All outcomes
Teams

Build a Hiring System That Works

12 weeks · 4 milestones

Milestone map

Milestone map

3 milestones

Audit current hiring process and identify failure modes

3 weeks

Conduct a structured audit of your current hiring process by gathering data on time-to-hire, offer acceptance rate, 90-day new hire retention, and hiring manager satisfaction. Interview at least three people who have been hired through the process (candidates) and three hiring managers to identify where the process breaks down from both perspectives. Produce a written audit report summarising the data and identifying at least three specific failure modes.

Proof required

Submit a hiring process audit report of at least 500 words covering: time-to-hire, offer acceptance rate, and 90-day retention data (or proxies if exact data is unavailable); interview findings from at least three candidates and three hiring managers; and at least three specific named failure modes in the current process. ADVERSARIAL VERIFICATION: a named HR or talent professional with 5+ years of hiring experience must read the audit and provide written confirmation that the failure modes identified are substantive — not surface-level descriptions of symptoms.

What gets checked

  • 500-word audit report with hiring metrics (even approximate), interview findings from ≥3 candidates and ≥3 managers, and ≥3 named failure modes
  • Failure modes are specific process breakdowns — not generic observations like 'the process takes too long'
  • Written confirmation from a named HR/talent professional (5+ years experience) that the failure modes are substantive

Common mistakes

  • Collecting opinions only from hiring managers and not from candidates — the most common hiring process failures (slow communication, vague expectations, disorganised interviews) are only visible from the candidate side
  • Identifying symptoms rather than failure modes — 'we lose candidates to competitors' is a symptom; 'our time-to-offer averages 3 weeks from first interview and we lose 40% of candidates in that window' is a failure mode
  • Not gathering any data and writing an opinion piece instead — even approximate data (estimated times, rough acceptance rates, 'we've lost 4 of 6 engineering hires in the first 90 days') grounds the analysis in reality

Resources

Foundationstart here

Depthgo deeper

What a verifier looks for

  • The audit must include both candidate and hiring manager perspectives — an audit based only on internal feedback has a significant blind spot
  • Failure modes must be named specifically — ask the submitter to describe the failure mode in one sentence with a causal mechanism; 'we lose candidates late in the process because…' is the right level of specificity
  • The confirming reviewer must have 5+ years in HR, talent acquisition, or people operations — a hiring manager who has participated in interviews is not a sufficient substitute

Design and pilot improved hiring process components

6 weeks

Design improvements to at least two of the three failure modes identified in M1, implement them as a pilot for at least two live hiring processes (not hypothetical), and document the results. Improvements may include: structured interview scorecards, standardised candidate communication templates, revised job description formats, take-home assessment design, or interviewer calibration sessions. Each improvement must be documented before implementation (what you changed and why) and measured after (what happened).

Proof required

Submit a pilot report covering: (1) the two process improvements designed and implemented, with the before/after process description for each; (2) documentation of at least two live hiring processes that used the improved components (role, dates, outcome); (3) measurement of at least two metrics comparing the pilot to the baseline from M1. ADVERSARIAL VERIFICATION: your named HR/talent reviewer reads the pilot design (before implementation) and provides written feedback on whether the improvements address the identified failure modes or just treat symptoms.

What gets checked

  • Two process improvements documented before and after implementation — not just 'we changed the interview format'
  • At least two live hiring processes that used the new components — hypothetical pilots or planned but not executed changes do not satisfy this milestone
  • Two metrics measured before and after — even approximate metrics are acceptable if exact data is unavailable

Common mistakes

  • Designing improvements on paper and not running them through a live hire — a process improvement that has never been used in a real hiring context has unknown friction points
  • Measuring outputs (did we make an offer?) rather than process quality (did candidates receive the decision within 3 days of the final interview, as designed?) — process improvements should be measured against the process they changed, not just against end outcomes
  • Making too many changes simultaneously and not being able to identify which improvement had which effect — pilot two components at a time, not all three failure modes at once

Resources

Foundationstart here

Depthgo deeper

What a verifier looks for

  • The reviewer's pre-implementation feedback is the primary adversarial check — ask them to confirm that the improvements address the failure mode mechanism, not just the symptom
  • At least two live hiring processes must be documented — ask for the role titles and outcome (hired, declined, withdrawn) for each
  • Before/after metrics are required — approximate values (estimated from memory or partial records) are acceptable; the absence of any measurement is not

Present hiring process improvement results with practitioner Q&A

2 weeks

Compile a hiring process improvement report covering your M1 audit, M2 pilot results, and a recommendation for the next improvement cycle. Present the report to a panel of at least one HR/talent professional (5+ years experience) and one hiring manager in a live session of at least 30 minutes. The panel poses adversarial challenges — questioning your metrics, your causal claims, or the generalisability of your pilot results.

Proof required

Submit your hiring process improvement report (M1 audit + M2 pilot results + recommendations) and a record of the challenge session: names and roles of panel members, date and duration (≥30 minutes), and notes on at least three challenges posed and your responses. A written statement from the HR/talent panellist (50+ words) confirming that your causal claims are well-supported by the evidence presented. ADVERSARIAL VERIFICATION RULE Level 2: live session required.

What gets checked

  • Hiring process improvement report covering audit, pilot results, and at least two specific recommendations for the next cycle
  • Challenge session record with ≥2 named panellists, date and duration ≥30 minutes, ≥3 challenges documented
  • Written statement from the HR/talent panellist (50+ words) confirming causal claims are evidence-based

Common mistakes

  • Presenting results without the underlying data — a good panel will ask to see the raw metrics and the before/after comparison; if the data is not in the report, the panel will not be able to evaluate the causal claims
  • Overstating causation from a two-hire pilot — the panel will challenge whether two data points constitute a process improvement; prepare to discuss confidence levels and next steps to validate at scale
  • Choosing panellists who were involved in the pilot as interviewers — panellists who were part of the process being evaluated cannot provide independent verification of the results

Resources

Foundationstart here

Depthgo deeper

What a verifier looks for

  • The HR/talent panellist must have 5+ years in talent acquisition or people operations — confirm their role before accepting the written statement
  • The session must be live and synchronous — email exchanges do not satisfy ADVERSARIAL VERIFICATION RULE Level 2
  • At least three challenges must be documented in the session record — a session where the panel only asked questions about the report format was not adversarial

We use analytics to improve Powstik. No ads, ever.