Quick answer: To find lookalike customers, import your existing customer or account CSV into Origami. The AI identifies common patterns across your best accounts — industry, size, geography, tech stack, buying signals — and finds hundreds of similar companies with direct contact information. One customer imported 12 accounts and got 847 lookalike prospects back in 3 minutes.

Why Lookalike Prospecting Works

Your best customers are the best signal you have. They already use your product. They already see value. Finding 500 companies that look just like them is almost always more valuable than sending mass cold email to a generic filtered list.

The traditional way to do this is painful: export your CRM, analyze the company attributes manually, figure out what patterns your best accounts share, and then build Apollo or ZoomInfo filters that approximate those patterns. It takes hours and the output is imprecise.

The Origami approach: paste your customer list, describe what you want, and let the AI find the patterns and the matching companies.

How the CSV Import Works in Origami

Origami doesn't have a formal "upload CSV" button the way a data warehouse tool does. Instead, you paste your list directly into the prompt — or describe your ideal customer based on what you already know about your best accounts.

Here's the workflow:

Option 1: Paste a company list

Tell Origami: "Here are my 12 best customers: [paste company names]. Find me 200 companies that look like these — same industry, similar size, in the US, with a similar tech stack."

Origami analyzes the companies, identifies the common thread (say: mid-market SaaS companies with 50–200 employees, using Salesforce, in the US), and returns 200 matches with direct contact information.

Option 2: Describe the pattern

If you already know what your best accounts look like, describe it: "My best customers are regional logistics companies with 20–100 trucks, operating in the Midwest, that are actively hiring dispatchers. Find me 150 companies that match this profile with owner or ops manager contacts."

Both approaches work. The CSV approach is faster if you have a list; the description approach is better if you already have a sharp ICP hypothesis.

What Lookalike Matching Looks At

When Origami analyzes your accounts and builds a lookalike list, it looks at multiple dimensions:

Industry and sub-vertical. A list of landscaping companies is not the same as a list of commercial landscaping companies. Origami picks up on the specificity.

Company size signals. Employee count, revenue estimates, number of locations, Google review volume — all proxies for company scale.

Geographic patterns. If 9 of your 12 best customers are in the Sun Belt, Origami can match on that.

Tech stack signals. If your customers tend to use Shopify, that's a signal. Companies actively using Shopify are a better ICP than "e-commerce companies generally."

Hiring signals. Companies posting jobs in specific roles (say, "Director of Operations") often signal specific growth stages that align with your ICP.

Web presence. A customer who has a strong company website and active social pages looks different from one that only has a Google Business Profile.

A Real Example

A sales rep at a field service software company came to us with 8 existing customers — all commercial cleaning companies in the Southeast US with 10–40 employees. They wanted to scale outreach.

They pasted their 8 company names into Origami with the prompt: "Find me 200 companies similar to these — commercial cleaning companies, 10–50 employees, Southeast US, with a website and active Google Business Profile."

Result: 213 commercial cleaning companies with owner or operations manager contacts. Of those, 156 had direct emails. The rep built an outreach sequence that same afternoon.

That's the whole workflow. For more on how to approach cleaning companies specifically, see finding cleaning company owners by city.

Enriching an Existing CSV

If you have a CSV of company names you've already collected — from a conference, a purchased list, or a manual research project — and you want to add owner contacts, Origami handles that too.

Paste the list into the prompt: "Here are 50 companies: [list]. Find the owner name, email, and phone for each one."

Origami enriches each entry against web sources, linking the data back to its source so you can verify accuracy.

See our full guide on enriching a company list with contacts for the step-by-step process.

Lookalike Prospecting vs List Buying

	Lookalike via Origami	Bought List
Based on your actual customers	✅ Yes	❌ No
Fresh data	✅ Live web sources	❌ Often 6-18 months stale
Customizable ICP definition	✅ Natural language	❌ Fixed filters
Direct owner contacts	✅ For local/SMB	⚠️ Varies
Price	$29–$129/mo	$500–$5,000 per list
Ownership of the data	✅ You own the export	❌ License, not ownership

Bought lists are not inherently bad — they work for certain ICPs. But for most SMB and local business prospecting, a lookalike approach based on your real customers consistently outperforms purchased lists in conversion rate.

According to research from Gartner, buyers are 3x more likely to respond to outreach when the message demonstrates understanding of their specific situation. Lookalike prospecting is how you build that specificity into your list before you even write the first email.

Building Your First Lookalike List

Pick your 5–15 best customers — the ones who churned the least, paid the most, or saw the most value
Go to useorigami.com
Paste the company names and say "find me 200 companies that look like these"
Review the results — filter out any obvious mismatches
Export and launch the sequence right inside Origami

The whole process takes under 15 minutes. You start with 1,000 free credits, which is enough for a meaningful first list.

For more on using signals to qualify the lookalikes you find, see our guide on what is signal-based prospecting.

How to Identify the Right "Seed" Accounts for Lookalike Matching

The quality of your lookalike list depends almost entirely on the quality of your seed accounts. Here's how to pick the right ones:

Use your highest-value, lowest-churn customers. These are the accounts that got real value from your product and stayed. They're a better signal than your highest-revenue accounts, which might have been closed on fit that wasn't really there.

Avoid outliers. If your seed list includes one giant enterprise account and 4 tiny startups, the lookalike model will be confused. Start with accounts in the same size band.

5–15 accounts is the sweet spot. Too few and there aren't enough patterns to identify. Too many and you risk over-constraining the match. If you have 3 accounts and they're not finding good lookalikes, add more seeds.

Use accounts from the same go-to-market motion. Don't mix inbound self-serve accounts with enterprise white-glove accounts in the same seed list. They look like very different companies and have very different buying triggers.

Enriching a Purchased List to Find the Real Decision Makers

Many teams buy a list of companies — from a conference, an association, or a list broker — but get company data without direct contacts. The list has company names, maybe industry and size, but no owner or decision-maker.

Origami solves this in one step. Paste the company names and say: "Find the CEO, VP of Operations, or head of [relevant function] for each of these companies." Origami returns direct contacts with emails and titles.

This is different from the lookalike use case but equally common. You already have the companies — you just need to get to the right person.

See our full guide on how to enrich a company list with contacts for the detailed workflow.

Lookalike Matching for Different Sales Motions

The lookalike approach looks slightly different depending on your sales motion:

PLG (product-led growth): Seed with your highest-activated free users, not paid accounts. Find companies that look like those users — they're your most likely trial-to-paid converters.

SMB sales: Seed with your stickiest small accounts. Find similar companies in adjacent geographies or industries you haven't penetrated yet.

Enterprise: Seed with your current enterprise logos. Build a target account list of companies that look like them. Enriching with buying committee contacts (CTO, VP of Eng, Head of Security) is usually the right move here.

Local businesses: Seed with your best local customers. Find similar businesses in other cities. This is where Origami particularly shines — traditional lookalike tools don't cover local businesses well.

Testing Your Lookalike List Quality

Before running outreach, do a quick quality check on your lookalike list:

Randomly sample 10 companies from the list
Look each one up manually — does it actually look like your best customers?
If yes, the list is good. If 3+ feel wrong, add context to your Origami prompt: "Exclude enterprise companies" or "Only include US-based businesses"

Most lists need one round of refinement. The second pass is usually much cleaner.

Rotate Your Device

How to Find Lookalike Customers by Importing a CSV (2026 Guide)