How to Find a Competitor's Customers: 10 Sources That Name Names

Competitors do not publish customer lists, but they leak them continuously, because every customer relationship creates public artefacts: a case study, a review, a job ad asking for the product by name, a contract award, a conference talk. Assembled together, ten sources produce a list that is usually more useful than the competitor’s own CRM export would be, because each entry arrives with the context that tells you whether the account is worth calling. One distinction worth holding onto before you start: these sources tell you who already buys from them. If the question is which segments they are chasing next, their ad creative names it outright, which is covered in how to find competitor ad spend.

1. 1Decide which slice you actually need. Each source answers a different question: who they serve, who is unhappy, who has already left. Running all ten at once produces a long list nobody ever calls.
2. 2Harvest the named logos. Work through their customers page, case studies, testimonials and press releases, and record the date each logo first appeared. Logos are marketing assets, so treat them as claims about the best relationships rather than a customer list.
3. 3Diff the logo wall against the archive. Open an older snapshot of the same page and list every logo that has since disappeared. A removed logo is a churn signal rather than proof, and it is one of the highest-value calls you can make.
4. 4Mine the review sites. Reviewer profiles carry company name, size and industry. Sort by the lowest ratings and read the recent ones: an unhappy named customer is worth more than fifty happy anonymous ones.
5. 5Read their customers' hiring. A job ad asking for hands-on experience with the competitor's product is strong evidence the employer runs it, and it is public, current and free.
6. 6Confirm on two independent signals. A logo plus a matching job ad, or a review plus a detected script, is a customer. One weak signal on its own is a hypothesis, and calling on it wastes the first thirty seconds of the conversation.
7. 7Qualify by trigger, not by logo. Prioritise the accounts with a reason to move now: a bad review, a public contract expiring, a change of leadership, or a migration mentioned in a job ad.

Every guide on this topic tells you to look at the customer page. Almost none tell you to look at what used to be on it, which is where the commercially interesting information sits.

Open their customers page in the Internet Archive at roughly six and twelve months ago, and list every logo present then and absent now. Companies redesign pages, rotate logos and reorganise by segment, so a disappearance is not proof of churn. What it is, reliably, is a shortlist far shorter and far better qualified than their current logo wall.

Cross-check each candidate against one of three things before you call:

• A recent review from that company, particularly a low-rated one.
• A job posting at that company describing a migration, a replatform, or a search for a new tool in the same category.
• The competitor’s script or tag no longer appearing on their public site, where the product is browser-visible.

1. Their logo wall and case studies

Start with the customers page, then case studies, testimonials, press releases and any “trusted by” strip on the homepage. Case studies are the richest of these: they name the company, the industry, the size, the problem and often the specific modules in use. Record the publication date, because a five-year-old case study says very little about the relationship today.

2. Archived versions of those pages

Covered in full above. This is the highest-value hour on the page.

3. Review sites

G2, Capterra and TrustRadius show reviewer company name, company size and industry on a large share of reviews. Filter by the lowest ratings and the most recent dates: one identified, unhappy customer is worth more to a sales team than fifty anonymous happy ones. Read what specifically failed, because that sentence is the opening line of your outreach.

4. Job postings naming their product

Search job boards for the competitor’s product name. A company advertising for someone with hands-on experience in a tool is almost certainly running it, and this yields far more names than a logo wall does because it captures ordinary customers rather than reference accounts. It is also current by definition: the role is open now.

5. LinkedIn profiles

People list the tools they use as skills on their professional profiles. Log in to your own LinkedIn account and search for the competitor’s product name, then look at where those people work. Follow the competitor’s company page and the profiles of their key executives as well, and review what they post: customer announcements, new office openings and role changes all surface there first. This is manual research on an account you already have, and it should stay that way.

6. Technology detection reverse lookup

BuiltWith and Wappalyzer can list public websites running a given technology. This works well when the competitor’s product leaves a browser-visible trace such as a chat widget, a tag or an embed, and not at all when it lives behind a login. Treat the output as a hypothesis list: detection lags removals, so some entries will be former customers. The same tools pointed the other way tell you what a company runs rather than who runs a product, which is covered in how to find a competitor’s tech stack.

7. Public contract awards and tenders

For any competitor selling to government, education or healthcare, award notices publish the buyer, the value and the contract term. The term is the part to record: a contract expiring in eleven months is a dated, public buying trigger, which is about as good as competitive intelligence gets.

8. Integration and partner directories

If the competitor is listed on a marketplace such as Salesforce AppExchange, the HubSpot ecosystem or a Slack or Shopify directory, the reviews and install counts there identify customers who have gone as far as connecting the product to their core systems. Those accounts carry the highest switching cost and are the hardest to displace, which is useful to know before you spend a quarter trying.

9. Events, communities and user groups

Conference agendas, webinar co-hosts, customer advisory board members and user-group speakers are all customers, and they are usually the enthusiastic ones. Community forums and subreddits are the mirror image: search the product name and read the complaint threads, where people frequently identify their employer while asking for help.

10. Your own CRM and win/loss records

Every deal you lost to this competitor is a customer of theirs whose name, budget, buying process and reason for choosing them you already possess. This is the best-qualified list in the entire exercise and it is sitting in your own systems. Run a win-back analysis against it rather than starting from their logo wall.

The two failure modes here are believing a logo and believing a detection tool. Both produce lists that look impressive and waste a sales team’s week.

1. 1Require two independent signals. A logo plus a matching job ad, or a review plus a detected script. One weak signal is a hypothesis, and opening a call on a hypothesis costs you the first thirty seconds of the conversation.
2. 2Date every entry. Ask when the evidence was published, not when you found it. A case study from 2021 and a review from last month describe very different relationships.
3. 3Separate customer from happy customer. These are different fields and only one of them predicts whether the account will take your call. Sentiment comes from reviews, forums and support complaints.
4. 4Check the scope of the relationship. A logo can represent a company-wide rollout or one team’s pilot that never expanded. Case studies usually say which; job ads and review text often imply it.

Selling to companies that currently buy from a competitor is ordinary commercial behaviour and is entirely legal. The constraints sit on how you obtain the list and how you contact people, and they are real.

• Do not use a list that was taken, not assembled. A customer list exported by a departing employee is a trade secret in most jurisdictions. Accepting one creates liability for your company as well as for them, and it is the single fastest way to turn a competitive win into litigation.
• Personal contact data is regulated even when the company is fair game. Names, work emails and phone numbers are personal data under GDPR and comparable laws, and they need a lawful basis and, in most cases, a notice. Targeting an account is not the same as being free to email everyone in it.
• Respect outbound marketing rules. CAN-SPAM in the US and the ePrivacy rules in Europe govern unsolicited commercial email, including opt-out handling and sender identification. These apply regardless of how well researched the list is.
• Never encourage a breach of contract. Approaching a company under contract with a competitor is fine. Advising them to walk away from a binding commitment, or offering to indemnify them for doing so, is a civil exposure known as inducing breach.
• Do not misrepresent yourself to confirm a customer. Calling a company’s support line pretending to be a user of the competitor’s product to check whether they are a customer is deception, and it is prohibited by the ethics code the competitive intelligence profession works to.

• The complete list. You will see the references, the vocal, the public-sector and the technically detectable. Quiet, satisfied mid-market customers are largely invisible. Proxy: the composition of what you can see, used as a sample rather than a census.
• Contract value and seat count. Almost never public outside government contracts. Proxy: employee count at the customer, multiplied by their published per-seat price, as a plausible upper bound.
• The renewal date. The single most valuable field, and the one least often available. Proxy: public contract terms where they exist, the case study publication date plus a typical term, and asking directly in a discovery call.
• Whether they are actually satisfied. Reviews are a self-selected minority, weighted to the delighted and the furious. Proxy: the trend in recent review sentiment rather than the average score, plus voice-of-customer signals from forums and support channels.

A customer list decays in two directions at once: their new wins are invisible to you for months, and their losses stay on the page long after the relationship ended. Both are fixed by watching changes rather than re-reading the list.

Manually, that means a monthly check of new reviews, new case studies and the logo diff, with each finding recorded against the account in your CRM so a rep sees it at the right moment. Prioritise the resulting list by trigger rather than by logo size. Flares runs the review and sentiment side of this continuously, which is the part that is tedious to do by hand and time-sensitive when it changes.