The Look-alike Notebook brings native, first-party look-alike modeling directly into BlueConic — no data science team, no third-party platform required.

Finding more customers like your best customers is one of the highest-impact moves a marketer can make — but it has traditionally required data science expertise or reliance on ad platform audiences built on cookie data. The Look-alike Notebook changes that by bringing this capability natively into BlueConic, trained entirely on a brand's own first-party data.

A marketer selects a target segment (the audience to find more of), a source segment (the broader population), and the profile properties to include. The notebook trains the model, publishes it to BlueConic, and writes a look-alike score from 0 to 100 to each profile. That score behaves like any other profile property — it can be used in segments, dialogues, connections, and lifecycles immediately. Built-in charts help marketers understand why certain profiles match and choose the right audience size for their campaigns.

Because the model is powered by BlueConic's continuous models support, look-alike scores stay current automatically — recalculated whenever a customer visits the website or an underlying profile property changes. Unlike ad platform look-alikes, there are no third-party data dependencies and no cookie reliance. The model is built on data the brand actually owns.

Benefits

• Expands high-value audiences without third-party data: start with your best customers and find more profiles like them across your full population.
• Enables publishers to monetize first-party data for ad sales — precisely targeted look-alike audiences command higher CPMs than broad demographic segments.
• Improves campaign targeting for ad spend, email, and on-site personalization toward profiles most likely to convert.
• Scores update continuously as profile data changes, so look-alike audiences always reflect the latest customer signals.

This notebook can be installed via this url.