Prompt

Find companies using a competitor tool but missing the piece you sell

Tools: ClaudeLusha

The short version: search for companies that run one technology and do not run another. The gap between the two is your pitch, already qualified. In our test, filtering for a core platform while excluding its adjacent module returned 201 companies at 201 to 1,000 employees, for 1 credit. Nothing revealed until you shortlist.

Technographic prospecting usually stops at “who uses X.” The more useful question is who uses X and is missing Y, because that gap is the reason to call. A company running a platform without the module that completes it either built something themselves, bought a point solution, or has the problem unsolved. All three are conversations.

Runs in Claude with the Lusha connector.

The example output below was pulled live via the Lusha connector on August 15, 2026. Company names are replaced with category descriptors.

The prompt

<context>
I sell [PRODUCT], which solves [PROBLEM]. Companies that already run
[COMPETING TOOL] don't need me. Companies that run [ADJACENT TOOL] but
not [COMPETING TOOL] have the problem and no solution for it.

My inputs:
- Technology they must be running: [TECH THEY HAVE]
- Technology that disqualifies them: [COMPETING TOOL]
- Company size: [EMPLOYEE RANGE]
- Geography: [REGION]
- Industry, if relevant: [INDUSTRY]
</context>

<task>
1. Before searching, resolve the exact technology names Lusha recognises
   for both my include and exclude lists. Show me what you matched.
   A technology name that does not resolve will filter nothing, silently.

2. Search for companies matching my size, geography and industry that
   run the included technology and do NOT run the excluded one.

3. Return a preview table, no fields revealed:
   Company | Employees | Industry | Location

4. Report the gap: how many companies match with the exclusion applied,
   versus how many match without it. That difference is the population
   already served by the competing tool, and it tells me how saturated
   this segment is.

5. For a shortlist I choose, reveal department headcount so I can see
   whether the team that would own my product actually exists there.

6. Then find the buying committee at those accounts.
</task>

<constraints>
- Preview only until I confirm the shortlist.
- Confirm every technology name resolved to something real before running
  the search. Report any that did not match.
- Technology detection shows what is observable, not a licence audit.
  Treat absence as "not detected" rather than "definitely not running."
- Do not reveal contact details at this stage. This play produces an
  account list, not a contact list.
</constraints>

What you'll get back

The situation: a seller whose product complements a widely deployed platform but competes with one of its modules. The question is which companies have the platform without the module.

SearchCompanies
Runs the core platform, 201-1,000 employeesFull population
Same, excluding those running the adjacent module201

A sample of what came back:

CompanyEmployeesIndustryLocation
[Security platform]201-500TechnologySan Francisco
[Workforce services]501-1,000Business ServicesHouston
[Healthcare marketplace]501-1,000HealthcareNew York
[Financial services group]501-1,000FinanceToronto
[Public sector software]501-1,000TechnologySan Francisco

Five of 201 returned. Live result via Lusha connector, August 15, 2026.

What it cost: 1 credit. Company search bills at 1 credit per batch of up to 25 records. The technology exclusion is free — you are narrowing a search rather than running two.

The step that turns an account list into a qualified one

Each company record carries fields you can reveal individually, at 1 credit each: department headcount, headcount by seniority, detected competitors, monthly website traffic, and estimated annual IT spend.

Department headcount is the one worth spending on. A company that runs the platform, lacks the module, and has forty people in the function you sell to is a different prospect from one with two. That single reveal separates a list of 201 companies into the ones with a team big enough to have the problem and the ones without.

At 1 credit per company, qualifying a 20-account shortlist this way costs 20 credits, against roughly 120 to reveal contact details across the same accounts. Qualify first.

What technology detection does and does not tell you

Detection reflects what is observable from the outside — tags, headers, public footprints. It is evidence, not a licence audit.

The practical consequence is that absence is weaker evidence than presence. A company detected as running something almost certainly runs it. A company not detected as running something may run it privately, run it in one business unit, or have bought it last month. Treat the exclusion side as a ranking aid rather than a guarantee, and confirm on the call.

If the search returns nothing

Check that your technology names resolved before assuming the segment is empty. Technology filters match against a controlled list, so a product name spelled differently, an old brand name, or a module name that does not exist in the taxonomy will filter nothing and return a result that looks legitimate. The prompt asks Claude to confirm what it matched for exactly this reason.

Why use Lusha in Claude

Technographic filtering exists in many data tools. What is usually missing is the exclusion, which is the half that qualifies. A list of everyone running a platform is a big list and an undifferentiated one. A list of everyone running the platform without the piece you sell is smaller, and every company on it has a specific reason to take the call.

Running it in Claude adds the step that normally requires a second tool: resolving the technology names before searching, so a typo does not silently return a plausible list built on a filter that never applied. That failure mode is quiet and expensive, because the output looks exactly like a successful search.

The reveal-by-field structure is what keeps this affordable at territory scale. You build the account list for one credit, qualify a shortlist on department headcount for one credit each, and only then spend on contact details for the accounts that survived both steps.

Data drawn from 290M+ verified contacts and 26M+ companies, sourced and processed under GDPR, CCPA, SOC 2 Type II, ISO 27001, ISO 27701, and TRUSTe Responsible AI. Full detail in the Trust Center and privacy notice.

FAQ

How do I find companies using a competitor’s product?

Filter by detected technology. The more useful version adds an exclusion: companies running one tool and not running another, which isolates the gap you sell into rather than returning everyone in the category. Resolve the exact technology names first, since a name that does not match the taxonomy filters nothing and returns a list that looks fine.

How many credits does a technographic search cost?

1 credit per batch of up to 25 companies. The exclusion costs nothing extra because it narrows a single search rather than running two. Revealing per-company fields such as department headcount or estimated IT spend costs 1 credit each, and contact details are separate again at 1 credit per email and 5 per phone.

Does technology detection mean they definitely use it?

Presence is strong evidence; absence is weaker. Detection reflects what is observable externally, so a company shown as running something almost certainly does. A company not shown as running it may still be running it privately or in one business unit. Use the exclusion to rank rather than to disqualify outright.

How do I know whether a company is big enough to need my product?

Reveal department headcount on the shortlist. Company size alone is a poor proxy — a 900-person business with four people in the function you sell to is a worse prospect than a 300-person one with thirty. At 1 credit per company this is the cheapest qualification step available, and it runs before any contact reveals.

Why did my technology search return nothing?

Almost always an unresolved technology name rather than an empty segment. Product names change, modules get rebranded, and the filter matches a controlled list. Confirm what each name resolved to before concluding the market is not there.

What should I do with the account list once I have it?

Qualify, then map, then reveal, in that order. Reveal department headcount to confirm the team exists, build the buying committee at the accounts that pass, and only then spend credits on contact details. Reversing the order means paying for contacts at companies you would have disqualified for a credit.

 

Ready to run this?

One data connection. Works in Claude, ChatGPT, your CRM, or any agent you build.