Field notes · July 16, 2026 · 3 min read
The real cost of re-typing a resume three times
> Editorial Note (August 2026): This post was written during an earlier phase of Voco, when our exploration focused heavily on building software for recruiting teams. While the observations about friction in those workflows remain true, Voco now operates more broadly to build custom AI systems for operational teams. You can read more about why we stopped starting with software.
Follow one candidate through a typical intake and count how many times a human retypes something that was already typed once.
Where the re-entry actually happens
They fill out an intake form — that's one. A recruiter reads their resume and adds notes into the ATS — that's two, sometimes with reformatting if the resume came in as an image or a non-standard layout. If the role goes to a client with its own submission template, the candidate's information gets reformatted a third time. If a second recruiter picks up the file later, they often re-verify details that were already confirmed, because there's no single place they trust to have the current answer.
None of this is anyone being careless. It's what happens when a workflow accumulates steps one at a time, each one reasonable on its own, until the total is four re-entries of the same six facts.
The cost isn't just time
The cost isn't just time, though the time is real — a few minutes per touch, dozens of candidates a week, adds up to hours nobody budgeted for. The bigger cost is error surface. Every manual re-entry is a chance for a phone number to get transposed, a certification date to get dropped, or a candidate's stated availability to silently fall out of sync between the notes doc and the ATS. Nobody catches these until a client calls about outdated information, and by then it looks like a candidate problem instead of what it actually is — a data-entry problem wearing a candidate's name.
Why this makes good candidates hard to find later
Here's the part that doesn't show up until months later: a candidate who looked strong for a role six months ago is only useful the second time if whoever's searching for them can actually find them, and find them accurately. A record that's been retyped four times, drifting slightly each time, is harder to trust than a record that was entered once and left alone. Status that never made it back into the ATS compounds this — a candidate marked "in process" from a placement that fell through six months ago reads as unavailable to anyone searching today, so a perfectly good match gets skipped without anyone deciding to skip them.
The candidate isn't lost. The trustworthy version of their record is.
The fix is a connection problem, not a purchasing decision
The fix isn't a new ATS. Most of the ATS platforms already in use are perfectly capable of being the one source of truth — they're just not being used that way, because the intake step that should feed them directly instead feeds a form, which feeds a notes doc, which someone eventually copies over. Closing that gap is a connection problem, not a purchasing decision, which is exactly the kind of thing that fits in five days: map where the resume actually enters the process, and close the re-entry loop at the first point it's technically possible to close it.
It's also why "search harder" isn't real advice when the record itself can't be trusted. Voco skips the search-the-ATS step entirely — you bring the role and the candidates you already know (a CSV export works), and it shows you who's worth another conversation, evidence included, before anyone gets written off as unreachable. Ask Voco.