sensecollect.com

SenseCollect returns public web data as flat records. Every tool runs a live collection against the source, so calls take seconds rather than milliseconds. Billing is per row returned, never per call. Credits are held against the limit you pass, then settled against the rows the run collected, so an over-large limit is not a penalty and a search that matches nothing is free. A failed run is refunded automatically. Every result reports credits_used and credits_remaining, and account_balance is free to call, so you can budget a long task as you go. Start with maps_search for local business data — it is what most accounts use. If a run outlives its request budget it returns a job_id; poll it with get_job rather than re-running the search, which would be charged again. Pass verify=true on a Maps search when the contact has to be reachable: the address is checked before the row is returned, rows without a deliverable one are dropped, and only the rows that pass are charged. Add personal_only=true to skip shared inboxes like info@. After acting on a lead, call claim_leads with its id. Claimed leads are left out of later searches and never charged for twice. For recurring work, create_watch monitors a niche and city and watch_changes reports what moved — both free, and far cheaper than re-running a search to diff it yourself.