl -H "x-api-key: $UNLOB_API_KEY" \
"https://api.unlob.com/search?q=Nebius+acquisition&exclude_site[]=nebius.com&collapse=story"The unit of analysis is the entity, not the query
Competitive work is a poor fit for a query-shaped interface because the question is rarely about words. It is about an organisation: what has been said about it, by whom, alongside which other organisations, and how that has moved. Expressed as searches, that becomes a dozen keyword variants and a manual merge, and the merge is where the analyst's time goes.
A dossier collapses that into one call — mentions, sources ranked by authority, and the entities that co-occur — which changes what the workflow can afford. When a brief costs one call rather than an afternoon, running it across forty companies weekly becomes reasonable, and the interesting signal in competitive intelligence is almost always in the change between runs rather than in any single brief.
Co-mention is a weak signal used well
Two companies appearing in the same passage means very little on its own; it means a great deal when the pattern is new. Co-mention discovery is worth running as a difference rather than a snapshot: the pairs that were not there last month are the ones worth a human look, and they surface partnerships, competitive framing and shared investor narratives before any of it is announced.
The limits are worth being blunt about. Entity resolution struggles with common names and with recent rebrands, and a merged or split entity produces a brief that is confidently wrong rather than empty. And the open web is not the whole picture — regulatory filings, licensed databases and paywalled trade press sit outside it, and on some questions they hold the answer. Use this to decide where to look, not to conclude that nothing is there.
Where this is not the right tool
- Entity resolution is imperfect. Companies with common names or recent rebrands may resolve poorly.
- The index covers the open web only. Filings, licensed databases and paywalled trade press are outside it.
Frequently asked questions
Can I track a company over time?
Yes — combine dossier for the current picture with published_from bounds for change. Store the passage ids you have seen to detect what is new.
Does this cover financial filings?
Only where they are published on the open web and crawled. For systematic filing coverage a dedicated data provider is the right tool.
Build it on the free tier
10,000 credits a month, no card. Every parameter above works on every plan.