Talent Pool

When a hiring round closes, most of the record disappears. The information about the one person who joined survives; everything else becomes a past application. Mixed into that everything else are people worth meeting again.

That happened at Pagedeck. Two candidates reached the final round for a design role and one was hired. The other wasn't short on experience overall — one specific item the role needed was missing. Six months later, when another role in the same discipline opened, nobody surfaced that person. Someone remembered the first name but not the surname, and Grace ended up digging through six-month-old email to find them.

The loss is larger on the sourcing side. Sourcing means finding candidates directly and contacting them rather than waiting for applications, and more than half the replies there are some version of "not right now." That half isn't uninterested; the timing was wrong. When that record lives only in one person's inbox, the list leaves with them.

A talent pool is the document built to keep that information in one place. It widens the possible outcomes of contacting a candidate from two — hired or discarded — to three, and the third one is where most of the contact actually ends up.

Six fields per person.

1. Identity and how to reach them. Name, discipline, contact method. Record the route that last worked, and keep a personal one alongside any work address, since a work address stops working the moment they change jobs.
2. Which role brought you into contact and how far they got. "Senior designer role, reached final interview." That one line tells the next reader where this person stood, which decides how much of the process needs repeating.
3. Why it didn't happen. This field is what separates a talent pool from a name list. Note whether the reason was on the company's side or theirs. There was one seat and two finalists; one required item was missing; the salary range didn't work — each of those implies a different condition for reaching out again.
4. When to revisit. Written as a date, not as a vague intention. Missing experience suggests a year out, allowing time to acquire it. No open seat suggests next quarter, since the constraint disappears as soon as headcount does.
5. What has to be different by then, on both sides. If the salary range was the blocker, the range has to have moved before the second conversation. Leave the condition unchanged and the second contact ends exactly like the first.
6. Consent and retention period. Keeping an applicant's personal data past the end of a hiring round requires their agreement, and the retention window gets set alongside it. A list kept without consent isn't an asset; it's exposure, and the exposure grows with every round that adds to it.

A list missing fields 3 and 5 is a collection of names, no matter how many of them there are. Whoever opens it six months later has to start by guessing why each person is in there, and entries that require guessing don't turn into outreach.

The person who ran the round writes it. A recruiter or someone in HR fills in fields 3 and 4, and whoever sat in the interview adds a line or two of reasoning. Written up the week after the round closes, memory has already faded and field 3 becomes something like "narrowly missed." Filling it in on the interview day, or when results go out, works better.

The reader is whoever opens that role next. That might be the same person a few months later, or a new hire in the seat. Write for the second case and the document actually gets used: someone with no knowledge of the previous round should be able to read an entry and make contact from it without asking anyone what happened.

There's a third reader — the judgment behind an offer, the package of terms presented to a candidate. When the list shows what salary range recent candidates for the same role declined at, the next offer has a basis.

Setting a refresh cadence is the only way this document survives. Open it once a quarter, pull anyone whose date in field 4 has passed, and either contact them or retire the entry. Without that cadence the list becomes a two-year accumulation that nobody opens, which is the state most talent pools are actually in.

"One item short, that's all. I'm adding them with a one-year revisit date."

Recording the rejection and setting when to look again, so the decision doesn't have to be reconstructed later.

"Two people declined on salary last quarter. Let's look at the band before anything else."

Using accumulated decline reasons as input to the terms rather than repeating the same offer.

"They're in the pool, but nobody knows why this person is here."

An entry with a name and no reasoning is unusable, however promising the person was at the time.

Pagedeck built the list after that trip through old email. Thirty-four people accumulated in the first year, most of them from sourcing replies rather than rejections. When a role opened last quarter, seven names were pulled and contacted, and four replied.

One of the four was the designer who had been one item short a year earlier. They had picked up that experience in the meantime, and this round they passed. What the list provided wasn't a candidate — it was time. Searching it took an hour, instead of the three weeks a search from scratch costs.

The most common reading is that a talent pool is a collection of names and contact details. The word list suggests it. But an entry holding only a name and an address can't be used six months out. Without a record of what was good about the person and what blocked it, the reader has to form a judgment from nothing, at which point consulting the list and starting a fresh search cost the same. Filling three lines per entry creates more value here than adding more entries, and a pool of twenty usable records beats one of two hundred names.

The other mistake is assuming accumulation alone makes an asset. Candidates' circumstances keep moving. Someone weighing a change a year ago has already moved; someone the salary didn't suit now sits in a higher range. An unrefreshed list is a historical record rather than information. The same holds for candidates who arrived through inbound applications, where people apply after seeing a posting, or employee referrals, where a colleague introduces someone they know. Whichever route they came in by, the circumstances recorded at the time fade with it.