Find accounts that look like your best customers

Example outputs in this play are illustrative — they reflect the structure, fields, and format of real Lusha connector output, but were not pulled from a live session. Run the prompt with your own customers and connectors to see live results. Personal details in any live examples are masked or abbreviated for privacy.

Your best customers share patterns you can’t always see manually — similar size, similar tech stack, similar signals that fired before they converted. This Claude prompt takes up to five of your best customers, runs them through Lusha’s lookalike engine, and returns the accounts most similar to them — scored by fit, verified with current contacts, and ranked by the buying signals firing right now.

The prompt

Open in Claude Open in ChatGPT 
<context>
I want to find new accounts that look like my best customers —
companies that share the same firmographic profile, signals,
and buying patterns as the customers I've already won.

My best customers:
1. [Company name or domain]
2. [Company name or domain]
3. [Company name or domain]
4. [Company name or domain — optional]
5. [Company name or domain — optional]
</context>

<task>
1. LOOKALIKE DISCOVERY — Use Lusha to find companies that
   match the profile of my best customers:
   - Same industry and sub-vertical
   - Similar employee count and revenue range
   - Similar funding stage
   - Similar geography

2. SIGNAL CHECK — For each lookalike account, check for
   buying signals firing right now:
   - Hiring surges in the target function
   - Funding events in the last 90 days
   - Executive moves in the last 30 days
   - Intent signals on relevant topics

3. CONTACT VERIFICATION — For each account, surface one
   verified VP+ contact via Lusha:
   - Current title and tenure
   - Verified email and direct dial

4. RANKED OUTPUT — Return a scored list ranked by signal
   strength and ICP fit. Highest fit and strongest signals
   at the top. Include a one-line reason why each account
   surfaced.

Return only accounts I can act on today.
</task>

What you'll get back

You give it five customers. It gives you the accounts most likely to buy from you next — scored by fit, ranked by signal strength, each one with a verified contact and a reason it surfaced.

Ranked lookalike accounts

Each account in the output looks like this:

#1 — [Company], Series B SaaS · 280 employees · Austin, TX

Why it surfaced: matches 4 of 5 customer profiles on industry, size, and funding stage. Active intent signal on your category.

FieldValue
IndustryRevenue Intelligence / SaaS
Employees260–300
Funding stageSeries B
HQAustin, TX
ICP fit scoreHigh

Buying signals firing now

SignalTypeScoreDate
Sales prospecting dataIntent82[date]
VP of Sales hired 6 weeks agoExecutive move—[date]
12 SDR roles postedHiring surge—[date]

Verified contact

FieldValue
NameR.M.
TitleVP of Sales
Tenure6 weeks ⚑ recent hire
Emailr.m@[company].com
Direct dial+1 512 555 ••••

Contact confirmed live via Lusha connector, [date]

#2 — [Company], Series A SaaS · 140 employees · New York, NY

Why it surfaced: matches 3 of 5 customer profiles. Funding event last month. Hiring surge in sales function.

The list continues ranked by fit score. Accounts with no signals are shown at the bottom — checked but not flagged as urgent. Accounts that couldn’t be matched in Lusha are listed separately so nothing silently drops.

Example outputs in this play are illustrative — they reflect the structure, fields, and format of real Lusha connector output, but were not pulled from a live session. Run the prompt with your own customers and connectors to see live results.

Built by: Lusha
Time to build: 1 min
Difficulty: Easy
Tools: Claude, Lusha
Type: Prompt

Why use Lusha in Claude

Most lookalike tools match on surface-level firmographics — same industry, similar headcount, comparable revenue. That gives you a list of companies that look alike on paper. What it doesn’t give you is the timing.

Lusha’s lookalike engine is built differently. It starts from your actual closed-won customers — the specific firmographic, technographic, and signal patterns that appeared in the accounts you’ve already won — and surfaces new accounts that match those patterns. Not a generic similarity model that applies equally to every company in your category. Your model, built from your data.

On top of that, every lookalike account is checked for live buying signals. A company that matches your best customer profile and just raised a Series B, hired a new VP of Sales, and posted 12 SDR roles in the last two weeks isn’t just a good fit on paper. Something is happening there right now. Lusha’s 1.2B+ data points processed daily and 7M new signals every week mean those patterns surface in real time — not after someone manually updates a spreadsheet.

The contact verification step closes the loop. Every account in the output comes with a verified VP+ contact — current title, direct dial, business email — so the list is ready to work the moment it lands. No separate enrichment step, no bounced emails, no calling a switchboard.

Lusha data is sourced and used in accordance with Lusha’s Privacy Policy and Terms of Use.

FAQ

  • How many customers do I need to give it?

    A minimum of three gives the lookalike engine enough signal to find meaningful patterns. Five is the sweet spot — enough data to identify what your best customers have in common without diluting the model with too many inputs. If your customers span very different industries or sizes, run the prompt separately for each segment rather than mixing them in one pass.

  • How is this different from the Find Lookalike Companies play?

    The Find lookalike companies play returns a list of similar companies based on firmographic matching. This play goes further — it layers live buying signals on top of the lookalike results, verifies a contact at each account, and ranks the output by signal strength so the accounts worth calling this week surface at the top. Same starting point, significantly more actionable output.

  • Can I use churned customers instead of best customers?

    Yes — and it’s a valuable use case. Run the prompt with your churned accounts instead of your best customers and you get a reverse lookalike: companies that match the profile of accounts that didn’t work out. Use that list to flag high-risk prospects before they enter the pipeline, or to identify patterns in why certain accounts churn so you can adjust your ICP going forward.

  • How often should I run this?

    Once a quarter is a good baseline — enough time for new companies to enter Lusha’s database and for signals to shift meaningfully. If you’re in a fast-moving market or your ICP is in a high-growth segment like early-stage SaaS, running it monthly keeps the list fresh. Save the prompt in a Claude Project with your customer list as a custom instruction and it takes 30 seconds to re-run.

  • What if one of my best customers is too large or too well-known to produce useful lookalikes?

    Large enterprise customers can skew the lookalike results toward accounts that are out of reach or out of ICP. If that’s happening, remove the outlier and replace it with a customer that’s more representative of your typical win. The lookalike engine is only as useful as the inputs you give it — the closer your seed accounts are to your actual target profile, the more actionable the results.

Related Prompts

Ready to run this?

Connect once, run anywhere. Works in Claude, ChatGPT, n8n, Clay, or any agent connected to Lusha.

Get free access Read the Docs