The Talent Intelligence Platform RFP checklist
Fourteen pillars, one head-to-head test, a scoring sheet. Copy it into your RFP, bring it to every demo, and score every vendor on the same page.
Same questions for every vendor. Dates, not adjectives.
How to use this checklist
A Talent Intelligence Platform is the single system of record for an executive search firm or a recruiting agency, holding all fourteen pillars natively. The checklist scores how much of that a vendor actually has today, in its own system, on its own data.
- Score each pillar for each vendor: 2 for native (built into the system of record by the vendor), 1 for bolt-on (a separate product you buy and integrate), 0 for absent (you supply it some other way).
- Ask the question in each row and write the date, number or place in the answer. "Coming soon" scores 0. "Has it" means generally available and usable today.
- Run the head-to-head test below on a live search before you total the scores. A platform at 28 that produces weak candidates has not passed; a feature set at 12 has not passed either.
The scoring sheet
| Pillar | What to verify | Ask the vendor | Score (2 / 1 / 0) |
|---|---|---|---|
| 1. Applicant Tracking System | Jobs, candidates, stages and activity in one record the rest of the system acts on | Was the ATS built around AI from the start or added later, and in what year? | ___ |
| 2. Recruiting CRM | One candidate record across every search, with every touch, transcript and outcome | Does a candidate keep one record for years across searches? | ___ |
| 3. Sales CRM | Accounts, contacts, deals and business development in the same system as the search | Where do client deals and business development live? | ___ |
| 4. Native sourcing | The vendor's own talent graph, searched inside the platform | Own graph or licensed? How large? Which generation of search? | ___ |
| 5. Contact enrichment | Verified email and phone found on the record | Is contact data inside the record, and since when? | ___ |
| 6. AI Chat | Questions answered from the whole firm's data, with cited sources | Can the AI see the whole firm or only the ATS, and does it cite? | ___ |
| 7. Outreach sequencing | Multi-step, multi-channel sequences run from the record, for candidates and clients | Do sequences run from the record for both sides of the desk? | ___ |
| 8. AI agents and agent workforce | Named agents that finish work, with their impact tracked | Which agents are generally available today, and where is their P&L? | ___ |
| 9. Full Agentic Mode | A search run autonomously from brief to shortlist | Can it run a search on its own, and where does it stop? | ___ |
| 10. API and MCP | Open API, MCP in the frontier models, a headless option | Is it open, and can we run our own instance on it? | ___ |
| 11. Revenue Operations | Fees, retainers, invoices and commissions from the search and the placement | Where does your billing start: the placement, or a spreadsheet? | ___ |
| 12. Hiring Manager and Client Portal | One live portal from intake to scorecard | Does the client see a live portal, or a document? | ___ |
| 13. Talent Insights | Size and shape of the addressable pool before the first call | Can you show me the market for this search right now? | ___ |
| 14. Memory | What the system keeps from past searches and uses on the next one | What does it remember from last year, and how does that help today? | ___ |
| The vendor | Tenure in executive search and recruiting agencies, generations shipped, roadmap direction | How many years in this segment, how many generations of each pillar, and what changed in your business model in the last twelve months? | Notes |
Maximum score: 28. Native in all fourteen is a Talent Intelligence Platform. Native in the first two with bolt-ons elsewhere is an ATS with a marketplace. Native in pillar 4 alone is a sourcing tool.
The head-to-head test
Run it before you total the scores. The same live search, in each system, judged on the candidates it produces and the steps it took.
- Pick a live search you are working this week and bring the brief.
- Create the job in each system. Count the steps and the minutes from the brief to a job the system can work.
- Run the longlist. Judge recall: how much of the real market did it find, and from whose data?
- Run the shortlist. Judge precision: would you present these candidates, and can the system explain why each one is here?
- Find the contact data and send the first sequence from the record, without leaving the system.
- Show the hiring manager the search in the portal they will log into.
- Score the other pillars on the sheet above, native, bolt-on or absent.
- Write down the dates in every answer. A platform's age is in its generations, not its logo.
Run the test with Loxo first
Tell Loxo a role you are working now. Loxo runs the live search before the call, brings real candidates with verified contact data, and scores itself on this sheet with you, row by row. Yours to keep, whatever you decide.
