Prospect Your Competitor's Customers: How to Build the Customers-of-X List
by Oscar Uribe
There is a moment in almost every demo we run. We ask what the person wants to search for, and after a pause someone says the name of a competitor. Not one of ours. One of theirs.
Then they lean in.
We have watched this happen with a property manager, an IT consultancy, a recruitment firm, a software vendor. The reaction is always some version of the same three beats: disbelief that the information exists at all, a request to test it against a name they know, and then a quiet "that's pretty accurate" — usually followed by someone pointing at the screen and saying "wait, that one is our customer today."
This post is about that list. Why it is the best-converting list most teams never build, where it is hiding in plain sight, how to build it by hand and how an agent builds it while you watch, and the four ways public sources quietly lie to you along the way.
Why this list is different from every other list
A normal prospect list answers the question who could buy this? You filter on industry, size, geography, maybe a hiring signal, and you get a few hundred companies that fit the shape of your customers. That is a perfectly good list — we have written a whole playbook on building one in the Nordics — but every company on it still has to be convinced that the category is worth paying for at all.
The customers-of-X list answers a different question: who has already bought this, from someone we know how to beat?
Three things come with every row on it, for free:
- Proof of budget. They pay for this category today. The conversation is never "why would we need something like this" — it is "why would we switch."
- A known incumbent. You know their supplier's pricing model, its weak spots, the objections its customers raise, and roughly how its contracts renew. You have handled that competitor a hundred times. Your best answers are already in your team's heads — or, if you have done this properly, in the system.
- A shared vocabulary. They already speak the category's language. Discovery starts three questions further in.
That is why reps lean in. A lookalike list is a hypothesis. A customers-of-X list is a set of companies that have already run the evaluation you are trying to get into, and chose someone else once. The second evaluation is always shorter.
Where the list is hiding
Nobody publishes a "customers of X" list. But both sides of every supplier relationship leak it constantly, because both sides have a reason to.
The supplier leaks it on purpose. Logo walls. Case studies. Press releases along the lines of "Company Y chooses X for its warehouse operations." Reference pages. Webinar and event line-ups, where the speaker's title includes the customer's name. Partner and reseller directories. Reviews on software marketplaces, where the reviewer's company shows. Every one of these exists to sell, and every one of them is a row on your list.
The customer leaks it by accident. Job ads: "experience with X is a plus" is a company telling you which system it runs. Public procurement records — in the Nordics, the tender databases are a goldmine for public-sector accounts, because the winning supplier is a matter of record. Annual reports and sustainability reports name systems and partners. Integration pages ("we connect to X"). Footers and badges ("powered by X"). Conference talks about "our journey with X."
The registries anchor it. All of the above gives you names, and names are ambiguous. "Nordic Logistics" might be three companies and a brand. The step most teams skip is matching every found name to an organisation number in the national registry, which is what lets you deduplicate, attach the financials, see the group structure, and then find the right person at the right legal entity. Without the anchor you have a list of strings. With it you have a list of companies.
Building it by hand
The honest version of the manual process looks like this.
- Pick one competitor. Open their website and go through the customers page, the case studies, the press room, and the events archive. Copy every company name into a sheet. Expect a couple of hours per competitor if they are proud of their customers.
- Search the competitor's name in quotation marks together with words like "chooses," "selects," "implements," "partners with," and the local equivalents. Search it together with "case," "reference," and "customer." Another hour.
- Search the competitor's name in job ads. Anyone requiring experience with the system is a user of it. Another hour, and this one ages fast.
- Match every name to an organisation number, kill the duplicates, drop the companies outside your market, and note where each name came from and when.
- Now find the decision-maker at each one, which is a separate afternoon we covered in the first hands-on episode.
Half a day to a day per competitor, before a single call, and the list is already ageing by the time you finish step four. Most teams do this exactly once, for their biggest competitor, in a burst of enthusiasm — and then never again, because nobody has a spare day every month.
Building it with an agent
The reason this has become a demo moment rather than a spreadsheet project is that the work in steps one to four is precisely what an agent that reads the public web is good at.
You name the competitor. The agent runs the searches — the logo walls, the press releases, the job ads, the procurement records, the references — and returns each company it finds with the sentence it found it in. "Y selected X as its platform in 2024." "Our client Z has used X since the merger." Every row carries its evidence and its source. You read the sentence, and you accept the row or you reject it.
Recently, during an onboarding, a customer named one of his competitors and watched the list go from four companies to twelve while we talked. Then he pointed at the screen: several of the twelve were his own customers today. That is not the agent being wrong. That is the agent surfacing something true — and it is the perfect segue to the part of this post that matters most.
Where public sources quietly lie
Everything above comes from the public web, and the public web is a record of what was true when someone wrote it. Four lies to watch for:
1. The customer who left. A former flagship customer stays on the competitor's case-study page for years after the contract ended, because nobody at the supplier wants to take down their best story. This is the same mechanism that makes contact lists go stale: websites preserve the past. Treat any undated reference as a hypothesis, and treat "case study, 2021" as a prompt to check whether they are still there.
2. The wrong legal entity. The case study names the brand; the buyer is the parent, or a subsidiary, or a franchisee who bought independently. The registry anchor fixes most of this, but it only fixes it if you actually look at the group structure before you call the wrong company's CFO about a system she has never heard of.
3. The relationship that is not what it looks like. Partners and resellers show up on customer pages. Pilots show up as customers. A consultancy that implements X for its clients appears to use X. Read the sentence. "Implemented X for its clients" and "runs X" are different rows.
4. Your own customers on their list. This is the one that stings, and it is the most useful. It means one of three things: the reference is stale and they switched to you (good — confirm it and move on), the account runs both suppliers in different units (a cross-sell you did not know you had), or you are about to be displaced and the competitor already has the case study drafted. Any of the three is worth a call this week.
The common rule: every row must carry its source and its date, and the list is a set of leads with a hypothesis attached — not a set of facts. An agent that returns company names without the sentence they came from is not saving you time. It is moving the verification to the phone call, where it costs the most.
What to say when you call
A quick word on tone, because this list is easy to misuse.
"I see you use X" as an opener is a gotcha, and buyers hear it as one. The list tells you where to call and roughly when. It does not give you the right to make the buyer feel surveilled.
What works is the category and the moment. You know they run something in this space; you know the incumbent's contracts tend to renew at a certain time of year; you know the two or three reasons customers of that incumbent typically look around. Open there. "We work with a number of companies that run X, and the conversation usually starts around the time the renewal comes up — I wanted to ask where you are in that cycle." That is asking rather than presenting, with a very well-informed question.
And time it. The list is ten times more valuable in the quarter before the incumbent's typical renewal, in the month after the incumbent raises prices or gets acquired, and the week a customer of theirs posts a job ad for someone to "own the migration." A colleague of one of our customers put it well: she was not planning a campaign at a competitor, she was planning a campaign at a competitor whose customers she knew were unhappy. The list is the who. The trigger is the when.
A note on the data
Everything in this post is company-level information that the companies themselves, or their suppliers, chose to publish. That is what makes it usable. The people you then call are found the way we always find them — from what the company itself publishes about who does what, not from a scraped profile of a private individual. Build the company list as aggressively as the public record allows, and build the people layer carefully.
How we do it
In Funnelfeedr this is an agent with the relationship type Customers of, pointed at any company in the registry. It reads the public web, matches every hit to the organisation number in the Nordic registries, and returns each company with the sentence it was found in. You approve the rows that hold up, the rest never reach a list, and the decision-makers for your target roles are found automatically on each approved company. From there the row is one click from a call. The whole flow lives on the Prospect seat, and if your team works in Claude rather than in a browser tab, the same search runs through our MCP.
But the tool matters less than the habit. Pick your biggest competitor. Build the list once, by hand if you have to, and note how many of the companies on it you have never spoken to. That number is your displacement pipeline, and it has been sitting in public the whole time.
Want to type a competitor's name and watch the list build itself — every company with the sentence it was found in? Book a demo and bring the competitor you'd most like to displace →