Hire to Exit

People lifecycle problem map - April 2026 | Expanded from 5 discovery sessions

Three teams are working on different segments of the same cost problem without a shared view. Bad hires (Chad/Bella) that can't be course-corrected (Stacey/Hannah) become expensive exits (Abby/Ashley). Each team's AI initiative is more valuable when framed as part of the lifecycle - and this is the strongest portfolio-level story for Carla's ~$1M AI investment question.

Click any card or heading to drill in
~100
Separations / yr
Across US, EMEA, Australia
0
Progressive discipline framework
No structure to course-correct
?
Quality of hire score
Not measured at Contentful
7
Recruiting KPIs tracked
Speed measured, quality is not
Employee Lifecycle
Stage 1 Hire
Chad (Head of TA) + Bella (Recruiting & Analytics)
7 KPIs tracked but all measure speed and efficiency. Quality of hire - the metric that makes faster hiring safe - is not measured.
View details
Bad hires
pass through
Stage 2 Perform
Stacey (PvP) + Hannah (L&D) + Aurelia (People Ops)
No progressive discipline framework. MTA spend is the proxy metric for systemic failure. Blocked until June offsite.
View details
Unmanaged
out via MTA
Stage 3 Exit
Abby (Employment Legal) + Ashley Adams (Legal Ops)
~100 separation agreements/yr. Separation agent in discovery. Only 2 surviving KPIs - lifecycle framing fills the metric gap.
View details
Hire - Full Detail
Sources: Chad 4/7, Bella 4/6
What's measured (7 KPIs)
  • Time to fill - role open to seat filled, daily via Tenfold
  • Time to hire - candidate enters funnel to offer signed, daily
  • Sourcing channel efficiency - 36+ pipelines beyond LinkedIn
  • Time to first qualified candidate
  • Diversity of hire - hires only, not candidates (legal constraint)
  • Cost per hire - not yet tracked, launching soon
  • Hiring manager NPS - relaunching June in simplified format
What's missing
  • Quality of hire - not measured at Contentful
  • Equation exists in concept: time in seat + performance rating + manager feedback
  • Data lives in CultureAmp but no calculation built
  • No forcing function - could drift indefinitely
Goodhart's Law risk - faster hiring without quality measurement
Bella's dual role
  • Recruiting execution + People Analytics lead (board-level reporting)
  • Historical KPI data back to 2019 - baseline for AI impact measurement
  • Owns analytics across entire People org (Jonathan, Lee, Richard, Stacy sections)
  • Explicit goal: eliminate recruiting coordinator role through AI automation
  • Hands-on technical: has information systems background, understands APIs, built multiple ChatGPT agents
Systems
CultureAmp Tenfold Greenhouse Juicebox (new)
AI already in flight
  • JD Generator - validated as first AI initiative, targets time-to-fill
  • Bella's ChatGPT agents: stack ranking, automated scheduling, social media
  • Juicebox: passive candidate outreach with engagement analytics
JD generator validated - first initiative
Chad: "Help construct a people function that utilizes AI to manage the tactical work or transactional work that we spend a lot of time on, providing space for us to spend more time on the work that matters most."
Perform - Full Detail
Sources: Stacey 4/9, Hannah (claude-mem)
What's broken
  • No progressive performance management framework exists
  • No early intervention mechanism for struggling employees
  • Managers lack tools for performance conversations
  • Employee engagement scores used as proxy but no action loop
  • Manager effectiveness = "most room for opportunity" (Stacey)
Systemic gap - no course correction before MTA
In flight
  • Manager 360s launching
  • Functional solutions exist: AE growth framework tied to MedPIC
  • June offsite will update priorities and framework
  • Chris consciously restraining from solutioning - discovery protected
Blocked - June offsite is the real planning moment
Hannah / L&D overlap
  • PvP and L&D initiatives are "very similar" (Stacey's words)
  • Hannah's vision: "automate entire tactical layer" so People team can serve as strategic partners
  • Aurelia runs People Ops globally - repetitive, cyclical, predictable processes
  • Nelson already built quarterly check-in agent for talent management
  • People Ops positioned as highest automation ROI in org
Solution architecture must account for overlap or risk duplication
Key metric
  • MTA / separation spend - the proxy for systemic failure
  • No dedicated performance management system today
Systems
CultureAmp (engagement)
Stacey: "It's just basically how much money we're spending on exiting people from the business because we don't do proper performance management."
Exit - Full Detail
Sources: Abby 4/3
Workload
  • ~100 separation agreements / year
  • Three jurisdictions: US, EMEA, Australia
  • EMEA has meaningfully higher complexity - different templates, severance rules by country + tenure (logic branching, not just localization)
  • Manual, consuming significant legal talent
Metric poverty
  • Outstanding litigation - top KPI (only one with teeth)
  • Legal ticket turnaround - healthy, not a focus
  • 5 candidate KPIs eliminated (matter resolution time, CSAT, legal cost as % of revenue, regulatory response time, annual state filings)
  • Legal spend tracking is informal - Ashley Adams flags verbally, no dashboard
Without lifecycle framing, this function is metric-invisible
Constraints
  • Strict access wall - must be isolated from commercial legal (compliance requirement, not preference)
  • Severance calculator must flag policy-limit violations and auto-reject out-of-bounds proposals
Key contacts
  • Vera Eberhard (Berlin) - EMEA templates and jurisdictional logic
  • Sam Haygreen - US/AU separation process
  • Ashley Adams - Legal Ops, existing Glean-Jira exploration
Prior art
  • Ashley/Tricia explored Glean connected to Jira for separation agreements
  • Tyler/Krissy built MTA request agent from Jira data
  • Annual state filings identified as quick-win fallback
Systems
Jira LinkSquares
Separation agent PRD in discovery
Abby: "Usually you hear about it when things aren't working well. But right now, I haven't... you kind of hear when people are not satisfied with the turnaround time."
- Reveals measurement is entirely complaint-driven, not data-driven.
Cross-cutting connections (6)
MTA = Separation Agreement
Stacey calls them Mutual Termination Agreements. Abby calls them separation agreements. Same process, same cost, different vocabulary. Any AI work on ER case management or spend analytics needs to treat these as unified.
Quality of Hire drives MTA volume
Chad: quality of hire is unmeasured. Stacey: no way to course-correct. Abby: processes ~100 exits/year. The volume Abby handles is a downstream consequence of gaps Chad and Stacey own.
PvP and L&D overlap
Stacey's PvP initiatives are "very similar" to Hannah's L&D work. Aurelia's People Ops is "repetitive, cyclical, predictable" - the highest automation ROI. Three teams (PvP, L&D, People Ops) must be coordinated or risk duplication.
Cost flows downhill
Each bad hire costs: recruiter time + ramp time (Chad/Bella), manager time + HR cycles + lost productivity (Stacey), legal time + severance package + external counsel (Abby). The total cost is invisible because no one tracks the full lifecycle.
Bella is the data backbone
Bella owns People Analytics across the entire People org - board decks, KPIs back to 2019, dashboards across all sub-functions (recruiting, L&D, total rewards, attrition). She is the single most important data access point for connecting all three lifecycle stages.
Zero Friction Lifecycle gap
Danny/Mikki's Tier 1 initiative (3/6/9-month CKO cadence) only covers the front end: headcount to onboarding. This problem map covers the full lifecycle. Perform and Exit have no Tier 1 coverage.
AI initiatives - concrete status (6)
Hire - Accelerate
JD Generator
Validated as the right first AI initiative. Targets time-to-fill (#1 KPI). Chain: better JDs, faster qualified applicants, reduced time to first qualified candidate, reduced time to fill.
Owner: Bella + Charlie
Validated - first initiative
Hire - Prevent
Quality of Hire Scoring
Automated real-time score from CultureAmp: tenure, performance rating, manager feedback. The metric that makes faster hiring safe. Equation undefined. No timeline.
Owner: Chad + Charlie
Not started - equation undefined, no forcing function
Perform - Catch early
Manager Enablement Tools
AI-assisted coaching, nudges, performance conversation tooling. Intercepts the problem before it becomes an MTA. Must wait for June offsite to clarify framework.
Owner: Stacey + Charlie
Blocked - June offsite defines framework
Exit - Reduce cost
Separation Agent + Severance Calculator
Integrated tool: severance calculator embedded in agreement generation, policy-limit flagging, auto-rejection of out-of-bounds. US + EMEA + AU. Access-walled from commercial legal.
Owner: Abby + Charlie | Vera (EMEA), Sam (US/AU), Ashley (Ops)
PRD in discovery
Exit - Existing
MTA Request Agent
Built by Tyler/Krissy from Jira data. Already exists. Represents organizational energy and prior art for separation workflow automation.
Owner: Tyler / Krissy
Already built
Exit - Quick win
Annual State Filings
Single-page filings, same information every year, same states. Auto-population candidate. Low stakes, high repeatability. Deferred behind separation agent.
Owner: Abby + Charlie
Deferred - fallback if separation agent needs runway
Timeline and sequencing
Now
Separation agent PRD in discovery - Exit stage moving fastest
JD generator validated as first Hire initiative
Cost per hire tracking launching soon - baseline opportunity
Apr 15
PRD Workshop (full team) - requirements thinking, feeds all three stages
Apr 20
Tyler returns for Glean deep-dive - separation agent could leverage his expertise
June
Stacey's PvP offsite - Perform framework gets defined (currently blocked)
Hiring manager NPS relaunches - coordination opportunity with Stacey's offsite
TBD
Quality of hire equation - no deadline, no forcing function. Critical gap that could drift indefinitely.
Ongoing
Zero Friction Lifecycle - 3/6/9 month CKO cadence (front-end only: headcount to onboarding)
Sequencing tension
The Exit stage (Abby's separation agent) is moving now while the Perform stage (Stacey's framework) is blocked until June. The stage that generates MTA volume won't have a framework until after the stage that handles the output is already being automated. This isn't necessarily wrong - reducing Exit cost has standalone value - but the "prevent upstream" narrative is aspirational, not yet operational.
Portfolio context - Carla's question
~$1M / year AI Solutions Partner cohort

Carla asked directly (4/9): "We're spending a million dollars a year... how we get visibility, how we actually think about where this gets in a year or two, and how we sort of maximize that." No consolidated ROI view exists. Executive-defined success metrics don't exist for any AI Solutions Partner. Mikki flagged this as an open question in real time.

Tier 1
Value streams
Cross-departmental initiatives like Hire to Retire, wall-to-wall sales, customer adoption. This lifecycle map is the strongest Tier 1 story available.
Tier 2
Domain partners
Charlie, Zuhoor, Veronica, Aben. Charlie's model is most mature and structured - being shown to cohort as candidate standard.
Tier 3
IC agents
Individual contributors building their own agents. Nelson's quarterly check-in agent, Bella's ChatGPT agents are examples.

Why this map matters for Carla: It connects three Tier 2 engagements (Chad/Bella, Stacey, Abby) into a single investment case. Fewer bad hires means fewer performance cycles means fewer separation agreements. The compounding story is what makes this portfolio-level, not three separate projects.

Vision tension to surface
Three different visions for the same transformation: Bella wants to eliminate the recruiting coordinator role ("we're trying to eliminate it"). Chad wants to free humans for relational work. Hannah says "AI allows us to be more human." Bella is talking headcount reduction; Chad and Hannah are talking role elevation. The portfolio narrative needs to reconcile or acknowledge this.
The unlock

Show the full lifecycle cost to Carla

Each team's AI initiative has a standalone ROI story. But the compounding story is what makes this a portfolio-level investment case: fewer bad hires means fewer performance management cycles means fewer separation agreements. This is the answer to Carla's ~$1M question.

Immediate actions:

1. Map MTA spend data (Stacey/Abby) to quality-of-hire signals (Chad). If even 20% of separations trace to identifiable hiring quality gaps, the full-lifecycle framing justifies investment at all three stages simultaneously.

2. Use Bella's historical data (2019-present) as the baseline before AI initiatives scale. Cost per hire launching soon; capture it before the JD generator changes the numbers.

3. Coordinate the June window: Bella's hiring manager NPS relaunch and Stacey's PvP offsite happen in the same month. Use that convergence to align the Hire and Perform stages.

4. Create a forcing function for Chad's quality-of-hire equation. It has no deadline and no owner pushing it forward. Without it, the prevention story remains theoretical.

Next steps with this document: Share with Chris to validate connections. Then use as the backbone of Carla's portfolio-level ROI narrative - connecting each team's AI initiative into a single investment case across the full employee lifecycle.