THE SCORECARD
Three VP Engineering objectives that define whether the roadmap ships.
The CTO owns architecture. EMs own squad velocity. You own the delivery machine between them. At 200-500 SaaS, three outcomes show up on every VP Engineering scorecard — and none of them are "manage engineers." They're commitments the board can measure. DORA supplies the benchmarks.
| Objective |
Key Result |
Benchmark |
Target |
| Become the predictable engineering org the CEO can commit the roadmap on O1 · The question behind every board meeting |
Ship 85%+ of what you commit per sprint |
Typical 60–70% |
> 85% |
| Deploy to production at least daily |
Top 15%: multiple/day |
≥ 1/day |
| Ship software the customer never sees break O2 · Where your EMs either earn trust or lose it |
Hold change failure rate below 12% |
Elite: 0–15% |
< 12% |
| Cut mean time to recovery |
Top 15%: < 1 hour |
< 45 min |
| Cut change lead time from commit to prod |
Top 15%: < 1 day |
< 24 hrs |
| Turn engineering capacity into a compounding asset O3 · The CFO's actual question when velocity drops |
Hold unplanned work below 20% of capacity |
Typical 35–45% |
< 20% |
| Cut on-call burden per engineer |
Typical 8–12 pages/wk |
< 4 |
Why O1 is where most VP Engineerings lose the seat
Most VP Engineerings optimize O2 and O3 because they're engineering-native — CFR, MTTR, toil. O1 is where the job actually lives. A VP who ships the roadmap at 60% predictability is harder to keep than one who ships 15% less surface at 90%. Product plans against your commitments. Sales sells against them. Marketing launches against them. When the commitments aren't real, every downstream team gets burned — and they remember whose fault it was.
STRATEGIC BETS
The three strategic bets inside every VP Engineering OKR stack.
Every sprint planning, squads bring new priorities — platform upgrades, tech debt, scope expansions, experiments, integrations. Backlog is infinite. Capacity isn't. Strategy at this stage is deciding what not to do so the three bets that actually move DORA numbers get the capacity. Those three bets: cut toil, build a real platform layer, make the roadmap honest.
More than half the initiatives below need something from another team — Product, Security, Finance. When dependencies slip, sprint predictability slips. The ones marked below are the silent killers decomposed in Section 6.
Strategy 1 — Cut the toil tax from 40% → 20%
→ OBJ-3 · OBJ-4
1.1
Audit every recurring manual ops task; automate or kill top 10 by week 6
Internal
1.2
Spin up a dedicated platform squad (3 engineers, 1 EM)
Recruiting + Finance
1.3
Roll out incident classification + auto-triage to cut page volume
Security
1.4
Mandatory postmortem on every P0/P1 within 72 hours; action items tracked as KRs
Internal
Strategy 2 — Fix the CI/CD pipeline
→ OBJ-1 · OBJ-2
2.1
Cut median PR merge time from 4 days → under 24 hours
Internal
2.2
Mandatory feature flags on every production deploy
Security review
2.3
Parallelize test suite; target full run under 8 min
Internal
2.4
Kill any deploy that requires more than 2 humans to ship
Internal
Strategy 3 — Make the roadmap real
→ OBJ-5
3.1
Force T-shirt sizing on every story above 2 points at refinement
Product
3.2
Add 20% buffer to every sprint commit — non-negotiable with Product
Product
3.3
Mid-sprint check-in at day 5: raise red early or shut up at retro
Internal
3.4
Kill the "stretch goal" culture — committed work only, or it's not on the board
Internal
ENFORCEMENT LAYER
Enforcement for VP Engineering OKRs — the 2 triggers that fire hardest in your seat.
ShiftFocus watches seven signals on every KR. All seven apply to you. Two define your daily pain as VP Engineering: Velocity Drop (Trigger 2) and Momentum Decay (Trigger 3). The roadmap slips you can't recover from are almost always one of these two, caught in week 10 instead of week 4.
The two that fire hardest in engineering
Trigger 2 · Velocity Drop — your 4-week warning
⚡ Fires whenSquad velocity falls below 50% of planned pace — measured from the KRProgressHistory table against baseline
▎ Why this matters
A squad at 30% in week 4 looks fine on a Jira dashboard. Velocity math projects them to land at 60%, not 100%. Most teams see that in week 11. ShiftFocus sees it in week 4 when Trigger 2 fires — while there's still time to descope, reallocate, or escalate.
▎ Example scenario
Squad committed to 240 points across 6 sprints. Sprint 3 close: 96 points shipped. Planned pace: 40/sprint. Actual: 32/sprint. Velocity = 0.80. Trigger 2 fires at 0.50 threshold — still green here. But momentum tells you where the slope is headed. See Trigger 3.
Trigger 3 · Momentum Decay — the slope trigger
⚡ Fires whenWeek-over-week velocity decelerates 2+ sprints in a row
▎ Why this matters
Velocity tells you pace. Momentum tells you whether pace is getting better or worse. A squad at 70% of commit with 0.8 momentum will land at 45% by end of quarter. The number on the dashboard doesn't show that — the slope does. Trigger 3 catches failing KRs 3-4 weeks before traditional sprint reporting does.
▎ Example scenario
Squad's sprint 1 velocity: 38 pts/sprint. Sprint 2: 34. Sprint 3: 30. Momentum = 30 ÷ 34 = 0.88, and 34 ÷ 38 = 0.89. Two consecutive decelerating sprints. Trigger 3 fires. KR flags with momentum-decay indicator. Your next staff meeting has this squad at the top — not at the QBR six weeks later.
The other 5 that also fire on your KRs
Trigger 1 · Missed Check-in
⚡ WhenEM misses weekly KR update. 48h clock, then escalates to you.
Trigger 4 · KPI Drift
⚡ WhenChange failure rate or deploy frequency crosses threshold — parent KR flags red.
Trigger 5 · Owner Absence
⚡ WhenKR ownership "shared" across EMs, or owner inactive 7+ days.
Trigger 6 · Dependency SLA
⚡ WhenProduct spec, Security review, or Design mocks past 48h SLA — blocking team's KR flags.
Trigger 7 · Projected Miss
⚡ WhenProjected end-of-quarter completion drops below 70% at week 6 — exec brief fires.
Why this works where integration-based tools fail
Lattice needs HRIS integration. Viva Goals needs Jira integration. Every time a customer changes sprint tools, the enforcement breaks. ShiftFocus enforces at the accountability layer — which is universal. Your squads can use Jira, Linear, Shortcut, or Notion. The enforcement works the same regardless of the tool.
ESCALATION DESIGN
The VP Engineering OKR escalation chain — 5 levels, all on a 48-hour clock.
Every trigger from Section 4 feeds into this ladder. The ladder climbs on time, not on human judgment. Nobody has to decide "is this bad enough to escalate" — the clock runs and the system moves it up.
Any trigger fires. KR owner gets a Slack DM + email with the KR status link and what changed.
L2 — Peer Flag
48h after L1
Owner hasn't responded. Initiative Lead + peer EMs notified. Peer review requested on the KR.
L3 — Manager Alert
48h after L2
You get the escalation brief: dependency map, blocker details, suggested reallocation. You own the next action.
L4 — Executive Brief
Week 6 auto-check
Projected completion below 70%. CTO/CEO receives one-page PDF: what's failing, why, recommended intervention.
L5 — Intervention
3 weeks before quarter end
Projected miss greater than 30%. War-room triggered. Resource reallocation or scope reduction required within 48 hours.
What this kills
The "nobody wants to be the bad guy" problem. In most eng orgs, slipping roadmaps surface at the quarterly review — when Sales and Marketing have already committed plans to dates that aren't happening. ShiftFocus surfaces the same drift at week 4 by auto-nudging when velocity signals fire. Nobody is a snitch. The clock is.
EXECUTION INTELLIGENCE
The three leading indicators every VP Engineering OKR needs.
Every VP Engineering reports the same number to their CTO: percentage-of-commit shipped last sprint. It's honest and it's useless. It describes what already happened, not what's about to. A squad at 80% with a momentum of 0.7 will land at 55% by sprint 6 and below 40% by end of quarter. The percentage hid the slope. The slope is what matters.
ShiftFocus calculates three numbers per squad, recomputed every sprint. These are the numbers you bring to your CTO instead.
What this looks like in practice
You're 3 sprints into a 6-sprint roadmap. The roadmap was scoped at 240 points across your 3 squads. You've shipped 96 points. Target pace: 40 points/sprint. Current pace: 32 points/sprint.
What the miss actually costs
The 48-point gap isn't the real cost. The damage compounds across the org and lands far outside engineering. Numbers below are sourced, not estimated.
Deals slipped because a feature missed Q2
3 enterprise deals committed on the launch date. 1 churned to competitor, 2 pushed 6 months. Median mid-market SaaS ACV $62K × 3.3× LTV.
1−$1.23M
Engineer attrition from the firefighting quarter
2 senior engineers leave. SHRM: replacement cost 100–200% of salary for specialized roles. Senior SWE base $170K × 150% midpoint.
2−$510K
Replacement ramp time
18–24 month productivity ramp for complex technical roles. 4 months × 50% productivity × $170K × 2 engineers.
3−$227K
Marketing launch with no feature to launch
Mid-market SaaS product launch cost: typical $300K–$500K in campaign spend, event commits, paid media already booked.
4−$380K
Valuation hit at next raise
Roadmap misses signal delivery risk. KeyBanc 2024: top-quartile SaaS valuation multiples drop 0.3–0.5× on missed guidance. Against $500M valuation.
5−$1.5M
Quarterly cost of running without enforcement
−$3.85M
1 Digital Bloom 2025 SaaS Pipeline Benchmarks — median ACV $62K at $10-20M ARR bracket. LTV multiplier from High Alpha/OpenView 2024.
2 SHRM + Work Institute 2025 Retention Report — 100–200% replacement cost for specialized roles. Senior SWE base from Exceeds.ai 2026 US SWE Salary.
3 Work Institute 2025 — 18–24 month productivity ramp for complex technical roles.
4 SaaS product launch benchmarks — typical B2B SaaS launch campaign budget ranges, Battery Ventures OpenCloud 2024.
5 KeyBanc 2024 Private SaaS Survey — valuation multiple sensitivity to delivery guidance.
The ROI math for a VP Engineering buying this internally
Quarterly cost: $3.85M. Annual: $15.4M. Stopping a single roadmap slip per year covers the tool cost many times over. That's the business case for your CTO — not "let's try another project tool."