# Search agents

Hunt for a needle in a haystack — blogs, news, pages on their site, and the wider web — by searching and reading what turns up, not only structured databases.

_Data source — Search agents Propensify uses: find mentions and news on their site, in the press, and on the wider web — then read what turns up._

HTML: https://propensify.io/data-sources/search-agents
Markdown: https://propensify.io/data-sources/search-agents.md

## Three questions we ask of the open web

1. **What did their nav hide?** — Unlinked PDFs, media kits, API docs, gated webinars, parked or dead domains.
2. **Who talks about them off-domain?** — Founder interviews, agency portfolios, registry crumbs, competitor case studies, title reconnaissance.
3. **What just happened in public?** — News language: layoffs, rebrands, acquisitions, budget freezes, agency-of-record appointments, launches.

No result means we did not find it — not that it is false.

## Finding is cheap. Verifying is the filter.

### Find

Search for pages and articles that might mention the company — on their site, in the press, and on the open web.

### Verify

Keep results that look like the right company and the right topic; discard mismatches before treating them as evidence.

## How search evidence stacks

Three layers — state the need, search where the answer likely lives, then decide whether a snippet is enough or a page is worth reading.

1. **Need** — What we’re looking for
2. **Where we look** — Hidden · off-domain · news
3. **Read** — Snippet vs full page

- **Sample, not census:** A bounded set of results — not the whole web or every article ever published.
- **Owned pages first:** Open-web search prefers pages that belong to the company when ownership matters.
- **News is a feed:** Press hits are validated article samples ordered by recency — not a news archive.

## Where we look

Same haystack — the jobs are the criteria. Whether this pass returned anything sits below.

## Where Chips

- **Hidden pages**
- **Off-domain**
- **News**

## Open-web presence

Whether search returns owned-page evidence for the information need — titles, snippets, and links worth following.

A thin snippet can answer a simple need. It is not proof of everything on the full page.

## Open Web Cards

- **Owned pages found:** Search returns pages on the company domain that match the need — sample of results, not a complete site index.
- **Snippet answers:** Title or description may answer a thin question without opening the page.
- **Page worth reading:** A result looks relevant enough to open and read for deeper evidence.
- **Nothing found:** This search pass returned no usable matches — unknown, not proof the topic never exists online.

## News & press

A recency-ordered sample of articles that cite or cover the company — standard coverage or topic-filtered when the need is specific.

Empty news feed means nothing turned up in this sample — not “never in the news forever.”

## News Cards

- **News hits:** Validated articles appear in the feed — sample coverage, not a complete press archive.
- **Topic-filtered news:** Articles intersecting a stated topic (funding, expansion, regulation, and similar) when requested.
- **No articles in sample:** No validated articles in this pass — coverage gap, not a permanent absence verdict.

## Honesty

Search can retrieve the wrong company or the wrong topic. Filters reduce that risk; they do not eliminate uncertainty.

## Honesty Cards

- **Wrong entity risk:** Results that fail ownership or domain checks are discarded from evidence — mismatch ≠ the topic never exists.
- **Wrong topic risk:** A page can mention the company without answering the need — relevance still has to fit.

## In a verified fit

Northline Analytics

## Findings

- **Owned pricing page found**
- **Recent press hit**
- **Nothing on niche regulation topic**

## Boundaries & method

- Results are a sample — not a full web index or complete news archive.
- Snippet answers are thin; they are not the full page.
- Owned-page preference discards many third-party mentions from open-web evidence by design.
- Nothing found ≠ the topic never exists online.
- Wrong-entity and wrong-topic mismatches can still occur — honesty over false certainty.
- Search agents ≠ Website Ranking traffic floor.
- Search agents ≠ Google Ads inventory.
- Search agents ≠ Website signals page-family captures (pair those guides).
- No query-syntax or vendor infrastructure details on this page.

## Want search agents on your verified fits?

We’ll hunt the needles that matter for your ICP — on their site, in the press, and on the wider web — and show what is worth reading.

- [Get 10 vetted companies →](https://propensify.io/#lead)
- [Book a call](https://propensify.io/book-a-call)


Part of https://propensify.io/data-sources. Full set: https://propensify.io/data-sources.md.