Reference Check Request and Collection
Reference checks are one of the last steps before an offer, but collecting them manually means a recruiter phoning or emailing each reference individually, playing phone tag across time zones, and then writing up notes in whatever format they happened to use that day. Inconsistent reference-check formats make it hard to compare candidates fairly, and the process frequently becomes the bottleneck that delays an offer by a week or more while a recruiter tries to reach a reference who's slow to respond — often at the exact moment a competing offer is putting pressure on the timeline.
STARTING PRICE
From €299
Standard tier · Multi-step workflow with AI extraction/decisioning and 2-3 integrations.
Get a quote →Saves roughly 2-3 hrs per candidate, faster time-to-offer.
How the automation works
We send structured reference requests to each named reference automatically, with a consistent question set tailored to the role so responses are comparable across candidates, and follow up on a set cadence rather than relying on a recruiter to remember who hasn't responded yet. References can respond via a short form or a scheduled call, with call transcripts summarized into the same structured format as written responses. Once all references respond, or a configurable minimum is reached, a compiled summary goes to the hiring manager and recruiter with responses organized by question, making it straightforward to compare candidates or spot a concerning pattern across multiple references.
Process flow
- 01
Reference check stage reached trigger
Candidate reaches the reference-check stage in the ATS, and named references are pulled from the candidate's submission.
- 02
Send structured requests output
Each reference receives a request with a consistent, role-tailored question set, offering both a quick-response form and a scheduled-call option.
- 03
Follow up on cadence ai
Non-responsive references are followed up automatically on a set schedule, without a recruiter needing to track who's replied and who hasn't.
- 04
Summarize call-based responses ai
For references who complete a call rather than a form, the transcript is summarized into the same structured format as written responses for consistency.
- 05
Compile comparison summary output
Once responses are in, a compiled summary organizes answers by question across all references, surfacing any concerning pattern that spans multiple respondents.
- 06
Update ATS status integration
Completion status and the compiled summary sync to the candidate's ATS record, unblocking the next hiring stage automatically.
Inputs
- Candidate-provided reference contact details
- Role-specific reference question set
- Reference response method (form or scheduled call)
- Minimum response threshold to proceed
Outputs
- Structured, comparable reference responses per candidate
- Compiled cross-reference summary
- Non-response follow-up log
- Time-to-complete reference check metric
Works with
Prefer a fully custom build instead of an off-the-shelf integration? We scope both options during your free consultation — most jobs like this one work fine on standard connectors, but higher-volume or non-standard systems sometimes need bespoke API work, reflected in the complex tier.
Where this goes wrong if you get it wrong
- A single generic reference question set applied to every role fails to surface what actually matters for a specific job — questions for a people-manager role should probe management style and team feedback specifically, while an individual-contributor technical role needs different questions.
- Treating a non-response as neutral or positive by default, rather than flagging it explicitly, lets a candidate who provided references likely to avoid answering slip through with an incomplete picture — an unreachable reference should be surfaced as a gap requiring follow-up, not silently treated as a pass.
- Summarizing a reference call transcript with AI risks softening or missing hesitation and hedging language that's often the most important signal in a reference check — summary generation needs to preserve notable tone and hedging cues, not just extract stated facts.
- Collecting and storing reference feedback that includes protected-characteristic-adjacent commentary a reference volunteers unprompted, without filtering it before it reaches the hiring manager, creates legal exposure — the compilation step needs a review pass for content that shouldn't factor into a hiring decision.
Frequently asked questions
How does this handle references who prefer a phone call over filling out a form?
References can choose a scheduled call instead of a form; call transcripts are then summarized into the same structured format as written responses, so both methods produce comparable output.
What happens if a reference doesn't respond at all?
Non-responses are followed up automatically on a set cadence and, if still unresolved, are surfaced explicitly as a gap in the summary rather than treated as a neutral or passing result.
Can questions be tailored differently for management roles versus individual contributor roles?
Yes, the question set is role-tailored, since what's worth asking a reference differs meaningfully between a people-manager hire and a technical individual contributor.
Does the summary preserve nuance like hesitation or hedging in a reference's response, or just facts stated?
Summarization is built to retain notable tone and hedging cues from call-based responses, since that context is often as informative as the literal statements made.