How do I find lookalike accounts to my best customers?

Find lookalike accounts by taking the attributes your best customers share and applying them, weighted by how strongly each separates your winners from everyone else, across the combined public, web, social, and review universe, so every net-new account that matches the profile surfaces. Lookalike prospecting is your ICP run at scale and weighted, which in practice is your Fit Score pointed at accounts you do not yet know.

This guide covers finding net-new prospects that resemble your customers. For defining the profile itself, see How do I build an ICP?

What makes a good lookalike model in healthcare?

Build it on traits that exist market-wide and can be verified, not soft signals missing for most accounts. Registries supply the structural ones (specialty taxonomy, size, ownership, setting), and they are one layer. Web and consumer sources supply the rest that exist for nearly every account too: the service lines listed on the company website, Google review volume and rating, an Instagram or LinkedIn presence, and hiring signals from job postings. If your best customers are independent multi-site specialty groups with deep service lines and strong review counts, that whole pattern can be matched across the market, not just the registry slice of it.

How is a lookalike different from a plain ICP list?

An ICP list applies your criteria once, pass or fail. A lookalike model weights the traits by how strongly they predict a win, drawing on third-party firmographics and, above all, your own win data, so the closest matches rise to the top rather than every match looking equal.

Illustrative example

Illustrative, based on how SignalRx customers work. A vendor whose best customers are 5-to-20-provider independent cardiology groups weights a mix highest: provider count and independence from the registry, plus a matching service-line list on the website and healthy Google review volume. A 12-provider independent cardiology group it has never contacted, with a full-service website and hundreds of strong reviews, scores near the top of the lookalike list; a 2-provider PE-owned practice with a thin site and few reviews scores low even though both are "cardiology."

Because lookalikes are net-new accounts with no history in your systems, SignalRx builds and runs this model on external data triangulated across sources, NPPES specialty, CMS size and ownership, the company website, reviews and social, job postings, and third-party firmographics, and writes the resulting Fit Score and tags onto both the account and its contacts as Revenue Plays and lists native in Salesforce. Where a CRM match arrives sparse, it fills the gaps by triangulating those sources so the account can still be scored rather than passed over, and inference decides which to trust when they disagree. Once a lookalike converts to a customer, SignalRx adds your own internal signals, revenue, support, and product usage tracked over time, which is the differentiator competitors cannot replicate.

Key takeaways

• Lookalikes are your ICP weighted by what separates winners from the rest, with your own win data as the weighting basis.
• Build them on market-wide traits from registries plus web, reviews, social, and hiring so they apply to the whole market.
• Rerun them as your customer base and the market change.

See it in action: Discover High-Fit Prospects with SignalRx.

Related: How do I build a Fit Score? and How do I build a target account list?