Install base from people · ask the people who run the tool

How to find companies that use Salesforce

The short answer

To find companies that use Salesforce, ask the people who run it. Search the job titles and job descriptions that name Salesforce, in current and past jobs, group the profiles by company, then keep a company only when one person's own words prove it used Salesforce internally. In a November 2026 run, past jobs alone revealed 41% of the US companies found.

Updated
Read
14 min
Video
2:40
On this page12 sections
  1. Why is a list of companies that use Salesforce hard to trust?
  2. How do people reveal which CRM their company uses?
  3. What do the Job Description and Past Experiences filters add?
  4. Why do past jobs find companies a current search misses?
  5. How do you prove a company really uses Salesforce?
  6. How many companies use Salesforce in the US?
  7. How do you find companies that use Salesforce, step by step?
  8. What can you do with a list of companies that use Salesforce?
  9. Does the same method work for HubSpot, Workday or any other tool?
  10. How do you search job descriptions and past jobs with the Find People API?
  11. Sources and method
  12. Video transcript

Say you want a list of companies that use Salesforce in the US. The usual way is pretty simple: open your data provider, tick Salesforce, export the list. And that’s it. Except your whole plan now rests on a checkbox you didn’t create, from a signal you probably know little about.

This guide shows another way to find companies that use Salesforce: build the signal yourself, from the people who run it. It follows the video, which presents two new filters in Find People, Job Description and Past Experiences, and the play they open. The figures come from a reference run the BlitzAPI team made on November 9, 2026.

Why is a list of companies that use Salesforce hard to trust?

Because you can’t see where each checkmark comes from. A Salesforce flag in a data tool is a signal someone else built: you don’t know where it was seen, when, or on what evidence. Maybe it’s good. Maybe it’s not. Your targeting and your pipeline now depend on a black box.

Your data providerTechnologySalesforceExport list
  • NorthwindSalesforce
  • GlobexSalesforce
  • AcmeSalesforce

Seen where?Since when?Proof?

One box ticked, one list exported. Each checkmark is a signal you didn't build, and the list doesn't say where it comes from.Watch this moment 0:20

There is a second reason, specific to a CRM. Most technographic tools read what a company shows in public: the tags and scripts of its website, the integrations it lists. That works for front-end tools. A CRM runs behind a login and leaves little public trace, so a website says little about whether a company runs Salesforce.

Definition Technographic data also called technographics
The software and tools a company works with, such as its CRM, its data warehouse or its support desk.

PreciselyTechnographic data records the technologies an organization uses. It is read from the public code of a company's websites, from integration and review data, or from job postings that name the tools a new hire will use. Each source sees a different part of the stack; job postings also show the tools a company is adopting now.

Your whole strategy rests on a checkbox you didn’t create.

How do people reveal which CRM their company uses?

They write it down. People describe what they did at work in their own job records, and Salesforce is a keyword they want the next recruiter to find. A Salesforce Administrator carries the tool in the job title. A seller writes “managed our pipeline in Salesforce”. A service agent writes “resolved cases in Service Cloud”.

That changes what you know. You are no longer guessing that a company uses Salesforce: someone is describing, in public, how they used it in their role there. Each matching profile is a witness, and a useful one:

  • First-hand. The person who writes “Salesforce Administrator” administers the org. Nobody infers the stack from the outside.
  • Specific. People name the brand and often the product: Sales Cloud, Service Cloud, CPQ, Pardot. They describe migrations too: “migrated from HubSpot to Salesforce”.
  • Dated. Every job has a start date and, once it is over, an end date. Every piece of evidence carries a time stamp.
  • Tied to a company. The job points to a LinkedIn company page, which becomes the key of your list.
  • Job title · Current jobSalesforce Administrator
  • Job description · Current jobManaged our pipeline in Salesforce.
  • Job description · Past job · until 2023Implemented Service Cloud.
Each one is a witness
  • A company page
  • A role
  • Dates
  • A quote
First-hand, specific and dated: they wrote it about their own job.
Three job records name Salesforce in the person's own words. Each one also brings a company page, a role, dates and a quote.Watch this moment 0:52

So the question changes. Stop asking which companies use Salesforce. Ask who uses Salesforce at each company, and who used it there before, including the people who left. The answer is an install base you built yourself, one witness at a time.

Definition Install base also called installed base
The companies that run a given product. The Salesforce install base is the list of companies that use Salesforce.

PreciselyAn install base, or installed base, is the set of organizations that use a given product. Only the vendor holds the full list. Everyone else infers it: from what a company exposes publicly, such as website code and integrations, or from the people who run the product and name it in their own job records, which also says when the company used it.

People tell you what they did at work.

What do the Job Description and Past Experiences filters add?

Three new places to look. Find People already searched the job titles of current jobs. It now also searches inside the job description, with job_description, and across past experiences, with include_past_jobs. A search for Salesforce can therefore run on four layers:

  1. Current job titles: the company staffs a role for Salesforce today.
  2. Current job descriptions: someone at the company works in Salesforce today.
  3. Past job titles: the company ran Salesforce until that job ended.
  4. Past job descriptions: someone worked in it there, until the end date.

Each layer is collected on its own, so each one can be judged on its own. In the reference run, titles were the cleanest signals and descriptions carried the reach: past descriptions alone reached 46,849 companies.

FieldCurrent jobinclude_active_jobsPast jobinclude_past_jobs
Job titlejob_titleJob title · Current jobExisting“Salesforce Administrator”20,587 profile matches8,231 companies reachedJob title · Past jobNew“Salesforce Developer” until 202448,139 profile matches15,045 companies reached
Job descriptionjob_descriptionJob description · Current jobNew“Managed our pipeline in Salesforce”38,704 profile matches15,982 companies reachedJob description · Past jobNew“Implemented Service Cloud” until 2023216,290 profile matches46,849 companies reached
Two fields, two time windows, four searches. The two new filters open three of the four.Watch this moment 1:11

One rule makes it work: the filters hold in the same job. When you search past jobs, the company filters (industry, size, type) describe the former employer. That is what turns a search for people into a search for companies.

Why do past jobs find companies a current search misses?

Because your current team is one snapshot in time, and former employees keep adding evidence. A company can have nobody who mentions Salesforce today, and three people who used it there two years ago.

Northwind
  • Past job · 2019 to 2023Salesforce Administrator
  • Past job · 2020 to 2024Sales Operations Manager“Built our forecast in Salesforce”
  • Past job · 2021 to 2025Account Executive“Managed my pipeline in Salesforce”
Current team: nobody mentions Salesforce
Northwind's team today says nothing about Salesforce. Three people who left say they ran it, until 2023, 2024 and 2025.Watch this moment 1:21

Witnesses pile up. A 30-person company may have one operations person who writes about the CRM, or none, while its former employees accumulate year after year, each describing the same org. In the reference run, past-job matches outnumbered current-job matches more than four to one: 264,429 against 59,291.

The result: 11,537 of the 27,893 validated companies were visible only through people who had left. A search of current jobs would have reached 16,356 of them at most. Switch past jobs off below to see where the gap sits.

Try it · what do past jobs add?

  1. 11–505,584
  2. 51–2007,612
  3. 201–5004,469
  4. 501–1,0002,805
  5. 1,001–5,0004,129
  6. 5,001–10,0001,105
  7. 10,001+2,189
Also seen in current jobsSeen only through people who leftEmployees, LinkedIn size band

Current and past jobs

27,893companies validated. 11,537 of them were visible only through past jobs.

At 11–50 employees, past jobs carry 57% of the companies found; at 10,001+, 20%. A small team has few current witnesses, and its former employees keep adding up.

Current jobs only

16,356companies at most. The other 11,537 never show up.

At most, because a company seen in a current job is not always proven by it. Turn past jobs back on to see what the search was missing.

Measured on the reference run of the Salesforce TAM playbook, November 9, 2026: people with a US location, companies of 11 employees or more, 27,893 companies validated as Salesforce users. A past job counts when it ended in 2022 or later.

The smaller the company, the larger the gap. At 10,001 employees and more, enough current staff write about Salesforce that past jobs add one company in five. At 11 to 50 employees, they add more than half. The same filter also finds the people who left a given company, as the page on former employees shows.

Past is not forever, though. A CRM can be replaced, so a past witness only counts when the job is recent enough, as the next part explains.

How do you prove a company really uses Salesforce?

Collect every profile, group them by company, then judge each company on its strongest witness, with a quote. Searching broadly finds the witnesses; judging each one strictly keeps the list honest.

Group by company. The key is the LinkedIn company page of the matched job. In the reference run, 323,720 profile matches collapsed into 55,132 companies. Profiles without a company page, and freelance contracts, drop out here: there is no company to prove.

Keep recent witnesses. A witness proves use at the time of the job, so its best possible confidence depends on when that job ended:

The best confidence a witness can reach, by when its job ended. A company validates above 0.90.
Matched jobAdjustmentBest confidenceCan validate
Current jobNone0.97Yes
Ended in 2022 or laterNone0.97Yes
Ended 2018 to 2021−0.100.87No
Ended before 2018−0.250.72No

A witness whose job ended before 2022 can never validate, so it is filtered out before any AI reads it. In the reference run, 37,422 companies had a witness recent enough, and 17,710 were seen only in older jobs. Those stay out of the list and are kept apart as former users: useful for win-back, never as proof of current use.

  1. 323,720profile matchesFour Salesforce searches, every page
  2. 55,132companiesGrouped by LinkedIn company page
  3. 37,422with a recent witnessA current job, or one that ended in 2022 or later
  4. 27,893companies validatedOne quote that proves internal use

17,710 companies were seen only in jobs that ended before 2022. They are kept apart, as a list of former users, not counted as Salesforce companies.

From profile matches to companies, then to companies proven by a quote. The older witnesses are kept apart, not thrown away.Watch this moment 1:37

Ask the right question. Each remaining witness goes through a lightweight AI validation layer. The question is not “does this person mention Salesforce?” It is “does this experience prove Salesforce was used internally at this company?”

Not the questionDoes this person mention Salesforce?

The questionDoes this experience prove Salesforce was used internally at this company?

  • A consultant implementing Salesforce for clientsThe work is for clients, not for the employerNo
  • A product that integrates with SalesforceIntegration is not internal useNo
  • Someone who managed their company's pipeline in SalesforceHands-on use, at this employerMuch stronger
The judge reads each sentence for what it proves about this employer, not for the word Salesforce.Watch this moment 1:45

A consultant implementing Salesforce for clients: no. A product that integrates with Salesforce: no. Someone who managed their company’s pipeline in Salesforce: that’s much stronger. The same reading sets aside “built a sales force of 40 reps”, a skills list that only names the tool, and Salesforce as a customer or an event sponsor. Of the 36,474 witnesses graded in the reference run, 32,054 were end users; the others were set aside one by one.

Keep one quote per company. Every verdict needs a real quote from the profile, 200 characters at most, and code checks it against the source text. A quote that can’t be found caps the confidence below the threshold, so the company can’t validate on it. Witnesses are graded strongest first, three per company at most, and grading stops at the first proof.

Yelpyelp.com · 1001–5000 employeesUses Salesforce
CRM
Salesforce · sales
Confidence
0.97
Proof
Salesforce Administrator, current job
Depth
1,351 profiles across 6 searches
Excerpt, verbatim
“Salesforce Administrator”
Found word for word in the profile
One row per company: the CRM, a confidence, the job that proves it and its exact words, checked against the profile.Watch this moment 2:06
Every verdict needs a real quote from the profile.

How many companies use Salesforce in the US?

Salesforce doesn’t publish a list, so every count of companies using Salesforce measures something different. Here is what the reference run measured on November 9, 2026, for people with a US location and companies of 11 employees or more:

  • 323,720 profile matches that name Salesforce in a job title or description. One person can match more than one search.
  • 27,893 companies validated as Salesforce users, each with one quote that proves internal use: 15,547 from a current job, 12,346 from a job that ended in 2022 or later.
  • 95 of 97 sampled verdicts held up in a blind audit by four AI graders working from the same written rubric, and the brand was right in 95 of 95. Because the graders are models too, the playbook reads the true rate as 95 to 99 in 100.

The list includes 3M, ABB, ABM Industries, A.P. Moller-Maersk and Yelp, each proven by someone in a current Salesforce role, and Abercrombie & Fitch, Accor and Airbus Helicopters, each proven by a Salesforce role that ended in 2022 or later.

Read it as a floor, not a census. A company where no current or recent employee with a US location writes Salesforce on their profile stays invisible, whatever it runs. And a company enters the list as soon as one witness in the US proves it, so global groups with US staff are in it: headquarters are not filtered.

How do you find companies that use Salesforce, step by step?

Six steps, from the search to the people you will contact:

  1. Search the four layers. Run Find People with Salesforce in the job title, then in the job description, on current jobs and on past jobs, with your geography. Keep the one word that names the product, and leave its double meaning (“a sales force”) to the judge.
  2. Set guardrails you have measured. Consultancies, staffing firms and solo profiles write Salesforce for work done for clients. Exclude a population only when a sample of it holds no end users: in the reference run, IT services firms held none for Salesforce, while software companies were kept because they include end users.
  3. Collect every page. A search stops at 50,000 results, so run each one per company size band, and split a band by company type when it still passes the ceiling.
  4. Group by company page and gate by recency. Key every profile by its LinkedIn company page, keep current jobs and jobs that ended in 2022 or later, and set the older witnesses apart.
  5. Judge the strongest witnesses, with a quote. Grade at most three per company, strongest first, stop at the first proof, and keep the excerpt with the verdict.
  6. Find the people to contact. The witness is evidence, not a lead: in 44% of the validated companies, that person has left. Start from the company and look for its current Salesforce administrators and RevOps people, as the decision makers guide explains.

What can you do with a list of companies that use Salesforce?

Sell into it, sell around it, or watch it move. Each company in the list carries more than its name:

  • Depth of use. The number of witnesses and the searches that reached each company stay on the record. A company with 284 witnesses and a Salesforce Product Owner runs a different org from a company with one person who logged calls in it. Service and marketing usage, at 2,674 and 1,290 companies, mark a multi-cloud estate.
  • Other CRMs and migrations. The judge records the other CRMs named alongside Salesforce, and the moves between them. HubSpot appears next to Salesforce at 1,154 validated companies; 375 describe a migration into Salesforce and 216 a migration out of it. For a vendor in the Salesforce ecosystem, those are timing signals.
  • A list that moves. People change jobs every day. Re-run the search every month and compare: a new current witness outside the list is a new install, a new Salesforce-dedicated title at a known account is an org that grows, and current witnesses gone while past ones remain is a possible migration.
  • Intent next to the install base. People describe what a company runs; job postings describe what it plans. A company hiring a Salesforce Administrator while absent from your list is implementing. The TAM guide shows how job postings answer that second question, on the same company key.

Does the same method work for HubSpot, Workday or any other tool?

Yes, for any tool that people name in their jobs: a CRM, an ERP, HR software, a data warehouse, a ticketing system. Nothing in the method is specific to Salesforce except the search term and the guardrails. The reference run applied it to six CRMs at once and validated 42,172 companies with one pipeline.

Two things change with the tool. The term, first: Salesforce names one product only, while a value like “Dynamics 365 Sales” matches its three words anywhere in a description and drowns in noise. Then the guardrails: excluding IT services firms removed no end users for Salesforce, but 3 of the 8 sampled for HubSpot. Measure them again for each brand.

Two new filters. A different way to build your own market signals.
Check yourselfA real Find People search of past job descriptions, run on November 12, 2026, returned these three people first. Each description names Salesforce. Which one proves the employer ran Salesforce?
Show the answer key
  1. An account manager at an IT staffing firm, until 2023, whose description lists “SalesForce BA” among the consultants they placed with clients.Work done for clients, not for the employer: an intermediary. The firm wasn’t filed under Staffing and Recruiting, so the guardrail let it through, and the judge sets it aside.
  2. An executive mentor at a leadership platform, until 2018, who “mentored executives from various industries/companies [i.e. Visa, Qualcomm, Tata, Salesforce +]”.Salesforce is a client here, not a tool. And the job ended in 2018: the recency gate sets it aside before any AI reads it.
  3. A director of enablement at a software company, until 2023, who “adopted and developed training for Salesforce CRM”.Hands-on use, at this employer, in a job that ended in 2022 or later. One quote like this is enough to validate the company.

For builders

Everything above runs on one endpoint, Find People. This part is for the people who will wire it up: the request, the four searches, a real call, and the record you keep for each company.

How do you search job descriptions and past jobs with the Find People API?

POST/v2/search/peopleAPI reference

Company and person filters in, one row per matching person out, each with the job that matched, current or past.

Send

people.job_title
include and exclude lists for the title of the matched job. Square brackets match a whole title exactly: [Salesforce Administrator].
people.job_description
New. Phrases searched in the description of the same job, every word in any order, in the first 10,000 characters. Up to 50 phrases to include and 20 to exclude.
people.include_active_jobs, people.include_past_jobs
New: past jobs. Both true searches both, one row per person, a current job preferred. include_active_jobs false with include_past_jobs true searches past jobs only.
people.location
Where the person lives, not where the company is.
company.industry, company.type, company.employee_range
Filters on the company of the matched job: in a past-job search, the former employer.
max_results, cursor
Up to 50 people per page. Send back the cursor until it is null; a search stops at 50,000 results.

Read

total_results
How many people match. Check it against the 50,000 ceiling before you page.
results[].experiences[]
The matched job only: company_linkedin_url, company_domain, job_title, job_description, job_start_date, job_end_date, job_is_current, job_contract_type.
cursor
The next page, or null at the end.

Each person returned is one record; plans and rate limits are on the BlitzAPI pricing page.

The four searches

Same body, two switches. Each search runs once per company size band, from 11–50 to 10,001+:

  • Current titles: people.job_title.include: ["Salesforce"], with the default job flags.
  • Current descriptions: people.job_description.include: ["Salesforce"], with the default job flags.
  • Past titles: job_title, with include_active_jobs: false and include_past_jobs: true.
  • Past descriptions: job_description, with the same two flags.

Why not TAM By People? It returns distinct companies with a count of matched people, which answers how many companies employ a persona today. This method needs two things it doesn’t return: past jobs, and the text an AI can read to confirm internal use.

One real call

On November 12, 2026, we ran the past-descriptions search for companies of 51 to 200 employees, with the guardrails of the reference run:

curl -X POST https://api.blitz-api.ai/v2/search/people \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d @past-descriptions.json
{
  "people": {
    "location": { "country_code": ["US"] },
    "job_description": { "include": ["Salesforce"] },
    "job_title": { "exclude": ["Freelance", "Freelancer", "Agency", "Ex", "Former", "Instructor", "Professor", "Partner", "Recruiter", "Recruitment", "Recruiting", "Talent", "Dean", "Coaching"] },
    "include_active_jobs": false,
    "include_past_jobs": true
  },
  "company": {
    "name": { "exclude": ["Salesforce", "HubSpot", "Attio", "Zoho", "monday.com", "Pipedrive", "Microsoft", "Upwork", "Fiverr", "Toptal"] },
    "industry": { "exclude": ["IT Services and IT Consulting", "Business Consulting and Services", "Staffing and Recruiting", "Advertising Services", "Marketing Services", "Venture Capital and Private Equity Principals"] },
    "type": { "exclude": ["Self-Employed", "Sole Proprietorship", "Self-Owned"] },
    "employee_count": { "min": 11 },
    "employee_range": ["51-200"]
  },
  "max_results": 3
}

It answered with total_results: 53005, above the 50,000 ceiling: before paging, this band has to be split by company.type, one search per type plus one for companies with no listed type. The first three people are the ones in the quiz above. One of the three proves Salesforce, which is why the judging step exists.

The record you keep

The judge sees only the matched job (company, title, description, dates, contract type, headline), never the search that found it, so it can’t lean on the keyword. It returns a code, an evidence tier, the brand, the CRM type, the other CRMs named, any migration, a verbatim excerpt and a one-sentence reason; the confidence comes from code, from the tier and the dates. In the reference run, two inexpensive models voted without seeing each other, and a third one, the Jev router by TypeSafe AI, graded only the witnesses they disagreed on.

One record per company, from the playbook’s appendix, with the proof profile left out:

{
  "company_name": "Yelp",
  "company_domain": "yelp.com",
  "band": "1001-5000",
  "uses_crm_confidence": 0.97,
  "crm_brand": "Salesforce",
  "crm_type": "sales",
  "brand_status": "current",
  "proof_signal": "SF_T",
  "proof_job_title": "Salesforce Administrator",
  "proof_excerpt": "Salesforce Administrator",
  "quote_check": "verified",
  "code": "E",
  "n_signals": 6,
  "signal_profiles": 1351
}

Before you ship

  • Collect the four layers as separate searches, so you can measure each one on its own.
  • Check total_results before paging. Split any search above 50,000 by size band, then by company type, and confirm none touched the ceiling.
  • Measure each guardrail on a sample of what it removes, brand by brand. Keep an exclusion only when the sample holds no end users.
  • Excluding the words around the term, such as "sales force": a multi-word exclusion drops every profile that has both words, real users included.
  • Gate past witnesses by end year before any AI call: a job that ended before 2022 can’t validate.
  • Keep the judge blind to the search that found the profile, and store its verbatim excerpt with the verdict.
  • Key every row by company_linkedin_url. Drop profiles without one, and freelance contracts.

Sources and method

Published
API calls run on

The reference run figures come from the BlitzAPI team's Salesforce TAM playbook: twenty Find People queries across six CRMs, people with a US location, companies of 11 employees or more, run on November 9, 2026, with Salesforce as the reference case. A profile match is one person matched by one query, so one person can count more than once.

The blind audit of that playbook was graded by four AI models applying the written rubric, blind to the panel's verdicts, on 200 graded witnesses, 97 of them Salesforce verdicts. Because the graders are models, the playbook reads the true rate as 95 to 99 in 100.

The validation panel used two inexpensive language models and, when they disagreed, a third model: the Jev router by TypeSafe AI. Jev is a product of TypeSafe AI, not a BlitzAPI feature.

The BlitzAPI team ran the Find People call shown in the builders section on November 12, 2026. The three people in the quiz come from it, described by their role, the kind of company and the year their job ended; their names, profiles and employers are left out.

Northwind, Globex and Acme are fictional, as in the video. The phrases in the diagrams are typical phrasings from the playbook, not individual records. The Yelp record comes from the playbook's appendix, without its proof profile.

Video transcript

Read the full transcript2:40 · 9 chapters

0:00Two new Find People filters: Job Description, Past Experiences

At BlitzAPI, we just released two new filters inside Find People. You can now search directly inside a Job Description… and across Past Experiences too. Two simple filters. But together, they open up some pretty creative GTM plays. Here’s one.

0:19How to find companies that use Salesforce: the usual way

Let’s say you want to find U.S. companies using Salesforce. The usual way is pretty simple. You open your data provider. Click Salesforce. You export the list. And that’s it.

0:31A technology checkbox you didn't create

The only problem is… your entire GTM strategy is now relying on a checkbox you didn’t create, from a signal you probably don’t know much about. Maybe it’s good. Maybe it’s not. Your entire strategy and performance linked to this black box.

0:47Build the signal from people's job records

So let’s think a little more like a GTM Engineer and build the signal ourselves. The idea is actually pretty simple. People tell you what they did at work. Someone writes: "Managed our pipeline in Salesforce." Or: "Implemented Service Cloud." That’s useful. Because now we’re not just guessing that a company uses Salesforce. We have someone publicly describing how they used it during their role.

1:10Four layers: current and past titles and descriptions

So with Find People, we search Salesforce across four layers: Current job titles. Current job descriptions. Past job titles. And past job descriptions.

1:20Past jobs: what former employees reveal

And this last part is where it gets really interesting. Because your current team is only one snapshot in time. Former employees keep adding more evidence. So a company might have nobody mentioning Salesforce today… but three people who used it there two years ago.

1:36Group by company, validate with AI

Now we collect all of those profiles. Group them by company. And send the strongest evidence through a lightweight AI validation layer. The question is not: "Does this person mention Salesforce?" It’s: "Does this experience actually prove Salesforce was used internally at this company?"

1:54Consultant, integration or real Salesforce user?

A consultant implementing Salesforce for clients? No. A product that integrates with Salesforce? No. Someone saying they managed their company’s pipeline in Salesforce? That’s much stronger. So every company gets a confidence score. Every verdict needs a real quote from the profile. And we only keep the strongest evidence.

2:14Results: US companies that use Salesforce

And here’s the part I really like. 323,000+ profiles found. 27,893 validated companies using Salesforce. 97.9% accuracy. All in under 5 minutes. So yeah… two new filters. But a completely different way to build your own market signals.

Questions this raises

Which companies use Salesforce?

Salesforce doesn't publish its customer list, but you can build one from the people who run it. In a November 2026 run, 27,893 US companies were each proven by someone describing their own Salesforce work, among them 3M, ABB and Yelp through current Salesforce roles, and Accor and Airbus Helicopters through roles that ended in 2022 or later.

How many companies use Salesforce?

Salesforce doesn't publish the number, and every list counts something different. One measure: in a November 2026 run, 27,893 companies were proven by at least one person with a US location who described running Salesforce there, in a current job or one that ended in 2022 or later. It is a floor, not a census: a company whose people never write the tool down stays invisible.

How to find out what CRM a company uses?

Look at the people who run it. Search the company's current and past employees for CRM names in their job titles and descriptions: a Salesforce Administrator, a HubSpot operations manager, someone who migrated the team from one CRM to another. Website scanners rarely see a CRM, because it runs behind a login. Job postings that name the tool are a second source.

Is this the same as technographic data?

It is one source of technographic data. Most technographic tools read what a company exposes in public, such as website tags, scripts and integrations, which shows front-end tools well and a CRM poorly. Reading the job records of the people who use a tool sees back-office software, and gives every company a date and a quote you can check.

Does a past job prove a company still uses Salesforce?

Not on its own. A past job proves the company ran Salesforce until that job ended, which is why a past witness counts only when the job ended in 2022 or later, with its year kept on the record. Replacing a CRM is rare and visible: in the six-CRM reference run, 216 validated companies described a migration away from Salesforce.

How do I find the Salesforce admins at those companies?

Start from the validated companies, not from the witness: in 44% of them, the person who proved Salesforce has since left. Find People takes up to 50 company LinkedIn pages per request, so a second search lists the current Salesforce administrators, RevOps and sales operations people of each account.

Data shouldn't be
the problem

Grab. Connect. Build.