Exa’s people search and contents endpoints handle bulk work well. We ran thousands of lookups in one batch with no throttling, and costs came in under our estimates. The contents endpoint returns a person’s work history as structured data with start and end dates, so we can check whether someone still works at a company with simple code instead of parsing text. Every result includes a citation and a crawl date, which lets us tell fresh evidence from stale evidence. That matters a lot when you clean CRM data. Errors are clear too: out-of-credit and no-data responses have their own status codes, so our job stopped cleanly and resumed from the right row. Review collected by and hosted on G2.com.

Coverage and freshness are the main limits. A noticeable share of the LinkedIn profiles we requested were not in Exa’s index, so those contacts got no answer. Cached profiles sometimes lag behind real changes. One person had rejoined a company after Exa last crawled their page, so the data showed them as gone. We also hit a few data quirks, such as a current role stored with an end date equal to its start date, and we had to handle those in our own code. Search answers can mix in a different person with the same name, so you need your own check that a result is the right person. Credits are prepaid, and our first run stopped partway when they ran out. An auto top-up or spend alert in the dashboard would help. Review collected by and hosted on G2.com.