← all playbooks

Put a dormant candidate database back to work, one segment at a time

One dormant segment re-qualified against what your desk recruits today: who still fits and why, which addresses still deliver, and a re-engagement message drafted and staged for the people worth a conversation. Nothing is sent and nothing in your CRM changes without your go-ahead.

use when
your agency owns a large candidate database that nobody opens any more, and you keep paying to source people you may already have
starts from
Your database

The prompt

Paste it into Claude Code or the Claude desktop app with Hyreflow connected. The first line tells your agent to use Hyreflow, so it reads the play, asks before it spends anything, and hands the work back to you.

paste this into Claude
Use hyreflow and work a dormant part of my <CRM_NAME> database again, so I
stop paying to source people I already own.

The segment: <SEGMENT, E.G. ADDED BEFORE A DATE, NOT CONTACTED SINCE> in
<DISCIPLINE>. Count it first. Leave out anyone marked do not contact and
anyone I placed in the last <MONTHS> months.

What my desk recruits today: <ROLES, SENIORITY, LOCATIONS, SECTORS>.
- Filter on my own fields first, that part is free.
- For the rest, find where each person works today and score them against
  what I recruit. Keep, maybe or drop, with a one-line reason.
- For the keeps, check that the email I hold still delivers. These are
  candidates: personal email and LinkedIn only, never a work address.
- Draft a short message from me for each keep, using what my record says
  about them. It is a reconnect, not a job pitch.

Run five first and tell me what the full run costs before you do the rest.
Stage it, do not send anything. Show me the changes before you write
anything to the CRM.

Replace every <PLACEHOLDER> with your own detail. Everything else can stay as written.

What you need first

  • A connected ATS or recruitment CRM
  • One dormant segment to start with, not the whole database
  • A short description of what the desk recruits today: roles, seniority, locations, sectors
  • A Hyreflow workspace with credits, and a connected sequencer if you want the drafts staged

Tools it can reach for

The agent picks per step from what your workspace has. Nothing here is required by name.

What happens when you run it

Free steps are marked free. Anything that spends credits is marked, and the agent asks before the first paid run of any size.

  1. 1

    Count the dormant segment

    free

    The agent reads your CRM and counts the segment you described: how many records, how many carry a LinkedIn URL, how many hold an email at all. It is a read of your own system, so it is free, and it tells you how far the paid steps can reach.

  2. 2

    Filter on what you already own

    free

    Your own tags, custom fields, stages and do-not-contact flags cut the segment down before anything is looked up. People you placed lately and people outside the disciplines you recruit today leave here, at no cost.

  3. 3

    Pilot five records and price the run

    credits

    Five records run end to end so you see the real output and the real cost. The agent then states the most the segment could cost and waits for your go-ahead.

  4. 4

    Find where each person works today

    credits

    An old record shows the job someone had when it was created. Each survivor is looked up through a chain of providers that runs in a set order and stops at the first one to return a dated work history, and an answer with no work history counts as a miss and is not billed.

  5. 5

    Re-qualify against what the desk recruits today

    credits

    The refreshed history, plus what your record holds about the person, is scored against the roles, seniority and locations you gave. Each person comes back as keep, maybe or drop with a one-line reason, so you can overrule it.

  6. 6

    Check the addresses and fill the gaps

    credits

    Emails you hold for the keeps are checked for deliverability, billed per address checked, and an address on an employer's domain is set aside because candidates are written to on personal email and LinkedIn only. Where no personal address exists, the personal-email chain runs in a set order and stops at the first hit.

  7. 7

    Draft the re-engagement message

    credits

    One short message per keep, written from you, that refers to how you know each other and to where they are today. It reopens the relationship and leaves the vacancy for the conversation that follows.

  8. 8

    Stage the sequence

    free

    Approved drafts are loaded into a paused campaign in your sequencer, each person carrying their own message, and someone with no email can be loaded on their LinkedIn URL alone. Starting that campaign is a separate, explicit action that you take.

What a run costs

Credits are spent per candidate the play actually works, and a lookup that finds nothing usually costs nothing. The two figures are the run where the first provider answers and the run where every lookup walks its full chain.

candidatesif the first provider answersif every lookup walks the chain
25$3.535 credits$12120 credits
100$14140 credits$47470 credits
500$70700 credits$2352350 credits
1,000$1401400 credits$4704700 credits

Free before anything is charged

  • Count the dormant segment
  • Filter on what you already own
  • Stage the sequence

What moves the number

  • The channel. This play buys personal email, because these are candidates, and a work inbox is the wrong place to approach one. A personal address costs more to find than a work one.
  • Coverage on work history and personal email. The chain stops at the first provider that answers, and only that provider bills.
  • How many candidates survive the free filters. Everything dropped before the paid steps costs nothing.
  • The scoring and drafting steps run on the metered agent, charged on what they read and write rather than per candidate, so they sit outside this table.
  • Providers you connect with your own key. Those calls bill your account, not your credits.

An estimate, not a quote, priced at the volume credit rate. Your agent sizes the run against your own workspace and tells you what it will cost before it spends anything.

You already paid for these people once

Every record in that database cost something: an advert, a sourcing licence, a consultant's afternoon. Then the consultant left, the desk changed focus, and the search box stopped returning anyone useful. So the team goes back to paid sourcing for people who are, quite often, already in there under a job title from years ago.

The rule this play follows is plain: check what you already own before you spend credits finding it again. It takes one dormant segment, works out who in it still matters to what you recruit today, and ends with a message ready to go to those people. It does not try to repair the whole database. It tries to get conversations out of one part of it.

What you get back

A table for the segment, one row per person: the title and employer your record holds, where they work today, a keep, maybe or drop call with the reason, and whether the address you hold still delivers. For the keeps, a drafted message each and a paused campaign with those drafts loaded. Everyone who was dropped is listed with the reason, so you can see the filter was fair. Suggested record updates, such as a current employer, wait for your approval.

Variations worth knowing

Start from a live role. If one vacancy is the reason you are looking, finding the candidates already in your ATS scores the database against that job, not against the desk in general.

Wait for a reason to write. If you would rather reconnect when something has happened in a person's career, getting back in touch after a promotion or a move is triggered by the change itself. This play needs no trigger: the segment is the reason.

LinkedIn only. Where no personal address can be found, the person stays on the list with their LinkedIn URL and can be staged for a LinkedIn-led cadence.

People you placed. Someone you placed years ago may be the one hiring today. List them separately: that is a client conversation, on a work address, and a different message.

Where this goes wrong

The address on file is a work address. Old records often hold the email of the employer someone had at the time. It is probably dead, and if it is alive it is the wrong channel for a candidate. Those addresses are set aside, not used.

Reading an unchanged profile as a confirmed one. Some people never update their profile. Unchanged means the profile has not moved, which is not the same as the person not having moved.

A job pitch to someone you have ignored for years. The first message reopens the relationship. Save the vacancy for the reply.

Running the whole database. Lookups are charged per person, and much of an old database is outside your market today. Pick the segment you would search this month, prove the output, then take the next.

Questions

Why not just email the whole database?

Because most of a large, old database is no longer your market, and a share of the addresses in it are dead. A mailshot to all of it bounces hard enough to damage the sending domain you rely on for live work, and the message has to be so general that it fits nobody. Re-qualifying first means fewer people get written to, and each of them gets a message that makes sense for where they are today.

What does it cost?

Counting the segment, filtering on your own fields, the review and the staging are free. Finding where someone works today is billed per profile found, and a lookup that returns no work history is not billed. Deliverability checks are billed per address checked, whatever the verdict. Scoring and drafting are metered by usage. The agent gives you a figure for your segment after the five-record pilot and before the full run.

Is it a GDPR problem to contact people from an old database?

It can be, and it depends on how the records were collected, what you told people at the time and how long you said you would keep their data. Records past your retention period should be removed by you, not revived, so leave them out of the segment. Do-not-contact flags are honoured before anything else runs. The decision about who you may lawfully write to belongs to you and whoever advises you on data protection.

How is this different from cleaning the database?

A clean-up is hygiene: current titles, duplicates paired, dead records flagged, and it stops there. This play refreshes only as much as it needs to decide who is worth a conversation, and it ends with messages staged for those people. If the segment is in a poor state, with many duplicates and few LinkedIn URLs, clean it first and the lookups here will match far better.

Can I run this every month?

Yes, once you have run it manually and trust the shape of the output. It converts naturally into a scheduled workflow that takes the next segment, re-qualifies it and leaves the list and the drafts waiting in your workspace for approval. A scheduled run never sends anything. Prove it by hand first, because a loose segment running unattended spends credits on people you would never have written to.