Approve a number, not a surprise
Every data tool will tell you what one lookup costs. Few will tell you what your run costs, because that depends on things nobody has looked at yet: how many people match, how many steps each of them passes through, which steps bill on a hit and which bill for trying. The usual way to find out is to run it and read the bill.
Hyreflow can answer before it spends. Databases report how many records match a search without selling you the records. Every method carries a listed rate and says what it bills on. A run can be previewed with no call made. Put together, that is a quote: a count, a list of steps marked free or paid, and the most the whole thing could cost.
Sizing a talent pool before you pitch uses the same counts to win an argument with a client. This play uses them to decide what you are willing to buy.
What you get back
A short quote. The count after each filter, with the source of each number. The steps of the full run, each marked free or paid and what it bills on. The most the run could cost, with any usage-metered step labelled as an estimate. The pilot record exactly as it came back, and what it was charged. Then one question: approve it, shrink it, or stop.
Variations worth knowing
Work backwards from a ceiling. Give a budget and ask what fits: fewer people, or the same people with the dearest step left out.
Price a list you already hold. For a list you already have the row count is known, so the quote is the rows that can be matched at the worst case per row.
Companies, not people. Company searches report totals the same way. Size the account list before you look for anyone inside it.
Where this goes wrong
An implausibly large count. A total close to the size of the whole database means a filter was dropped without an error. Add filters singly and watch the number fall. If it does not, the search is wrong. The market is not that large.
Reading the maximum as the bill. Most lookups charge on a hit, so the real cost usually lands under the quote. Budget to the maximum anyway.
Empty results that still cost. A per-request search charges for an empty page, and a search no database can answer ends in paid web searches.
Counts are a floor. A database counts the people it holds, under the titles they chose. The number sizes the run. It is not a census.
Counting in one place and pulling from another. Each database matches titles its own way, so size the run on the source that will do the pull.
