> ## Documentation Index
> Fetch the complete documentation index at: https://docs.revenuebase.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Company Description SmartSearch
> Find companies by what they actually do, described in plain language, instead of by industry code.
export const RBCardGroup = ({cols = 3, children}) =>
}
{tag && {tag}}
;
};
Describe what a company **does**, in your own words. The match runs against each company's own description, so you find businesses by actual activity rather than by an industry label somebody picked from a dropdown years ago.
```json theme={null}
{ "type": "semantic", "on": "company",
"text": "sells payroll software to small businesses",
"match": "similar" }
```
Works on both surfaces — [Contact Search](/api-reference/v2/contact-search-overview) to find people at those companies, [Company Search](/api-reference/v2/company-search-overview) to get the companies themselves.
## Why not just use industry?
Industry taxonomies are too coarse for most go-to-market work. `Financial Services` contains a retail bank, a crypto exchange, a mortgage broker, and an insurance claims processor. If you sell to one of those, the label is nearly useless as a filter.
Description search separates them, because it reads what the business says about itself:
| You want | Industry gives you | Description search gives you |
| ----------------------- | ------------------------------------- | ------------------------------------------------------- |
| Crypto exchanges | All of Financial Services | Companies that describe trading digital assets |
| Cold storage operators | Logistics, or Warehousing, or neither | Companies that describe running refrigerated warehouses |
| Clinical trial software | Software Development | Companies that describe serving pharma research |
We recommend against filtering on industry columns for targeting. Their values carry both false positives and false negatives at a rate that will quietly distort your results. Describe the business instead.
## Control how close a match must be
Two ways to do it. Pick one — sending both `match` and `maxDistance` is a `422`.
### Option 1 — a named preset
Send `match` with one of four names. Each is a similarity floor calibrated for company descriptions specifically.
| `match` | What it means here | Similarity band | On target |
| --------- | ----------------------------------------------------------------- | --------------- | --------- |
| `exact` | Only companies that are precisely what you described | 60–100% | 86% |
| `similar` | Companies clearly in the space you described | 50–60% | 56% |
| `related` | Companies in and around the space you described | 45–50% | 42% |
| `broad` | Anything with a real connection. More results will require review | 40–45% | 24% |
**On target** is the measured share of matches at that stop that genuinely fit. Presets are cumulative — `broad` returns everything `related` would, plus a looser band beneath it.
### Option 2 — your own similarity threshold
Send `maxDistance` instead and put the floor wherever you like:
```text theme={null}
maxDistance = 1 − similarity
```
A 55% similarity floor is `"maxDistance": 0.45`. A raw `maxDistance` means exactly what it says and skips our calibration.
To find your own number, preview with `"includeScores": true` and read `_companySimilarity` on each row — they come back best-match first. Scroll to where results stop being useful and set your floor just above that.
### If you send neither
You get **`similar`** — a 50% similarity floor. Sending nothing is exactly the same search as sending `"match": "similar"`, down to the same rows.
**Short queries widen `exact` automatically.** A one- or two-word query ("airlines", "staffing agencies") embeds with a lower ceiling than a descriptive phrase, so `exact` would starve — 6 of 8 major airlines sat below its normal floor. On queries of two words or fewer, `exact` runs at a 55% floor instead of 60%. This applies to the preset only; a raw `maxDistance` always means exactly what it says.
Company descriptions vary enormously in length and style, which is why these floors sit far lower than the [job title](/api-reference/v2/smartsearch-job-title) ones. **Do not carry a number across from another target** — 50% here is a solid match; 50% on a job title is noise.
Accuracy falls off sharply past `similar`. At `broad`, roughly three in four results need review. Use `related` and `broad` to discover companies you hadn't thought of, then tighten before you build a list.
## Write a good description
The phrasing matters more here than on the other two targets. A phrase describing what the business makes, who it sells to, and how it operates works well. A company name, a product brand, or a single abstract noun works poorly.
**Better:**
```json theme={null}
{ "type": "semantic", "on": "company",
"text": "regional passenger airlines operating short-haul routes",
"match": "exact" }
```
**Weaker, but handled:**
```json theme={null}
{ "type": "semantic", "on": "company",
"text": "airlines", "match": "exact" }
```
Short queries are compensated for automatically. A one- or two-word category noun embeds with a lower ceiling than a descriptive phrase, so `exact` widens to keep its promise. You still get better results from the fuller phrase, so write one when you can.
## Examples
```json theme={null}
{ "type": "semantic", "on": "company",
"text": "operates cold storage and refrigerated warehousing", "match": "exact" }
```
```json theme={null}
{ "type": "semantic", "on": "company",
"text": "provides clinical trial management software to pharmaceutical companies",
"match": "similar" }
```
```json theme={null}
{ "type": "semantic", "on": "company",
"text": "develops and finances renewable energy projects", "match": "related" }
```
## Combine several descriptions
Up to three company-description nodes per filter. Join them with `OR` to cover a market that doesn't reduce to one sentence — the piece no keyword tool can do:
```json theme={null}
{ "type": "group", "op": "OR", "children": [
{ "type": "semantic", "on": "company",
"text": "third-party logistics and freight forwarding", "match": "similar" },
{ "type": "semantic", "on": "company",
"text": "customs brokerage and trade compliance services", "match": "similar" }
]}
```
See [SmartSearch](/api-reference/v2/smartsearch) for combining this with job titles and look-alikes.
## Next steps