Contact Lookalikes can be used to retrieve a ranked list of people who are most similar to a given reference person. The recommendations are generated by a machine learning model that compares the reference person's profile to other contacts. The profile includes person characteristics such as title, seniority, department, as well as company characteristics such as industry and company size. These lookalike contacts can be used to identify additional stakeholders, expand buyer networks, or build targeted prospecting lists based on known high-priority contacts.

To find Contact Lookalikes, select the referencePersonId, which identifies the person whose profile you want to use as the basis for finding lookalikes. You can optionally provide a targetCompanyId to constrain the search to a specific company. If targetCompanyId is not provided, the model will search for similar contacts across all companies in the Zoominfo database.

Behind the scenes, the model uses a semantic vector representation of the reference person's profile to efficiently find similar contact profiles in the ZoomInfo database. It then applies a re-ranking algorithm to the set of similar contacts found, in order to boost relevance in the final return list.

The endpoint returns up to 100 lookalike contacts ordered from the most similar contact to the least similar contact (descending order by similarity score attributes.score). To control the number of lookalikes returned, use the page[size] parameter with any integer value between 1 and 100. Each lookalike contact contains additional metadata (meta) that describes the reference person used as the basis of the lookalike.