More Resources
Prompt

Build a target list around real engagement

Tools: Claude ▪ Lusha

To build a target list from buying signals rather than filters, you start at the other end: find companies already engaging with your category, then check whether they fit. This play runs in Claude with the Lusha connector and the LinkedIn Activity Intent signal.

The short version: ask Claude for companies showing LinkedIn Activity Intent, add a headcount band, and return the top 20 with industry and location. The search costs 1 credit per batch of up to 25 results. No reveal credits are spent until you ask for contact details. Every company that comes back is there because its own employees are publicly active on a relevant topic, not because it matched a filter thousands of other vendors are running.

How to set it up

01

Connect Lusha to Claude

Open the Lusha connector on Claude, or go to Settings → Connectors in Claude and add Lusha directly.

02

Start from the signal, not the filter

Ask Claude for companies showing LinkedIn activity intent first, then narrow by size or geography if the list needs trimming.

03

Export a list built on engagement, not a guess

Every company on the list is there because of a real, dated signal, not because it matched a generic firmographic filter everyone else is also targeting.

The prompt

Open in Claude Open in ChatGPT 

Use Lusha to find companies showing LinkedIn activity intent, 51-200 employees.
Return the top 20 with industry and location.
Do not reveal contact details yet — preview only.

 

Two words doing real work here. “Use Lusha” names the tool explicitly, so the call is deterministic if you have several data connectors enabled. “Preview only” keeps the run at search cost and stops an agent spending reveal credits on twenty companies you have not qualified yet.

What you'll get back

The situation: you need a list for next quarter and the usual approach is industry plus headcount plus geography. That list is identical to the one every competitor in your category is building this week.

The output: this filter alone matched 191,567 companies. Two from the result set:

CompanyIndustryLocationSize
it|v##ureTechnology, Information & MediaLondon, UK51–200
PHP A##ncyFinanceAddison, TX51–200

What it cost: 1 credit. Company search bills at 1 credit per batch of up to 25 results, and preview-only means no reveal credits were spent. When you shortlist and ask for contacts, email reveal is 1 credit per contact and phone reveal is 5.

If it comes back thin or empty: that is usually the filter, not the coverage. Drop the headcount band first and rerun. Signals fire on company activity, so a narrow size range in a narrow geography can legitimately return very little even where the companies exist.

Live result via the Lusha connector, August 12, 2026. Company names masked.

Why use Lusha in Claude

A list built purely on firmographics is a filtered guess, and it is the same guess your competitors are making from the same public attributes. Ask Claude to build one without Lusha connected and that is what you get, with no way to tell who is actually paying attention. With the Lusha connector on, the starting point inverts: the list begins with companies whose employees are publicly engaging on a relevant topic, refreshed weekly, and gets narrowed by fit afterwards. Smaller list, sharper list, and every account on it has a dated reason to be there. Signals describe company-level business activity, sourced and processed under Lusha’s privacy notice and compliance framework.

FAQ

  • The LinkedIn Activity Intent filter alone matched 191,567 companies in this run, before any size or geography narrowing. You are sampling from that pool, not scraping it, so the practical question is how tightly you narrow rather than whether there is enough volume.

  • 1 credit. Company search bills at 1 credit per batch of up to 25 results. Because the prompt specifies preview only, no reveal credits are spent. Revealing contact details later costs 1 credit per email and 5 credits per phone number, per contact.

  • Yes. Ask Claude to format the result as CSV, or push it directly if you have a CRM connector enabled in the same conversation. Shortlist before you enrich, so you are only spending reveal credits on accounts you have qualified.

  • Yes, deliberately. This play is for prioritising who to work now, not for defining your total addressable market. Run it alongside a firmographic list rather than instead of one: the signal list tells you where to spend this month, the firmographic list tells you where the ceiling is.

  • LinkedIn Activity Intent refreshes weekly. That is fast enough for a monthly or fortnightly prospecting cadence and it means a list you built three weeks ago is stale enough to be worth rebuilding rather than reworking.

  • Almost always the filter combination rather than the data. Filters compound, and each one can independently exclude a company that exists. Remove one at a time, starting with the narrowest, and watch the count move before concluding the coverage is not there.

Related Prompts

Ready to run this?

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

Start for free Read the Docs