Candidate relationship management (CRM) is the practice of building and keeping relationships with future candidates before a role is open, so that the next vacancy starts with a warm shortlist instead of a blank search. It covers who you track (applicants, silver medalists, sourced profiles, referrals), how you stay in touch, and the software that holds it together, usually called a recruiting CRM or talent CRM. It differs from an applicant tracking system (ATS), which manages people applying to a specific job, and it only works with a legal basis for storing the data.
What does candidate relationship management mean in practice?
Marketing has customer relationship management; recruiting has the same logic for talent. In practice it means three things: a pool of people you would like to hire someday, a rhythm of relevant contact, and one record per person with the full history. If a recruiter cannot answer “who spoke to this person, when, and what did we promise?” in ten seconds, there is no candidate relationship management yet, only a list.
The term also names the tool category, which causes confusion. Here is how the three things people mix up differ:
| ATS | Candidate CRM | Talent pool | |
|---|---|---|---|
| Focus | People who applied to a job | People who might apply or be approached | The stored group of people itself |
| Trigger | An open requisition | No vacancy needed | Result of both |
| Core job | Process, compliance, hiring decision | Contact history, nurturing, segments | Fast shortlists for future roles |
| Typical failure | Rejected candidates disappear | Nobody keeps the data current | Turns into a graveyard |
A longer comparison is in Talent pool, ATS or CRM: which does what?, and what a CRM should own beyond the ATS is covered in Recruiting CRM: what it should own beyond the ATS.
Why it matters: time, cost and the next hire
SHRM’s 2025 benchmarking report puts the median time-to-fill at about 45 days for executive and non-executive roles, with a median cost-per-hire of $1,200 for non-executive and $10,625 for executive roles (2025). Executive cost-per-hire has more than doubled since 2017. Every week you save at the start of a search counts, and a ready pool is the only way to save it without cutting corners.
A worked example with stated assumptions
A team makes 60 hires a year. Assumptions (yours will differ): 30 % of roles can start from the existing pool, that is 18 roles; each of them saves about 15 recruiter hours of initial sourcing; a recruiter hour costs 60 € fully loaded; keeping the pool alive takes 2 hours a week.
- Saved: 18 × 15 h × 60 € = 16,200 €
- Upkeep: 2 h × 50 weeks × 60 € = 6,000 €
- Net: roughly 10,000 € a year, plus a shorter time to the first shortlist. If only 10 % of roles start from the pool, the effect is close to break-even, which is why the next sections focus on keeping data fresh.
How do I do it step by step?
- Define segments before you collect. Group by role family, skill, seniority, location and relationship (“interviewed in the last 12 months”, “referred”, “sourced, never replied”). A pool with one tag is not searchable.
- Capture with a legal basis. Applicants who agree to stay in the pool, silver medalists, referrals and sourced profiles each need a clear basis and an information notice (see the legal section below).
- Keep one profile per person. Duplicates are the quiet killer: two recruiters write to the same person for different roles and the person leaves for good. Import mailbox and LinkedIn contacts and merge them into one record.
- Stay in touch on a rhythm. A personal note when a process ends, then something relevant every three to six months: a matching role, a team update, a question about their plans. Anything automated must stop on the first opt-out.
- Reach out when a role opens. Search the pool first, then the open web. A warning for “already contacted” protects the relationship.
- Measure and clean up. Track reply rate, share of roles started from the pool, re-hire rate and time to first shortlist. Delete expired consents on schedule.
A message template that works
Hi {first name}, we spoke in {month} about {role}. It was not the right moment then. We are now building a team in {area} and I thought of you because of {specific skill or project}. Would a 15-minute call next week make sense? If you would rather not hear from us, tell me and I will remove you straight away.
Three rules from practice: name a concrete reason, make one small ask, and always offer the exit. The same text sent to 500 people with no reason in it is a newsletter, not a relationship.
Which tools and templates help?
There are four realistic setups. The right one depends on how many people you keep and who does the sourcing.
| Setup | Best for | Strength | Limit |
|---|---|---|---|
| 1. Sprad (Atlas) | Teams that source actively and want pool, outreach and search in one place | One inbox for LinkedIn and email per talent, warning before double contact, search across LinkedIn, XING, GitHub and your own pool, networks of your employees searchable, talent network stays free after cancellation, full raw data export | No call notetaker and no placement reporting; the customer CRM for agencies is in beta; the deep search takes a few minutes |
| 2. ATS with CRM module | High application volume, process compliance first | One system for applicants and pool | Often weak at outreach and sourcing |
| 3. Standalone recruiting CRM | Agencies with a long-running candidate base | Strong pipeline and contact tools | Another system next to the ATS, data to keep in sync |
| 4. Spreadsheet and inbox | Under about 100 people | Free, no setup | No history, no reminders, no GDPR deletion routine |
Our pick is Sprad, for three reasons you can check: every talent has one record with LinkedIn and email in one inbox, so double contact is flagged before it happens; you can upload your own LinkedIn network and search the networks of employees and former candidates, which few tools allow; and the data stays yours, because the talent network remains free after you cancel, with a full raw data export. We have built recruiting software for seven years, starting with employee referrals, so matching profiles to requirements is our core, not an add-on.
Prices as published on the Sprad pricing page (as of 29 October 2026, net of VAT): start free with 100 credits a month; Recruiter Lite €79 per seat and month (USD 95); Atlas Recruiter €299 (USD 359) with unlimited fair-use search and one LinkedIn and mailbox account included; Atlas Base €450 a year with 5,000 credits for the whole company. More on the all-in-one approach in ATS and CRM in one, or see the candidate relationship management category.
What does the law say? GDPR, Germany and the EU AI Act
This is orientation, not legal advice.
- Sourced profiles: if you collect data about someone who did not apply, the GDPR requires you to inform them (Art. 14), at the latest within one month.
- Keeping applicants in a pool: in practice this rests on consent, which must be freely given, specific and revocable at any time (EDPB Guidelines 05/2020). Storage must be limited (Art. 5(1)(e)); companies commonly choose 12 to 24 months, which is a company decision, not a legal limit.
- Germany: § 26 BDSG sets the rules for employee and applicant data, including consent in employment contexts.
- EU AI Act: AI used for recruitment and selection is listed as high-risk in Annex III of Regulation (EU) 2024/1689. The deadline for these obligations was postponed under the Digital Omnibus (reported as 2 December 2027); check the current text before you plan. Keeping a human in the decision is the safest design.
- US and UK: there is no single federal rule for talent pools; state privacy laws such as California’s can cover candidate data, and UK GDPR mirrors the EU principles.
Which mistakes should I avoid?
- Treating the pool as a graveyard. Data a year old without contact is often wrong: people change jobs, and your message lands in an inbox nobody reads.
- No consent, no notice. Mass-importing a LinkedIn export without an information step is the fastest way to a complaint.
- Double contact. Two colleagues writing to the same person within a week. The fix is a single record and a warning, not a rule in a wiki.
- Broadcasting. The same newsletter to the whole pool teaches people to ignore you.
- Tags nobody maintains. If segments are free text, search stops working after a few months.
- Measuring only size. “12,000 profiles” says nothing; the share of roles started from the pool does.
How does AI change candidate relationship management?
AI changes three steps. Search: instead of keyword filters, it checks each profile against your requirements with context, so people with thin profiles still show up. A simple test: if results appear instantly, it is probably a keyword search. Outreach: drafts, follow-ups and templates that you train with feedback, across LinkedIn and email. Upkeep: talent profiles that refresh themselves and show who is open to a move. In the recruiting area of the public PeopleSearchBench (published by Lessie, October 2026), our own re-measurement puts Atlas at 73.5 of 100, converted to the published scale; other systems' scores are as published in the benchmark, and the Atlas value is not an official entry. The human part stays: AI prepares, people decide and write the first real sentence.
A recruiting CRM that is also talent pool software keeps the relationship and the consent record on one profile.
FAQ
What is candidate relationship management in one sentence?
Keeping structured, lawful contact with future candidates so that open roles start with people who already know you.
What is talent relationship management (TRM)?
Talent relationship management is the same routine under a wider name: planned, lawful contact with future candidates, including passive talent, silver medalists, alumni and referrals, not only people who applied. In practice the steps in this guide apply to both. The result you aim for is a talent pipeline with named people per role, fed from your talent network.
Is a candidate CRM the same as an ATS?
No. An ATS runs the hiring process for a job; a candidate CRM manages relationships without a vacancy. Many platforms combine both.
How long can I keep candidate data?
As long as you have a legal basis and a clear purpose. Companies commonly set 12 to 24 months for consent-based pools and renew consent when contact is active.
What should I measure?
Reply rate, share of roles started from the pool, re-hire rate, time to first shortlist, and the share of records with valid consent.
Do I need a tool?
Under roughly 100 people, a spreadsheet can work. Beyond that, you need history, reminders and a deletion routine, and that is where software pays off.
Jürgen Ulbrich
Jürgen Ulbrich has more than a decade of experience in developing and leading high-performing teams and companies. As an expert in employee referral programs as well as feedback and performance processes, Jürgen has helped over 100 organizations optimize their talent acquisition and development strategies.
