How to search companies by the software holding their data
Find companies holding support tickets, sales records or usage data by searching the software that produces it, minus companies you already contacted.
To find companies that hold a kind of data, search for the software that produces it. Support tickets live in a help desk such as Zendesk, Freshdesk, Gorgias or Help Scout. Sales records live in a CRM such as Salesforce or Pipedrive. Product usage events live in an analytics tool such as Amplitude or Mixpanel. In TheirStack, open a company search, set Technology to "is any of" those tools, set Keyword confidence to High and Last date found to a recent date, then exclude the companies you've already contacted with Company list "is not any of".
If you source data from other businesses to train AI models or build a data product, this is how you find the companies worth asking. On 3 November 2026 that search, run for help desks, returned 715 US direct employers with 25 to 500 employees.
What will you get, and what do you need?
You get a list of companies that run at least one of the tools you picked, with a column per tool showing how sure we are they use it, how many of their job posts mention it, and when we first and last saw it. You can export that list to CSV or Excel, or send it to a webhook.
We find technologies in job postings. When a company names Zendesk in a support role or Salesforce in a sales ops role, we record it, and the more posts mention it, the higher the confidence. The search tells you which software a company runs. How much data sits in it, and whether the company would share it, is for your outreach to find out.
You need:
- A TheirStack account. The free trial includes company credits, so you can run this search on your own market before you pay. Seeing a company's name and domain costs 1 company credit, and you can see and export it again for 90 days at no extra cost (how credits work).
- A short list of tools per kind of data. Pick tools from the same category, so each search stands for one kind of data.
- The companies you've already contacted, either as revealed companies in TheirStack or as a list of domains.
How do you run the search in the app?
-
Open a new company search. Go to Company Search and click Search Companies.
-
Choose the tools that produce your data. Click Add filter, pick Technology, keep the operator on is any of and select the tools. For support tickets: Zendesk, Freshdesk, Gorgias and Help Scout. "Is any of" matches companies running at least one of them. "Contains all of" would only return the rare company running all of them at once.
-
Keep only strong, recent evidence. Add Keyword confidence set to High, and Last date found set to the last 90 days. High confidence means we're most sure the company uses the tool, judged from how many of its job posts mention it, how often compared with similar tools, where the mention appears and how recent it is (how confidence works). A recent last date found drops companies whose last mention is old, which can mean they've switched tools since.
-
Scope it to your market. Add HQ Country, Employees and Company type set to Direct employer, so recruiting agencies hiring on behalf of others stay out.
-
Exclude the companies you've already contacted. This keeps you from paying for the same company twice, and how you do it depends on where those companies come from:
- You found them in TheirStack before. Every company you reveal goes into the automatic Companies Seen list, and it keeps them for good, even after the 90-day reveal window ends. Go to Company Lists, open the three-dot menu on Companies Seen, click Duplicate and rename the copy with today's date, for example "Seen until 2026-11-03". The copy is a snapshot, so companies you reveal in this search won't disappear from it while you work.
- They come from your CRM or another tool. Create a list (for example "Already contacted") and add the companies to it from search results with Add to list, or by domain through the MCP server's
add_companies_to_listtool. For a short list, the Website domain filter set to exclude those domains works too.
Then click Add filter, pick Company list, change the operator from is any of to is not any of and select the list. The companies in it no longer appear, so you can't reveal or export them again by mistake. The same steps are in How to exclude specific companies from a search and How to avoid downloading the same companies twice.
The full filter set for the help desk search:
Technology: is any of Zendesk, Freshdesk, Gorgias, Help Scout Keyword confidence: High Last date found: last 90 days HQ Country: United States Employees: 25 to 500 Company type: Direct employer Company list: is not any of your exclusion list -
Check the evidence before you reveal. Each tool you filtered by gets its own column with its confidence (High, Medium or Low). Hover over a cell to see how many job posts mention the tool, with a link to them, and the first and last date we saw it.
-
Reveal, then export. Click Reveal on the companies you want, or click Export to download the whole result as CSV or Excel, or push it to a webhook. Each option shows its cost in company credits before you confirm.
-
Repeat for each kind of data. Run a separate search per category (help desk, CRM, product analytics), so each export stands for one dataset. Keep the same Company list exclusion on all of them.
How do you run the same search with the API or the MCP server?
Click API in the app to get the cURL command for the search on screen, or send it yourself to POST /v1/companies/search (API reference). This is the help desk search from the steps above:
curl -X POST "https://api.theirstack.com/v1/companies/search" \
-H "Authorization: Bearer <YOUR_API_KEY>" \
-H "Content-Type: application/json" \
-d '{
"company_technology_slug_or": ["zendesk", "freshdesk", "gorgias", "help-scout"],
"tech_filters": {
"confidence_or": ["high"],
"last_date_found_gte": "2026-08-05"
},
"company_type": "direct_employer",
"company_country_code_or": ["US"],
"min_employee_count": 25,
"max_employee_count": 500,
"company_list_id_not": [12345],
"expand_technology_slugs": ["zendesk", "freshdesk", "gorgias", "help-scout"],
"limit": 100
}'Replace 12345 with the ID of your exclusion list, and 2026-08-05 with the date 90 days before today. expand_technology_slugs adds a technologies_found field to each company, with the confidence, job count and dates for each tool. Each company returned costs 3 API credits, and over the API the same company costs credits again every time a request returns it, which is one more reason to keep company_list_id_not in every call. To check the size of a search before paying for it, see counting results.
With the TheirStack MCP server connected to Claude, ChatGPT or another assistant, you can ask in plain words. The assistant uses search_companies, and find_technologies_and_topics to look up the tool slugs, at no credit cost for the lookup (MCP tools).
Using the TheirStack MCP, find US direct employers with 25 to 500 employees that use Zendesk, Freshdesk, Gorgias or Help Scout with high confidence, seen in the last 90 days, excluding my 'Already contacted' list.
What does a real search return?
We ran the help desk search on 3 November 2026: Zendesk, Freshdesk, Gorgias or Help Scout, US headquarters, 25 to 500 employees, direct employers.
| Filters | Companies |
|---|---|
| Technology only | 8,841 |
| + Last date found since 2026-08-05 | 1,746 |
| + Keyword confidence High | 715 |
The two technology filters cut the list by more than 90%, down to the companies with strong, recent evidence. The five companies we looked at each had Zendesk at High confidence, mentioned in between 87 and 217 of their job posts, with the last mention between late August and late October 2026.
The same filters for other kinds of data, on the same date:
- Sales records (Salesforce or Pipedrive): 6,940 companies.
- Product usage events (Amplitude or Mixpanel): 207 companies.
How do you narrow the results?
- Jobs mentioning keyword: set a minimum number of job posts per tool to keep companies that rely on it heavily, such as a large support team.
- First date found: keep companies that adopted the tool recently, if new data is less useful to you than an established history.
- Industry and Revenue: when your data buyer needs a vertical, such as e-commerce support tickets.
- Keyword category: filter by a whole category (for example, Helpdesk) instead of naming tools, when you don't mind which vendor holds the data.
What are the limits, and what if the results look wrong?
- No contacts. The export lists companies, not people. To reach the person who owns the data, select companies and click Find People to open them in Apollo, LinkedIn Sales Navigator, ContactOut or LinkedIn Recruiter (find people).
- Companies that don't hire, or don't name their tools, won't show up. We only see software that appears in job posts, so small or quiet companies are under-counted.
- A tool isn't one kind of data. Some products span categories: we file HubSpot under marketing automation, and Intercom under lifecycle marketing, so a match on them doesn't always mean CRM records or support tickets. Prefer tools whose category matches the data you want.
- Small categories run thin. As of 4 November 2026, Help Scout appears at 1,377 companies in our catalogue, against Zendesk's 49,032. A strict confidence filter on a niche tool can leave few companies, so loosen confidence to Medium before dropping a tool.
- A tool you need is missing. Ask us to add it with the MCP server's
request_technologytool. - Still seeing companies you contacted? Check that the exclusion list holds the same company, and not a different entity with the same domain, and that the operator reads is not any of. For domains you exported before, the Website domain filter catches what the list misses.
What does the export contain?
The file always carries the firmographics: name, website, LinkedIn, country, city, industry, employee count, founding year, revenue and description. On top of that, it carries one set of columns per technology you filtered by, with its confidence and job count. That tells you, row by row, which tool made each company match and how strong the evidence is, so you can sort by it before you write to anyone.
Two things to know before you export (what the file contains):
- Only the technologies you filtered by, or added under Columns, get columns. Add the others there first if you want them.
- Exporting a company list instead of a search gives you the firmographic columns only, without the technology ones. Export from the search.
Open the help desk search in TheirStack to see today's count, add your Company list exclusion, and export.
TheirStack: integrations with AI assistants, automations and CRMs
TheirStack works with the AI assistants, automation tools and CRMs you already use: ask Claude or ChatGPT through the MCP server, send new matches to Clay, n8n, Make or Zapier by webhook, and create leads in Salesforce or HubSpot.
AI assistants
Ask Claude, ChatGPT, Cursor or any MCP client to search jobs, companies and tech stacks through the TheirStack MCP server.
MCP server
https://api.theirstack.com/mcpAutomations
TheirStack sends every new matching company by webhook as soon as it finds it, to your automation platform or through it into your CRM and sales tools.
CRMs and sales tools
Find the companies that hold the data you need
Search companies by the help desk, CRM or analytics tool they run, and leave out the ones you already contacted.
Start free: 200+ free credits.
How is this guide?
Last updated on
How to Scrape Job Data for a List of Company Domains
Learn how to get job posting data for hundreds of company domains at once — without building scrapers. Query TheirStack's pre-indexed database of [[total_jobs]]+ jobs from [[total_job_sources]]+ sources by domain via the app or API.
How to Search Job Postings Across 328,000+ Sources
Complete guide to searching millions of job postings from 328,000+ sources. Learn how to combine filters for sales prospecting, recruiting, and market research — with full walkthroughs for each use case.

Hermes
OpenClaw
Clay
n8n
Make
Zapier
Databar
Salesforce
HubSpot
Attio
Pipedrive
Apollo
Instantly
ContactOut