Most small businesses treat candidate information as temporary. The application, the interview notes, the decision: they live in email and memory for long enough to make the hire, then disappear. The problem is that every search produces data that would make the next one better, and almost none of it gets kept.
The instinct is that tracking is overhead. If you’re a small business hiring two or three people a year, a formal candidate database sounds like enterprise software you don’t need. That instinct is wrong in a specific way: the tracking doesn’t have to be formal. It has to be consistent.
Five fields, captured for every candidate, after every search, produces a record that compounds. By the fifth or sixth hire, you know which channels produce candidates who stay. You know which interview signals predicted success, and which profiles looked good in the room but failed on the job.
The Candidate Outcome Record gives you that tracking structure. It connects directly to the pipeline work in How to Build a Talent Pipeline Before You Have an Urgent Opening: the same Candidate Collection that holds the pipeline also holds the tracking record.
Why Tracking Feels Like Extra Work — and Why That’s Backward
The obvious objection: you just finished a stressful hiring process and the last thing you want to do is document it. The hire is made. Move on.
The less visible cost is what that “move on” produces over five years. A business that has hired twenty people without tracking has no idea which channels work, which interview questions matter, or why certain profiles keep failing. Each search starts from the same intuition the first one did.
The deepest problem is that tracking while you’re hiring costs almost nothing. Tracking after the fact, six months later, when you’re trying to understand why a hire didn’t work, is expensive and often incomplete. The data is most accurate closest to the event.
The five fields in the Candidate Outcome Record take three minutes per candidate. Over a ten-person search, that’s thirty minutes. That thirty minutes produces a permanent record that informs every subsequent search.
The Candidate Outcome Record
The record has five fields. Each is captured for every candidate who reaches the interview stage, and updated when the search closes.
Field 1 records the sourcing channel
Where did this candidate come from? The job board, the specific referral source, the direct approach, the prior-applicant list, the pipeline. One line.
Over time, this field tells you which channels produce candidates who reach the offer stage and which screen out at the first interview. Combined across ten searches, it tells you where to put your sourcing time.
Field 2 records how far the candidate progressed
Did they screen out at the resume stage, the first interview, the second, or the final round? One line.
Combined with Field 1, this shows where a channel’s candidates drop out. A channel can look strong at the screen stage but weak at the offer stage. The same sourcing channel can look strong at the screen stage and weak at the offer stage.
Field 3 records the key signal from the interview
One or two sentences. What was the thing that most distinguished this candidate, positively or negatively? Not a summary of the interview. The one signal that mattered most.
Over time, these signals compound into a pattern. Candidates who gave specific examples of outcomes they owned consistently advanced to offers. Candidates who described what they would do rather than what they had done consistently struggled in the role. That pattern is invisible without Field 3.
Field 4 records the final decision and the primary reason
Hired or not hired, and in one sentence, why. “Not hired: weak on building-from-scratch criterion, strong on everything else.” “Hired: only candidate who could name a specific process she’d built and what it replaced.”
This is the field you’ll want to read when evaluating a candidate in the same role two years later. It tells you what this business has decided to care about in this function.
Field 5 records the outcome at 90 days
Added after the hire, not at the time of the search. One line: working well, struggling, or departed. And if struggling or departed, what the gap turned out to be.
Field 5 closes the loop. It tells you whether the signal in Field 3 actually predicted what happened. Over enough hires, it shows which interview signals are genuinely predictive and which are noise.
How the Living Library Updates Your Candidate Record
Without tracking, each search starts with the same blank intuition. With a maintained Candidate Outcome Record in the Living Library, each search starts from what the previous ones actually showed.
The Living Library is the part of the Kiluma platform that holds your accumulated knowledge and keeps it organized over time. When you add candidate notes, interview records, and outcome observations to the Candidate Collection, the Library reads them and maintains the Candidate Outcome Record.
When the next search opens, the record is already current. It reflects the last six hires: which channels produced offer-stage candidates, which interview signals predicted success, and which overlooked concerns mattered later. Not a summary you compile. A record the Library has been maintaining.
The Conductor, Kiluma’s context-aware AI, can read the tracking record to answer specific questions about the upcoming search. Which sourcing channels have the strongest track record for this role type. What the 90-day success pattern looks like across past hires in this function. The tracking record makes those questions answerable rather than approximate.
Track the Last Three Searches Retroactively
Before you set up the system for future searches, fill in what you can from memory for the last three hires. It doesn’t have to be complete. Source, outcome, and the one thing you remember most about why you made the decision you made.
That retroactive record is an imperfect starting point and still better than nothing. The next search adds to it. The one after that adds more. By the fourth search, the record is genuinely useful.
Start imperfect. Improve from there.
The Search That Compounds
The business that starts each hire from the same intuition stays roughly as good at hiring as it was on the first search. The business that tracks gets measurably better with each one. The Candidate Outcome Record is what makes that possible. Try Kiluma free for 14 days at kiluma.ai.