Reading the Outcome Attribution Report — Executive Track (Module 2)

In this module

What the Outcome Attribution Report (OAR) is
The four sections you'll see on every OAR
The three questions to ask every quarter
What NOT to do — how to misread the OAR
Reading drift alerts — your weeks-not-quarters early warning
A worked example — a real-shaped OAR with annotations
Discussion prompts

What the OAR is

The Outcome Attribution Report is the closed-loop platform's primary executive surface. One page per strategic objective. Generated automatically. Refreshed continuously.

The OAR answers four questions about each strategic objective:

Are we on track? (current trajectory vs. committed target)
What's contributing? (which shipped Briefs are moving this objective, by how much)
What's missing? (predicted impacts that didn't materialize)
What's the team's actual capability? (recalibrated AR(1) priors based on this objective's history)

The OAR is your quarterly business review, but with the heavy work pre-done. Instead of spending a week assembling decks from disparate spreadsheets, you spend an hour reading and annotating. The data is already there; you edit the narrative.

This is the single biggest time savings the closed-loop pattern produces at the executive level. Per McKinsey's product operating model research, high-maturity product orgs show 60% higher shareholder returns and 16% higher operating margins — and the differentiator is operational discipline around outcome measurement, not throughput.

The four sections on every OAR

The structure is consistent across objectives, which lets you read multiple OARs in parallel without context-switching cost:

Section 1 — Objective status (the headline)

The one-glance summary:

The objective statement (e.g., "Reduce TTFCV by 30% by end of Q4")
Current value vs. target (e.g., "22% reduction achieved; 30% committed")
Forecast trajectory: AR(1)-projected end-of-period value (e.g., "27% by EOQ at current Brief throughput")
Confidence interval on the forecast (e.g., "23-31% within 80% CI")
Status flag: ON TRACK / AT RISK / OFF TRACK / DRIFT

Section 2 — Top contributing Briefs (what's working)

A ranked list:

Brief title and short identifier
Predicted impact on objective (declared at filing)
Realized impact (measured at the configured window after ship)
Variance — predicted vs. realized, with confidence framing

This section is where you celebrate calibrated wins — Briefs that moved the metric within their confidence band. Calibrated wins are evidence the team has accurate predictive models, not just lucky bets.

Section 3 — Misses and surprises (what isn't working)

This is the section most leaders skip. Don't.

A ranked list of:

Briefs that shipped but didn't move their predicted metric
Briefs whose predicted-vs-realized variance exceeded the confidence band
Briefs whose measurement window hasn't closed yet (still in flight)

Each miss has a diagnostic field: wrong metric chosen, wrong target value, wrong measurement window, confound from concurrent work, or feature still gated. The diagnostic isn't automatic — it requires a 5-minute human investigation per miss. That investigation is the single highest-leverage learning activity in the quarter.

Section 4 — Recalibration history (what your team actually does)

The temporal record:

AR(1) prior for this objective at start of quarter vs. now
How the prior moved with each shipped Brief
Workspace-specific predictive accuracy compared to industry-default priors

This section is the one that earns the platform its keep over multi-quarter horizons. After 100+ shipped Briefs against a class of objectives, your workspace prior beats any industry-default prediction model. You have organizational knowledge that wasn't capturable before.

The three questions to ask every quarter

For each OAR, three questions land:

1. Are we hitting our objectives?

The honest answer comes from Section 1's status flag. ON TRACK or AT RISK is good news if you trust the forecast accuracy. To trust the forecast, look at Section 4 — has the AR(1) prior been stable over the last 4-8 weeks, or is it still moving? Unstable priors mean the model is still learning your team; the forecast carries more uncertainty than the confidence interval suggests.

2. Where is the gap between predicted and actual?

Section 3 is the answer. The pattern in misses matters more than any single miss. Three patterns to watch for:

Systematic over-prediction (Briefs ship but metrics consistently move less than predicted) → the team's optimism is built into the AR(1) prior; recalibration will tighten the model
Confound clustering (multiple misses share a common confound — feature flag, concurrent work, measurement window) → the measurement layer needs a fix, not the work
Random scatter (no pattern, just noise) → the underlying variance is genuinely high; consider increasing the confidence interval rather than the sample size

3. What does recalibration tell us about our team's actual capability?

Section 4. Workspace-specific priors are the most honest organizational self-assessment your leadership team has access to. They show what your team actually delivers, not what they say they'll deliver.

Three things to read from the priors:

Are predictions converging (variance decreasing over time)? If yes, the team is building organizational learning. If no, something is interfering with the loop.
Are predictions accurate (realized within band most of the time)? If yes, the team has good intuition. If no, the team has overconfidence or motivated reasoning.
Are predictions improving (this quarter's accuracy higher than last quarter)? If yes, the closed loop is functioning. If no, the loop isn't actually closing — usually because outcome reviews aren't happening or recalibration isn't being acted on.

What NOT to do

Three common misreads:

1. Cherry-pick high-attribution Briefs to declare victory

The OAR shows the WHOLE picture for a reason. Section 2 wins and Section 3 misses are equally informative. An OAR that only gets read for Section 2 is a vanity dashboard. Read Section 3 first if you have to ration time.

2. Treat misses as failures rather than data

Misses are not performance issues. They are the highest-leverage learning data in the platform. Punishing teams for misses creates incentive to game the predictions — to file Briefs with conservative predicted impacts so the realized variance always looks favorable. That destroys the closed loop.

The principle: predictions should be calibrated, not pessimistic. A team that ships a Brief predicting 20% lift and realizes 22% is well-calibrated. A team that ships a Brief predicting 5% lift and realizes 22% has a worse predictive model, even though the realized outcome is identical.

3. Read the OAR as a static report

The OAR is alive. Section 4's recalibration history changes weekly as Briefs land. Drift alerts in Section 1 trigger within the quarter, not at quarter-end. Treating the OAR like a static document means missing the in-quarter signals that are the whole reason the loop is closed.

Reading drift alerts

Drift alerts are your weeks-not-quarters early warning.

When a forecast trajectory diverges from a target by more than the confidence interval allows, the OAR flags drift. The alert tells you:

The objective is now forecast to miss target by X
This is Y weeks earlier than you'd have noticed at quarter-end
The Brief throughput needed to get back on track is Z additional Briefs in the remaining window

A drift alert is an executive cue. It doesn't tell you what to do — it tells you the situation needs your attention. The right response is usually one of three:

Re-prioritize: pull additional Briefs into the strategic objective's lane (if capacity exists)
Reset target: if the objective was set unrealistically, recalibrate (and be honest about why)
Investigate: if the team is shipping but the metric isn't moving, Section 3 will tell you why; sometimes the right answer is "the work is fine, our hypothesis was wrong"

The 12-month difference between orgs that read drift alerts and orgs that don't: the former course-correct mid-quarter; the latter explain the miss at the QBR.

A worked example

Here's the shape of a real OAR (sanitized; numbers are realistic for a mid-stage SaaS company):

OBJECTIVE: Reduce Time-to-First-Customer-Value (TTFCV) by 30% by EOQ
STATUS: AT RISK (drift detected 2026-04-15)
Current: 22% reduction achieved
Forecast EOQ: 27% (80% CI: 23-31%)
Recalibration delta: prior was 33%, now 27% — model is more conservative than at start

TOP CONTRIBUTING BRIEFS (last 8 weeks):
  BRIEF-2403 (onboarding email sequence) — predicted 15% lift, realized 12% — within band
  BRIEF-2391 (sandbox auto-provision) — predicted 8% lift, realized 11% — within band
  BRIEF-2387 (first-API-call dashboard) — predicted 5% lift, realized 4% — within band
  Total attributed: 27% of total TTFCV improvement
  Unattributed (random walk + confounds): 73%

MISSES (last 8 weeks):
  BRIEF-2412 (email verification speedup) — predicted 4% lift, realized 0% —
    diagnostic: shipped behind feature flag; metric measurement window opened before flag
  BRIEF-2398 (signup form simplification) — predicted 3% lift, realized -1% —
    diagnostic: A/B test confounded with concurrent SEO push; variance dominated by traffic mix

RECALIBRATION HISTORY:
  Q1: AR(1) prior μ = 0.10 (10% predicted lift per onboarding-area Brief)
  Q2 in-flight: AR(1) prior μ = 0.07 (recalibrated down based on Q2 misses)
  Confidence: variance has decreased 18% over 2 quarters — model is converging

RECOMMENDED ACTION:
  Forecast says 2 additional onboarding-area Briefs needed this quarter to hit 30%
  Capacity-aware scheduler shows: capacity exists if 1 lower-priority Brief defers
  Strategic call required: defer BRIEF-2429 (analytics dashboard refresh) by 1 quarter?

Reading this OAR in 5 minutes:

The 30% target is at risk; the team is forecasting 27%
The misses are diagnosable and not catastrophic — measurement issues, not capability issues
The recalibration is healthy — variance is decreasing
The decision is whether to defer BRIEF-2429 to hit target

This is the executive moment. The platform did the work; you make the call.

Discussion prompts

Your current QBR effort: how many person-weeks does it take to assemble your current quarterly business review? What fraction of that is data gathering vs. narrative writing? The closed-loop pattern eliminates most of the data gathering.
Your current drift-detection latency: how late in a quarter do you notice an objective is going to miss? If the honest answer is "at the QBR" or "from the team's hedging language in updates," you're 12 weeks behind.
Your current attribution depth: pick the highest-priority strategic objective from last quarter. How many specific shipped work items can you attribute movement to, with quantified evidence? If fewer than 3, you have an attribution-depth gap, not a strategy gap.
Your reaction to misses: when a team reports a miss against a predicted metric, what's the cultural response? Curiosity or blame? The closed loop requires curiosity to function.