Prioritization Matrix & Sequencing Logic

Combines Step 2 evaluation scores and Step 3 AI-readiness gate results into a weighted queue. Every candidate lands on a quadrant, gets a priority score, and receives a sequence position.

1. Priority Score Formula

// Base from Step 2 (0–100) priority = composite_score // Modifiers (± points, applied in order) priority += strategic_deadline ? +15 : 0 priority += is_enabler ? +10 : 0 priority += ai_conditional ? −10 : 0 priority += has_blocking_dep ? −15 : 0 priority += ai_fail_reroute ? −30 : 0 // deterministic track; scored separately // Final priority clamped to 0–115 priority = clamp(priority, 0, 115)

Modifier	Trigger	Adjustment	Rationale
Strategic Deadline	Hard regulatory, contractual, or board-committed deadline exists	+15	Missing a deadline has cost, penalty, or reputation consequences that override pure ROI ranking
Enabler Bonus	This process unblocks 2+ downstream pipeline items	+10	Bottleneck-busting has multiplicative value — sequencing it early unlocks capacity for multiple follow-ons
AI Conditional	Step 3 gate returned CONDITIONAL	−10	Guardrail work must be completed before go-live; this adds schedule risk. The process remains eligible but carries a drag
Blocking Dependency	Process depends on another pipeline item that is not yet built or deployed	−15	Cannot start until the dependency resolves. Score penalty ensures it doesn't jump the queue ahead of its blocker
AI Fail — Reroute	Step 3 gate returned FAIL	−30	Routed to deterministic/RPA track with separate sequencing. The penalty removes it from AI-track contention

Clamping: The ceiling of 115 ensures a process can't score above a theoretical perfect candidate with all bonuses. The floor of 0 prevents negative scores. Processes rerouted to the deterministic track are ranked in a separate queue — their modifier penalty keeps them out of the AI-track top slots but doesn't make them invisible.

2. Effort × Impact Quadrant

Q1: Quick Wins — High impact, low effort. Sequence first to build momentum and credibility. Target: 30–40% of active pipeline.

Q2: Strategic Bets — High impact, high effort. Sequence second — after the governance rhythm is proven and the team has delivery confidence. Target: 40–50% of active pipeline.

Q3: Fillers — Low impact, low effort. Backfill when capacity opens between larger items. Useful for onboarding new team members. Target: 10–20%.

Q4: Hold — Low impact, high effort. Do not start. Revisit quarterly — a strategy shift or new technology may reclassify them.

Axes: Impact = avg(Financial Impact score, Strategic Alignment score) from Step 2. Effort = Implementation Ease score from Step 2 (inverted: 5 = easiest). Dot color = AI Readiness Gate result (Step 3). Numbers ①–⑩ map to the sample batch below.

3. Sequencing Rules

Processes are ordered by priority score within their quadrant, then interleaved according to these rules. The goal is a queue where every slot has strong rationale — not a mechanical sort.

1 Lead with unblocked Q1 PASS Highest-priority Quick Wins with clean AI gate results ship first. This builds delivery credibility and stakeholder trust. Aim for 1–2 go-lives in the first 4–6 weeks.
2 Pull in Q1 CONDITIONAL Quick Wins with AI guardrails. Start the guardrail work immediately (parallel track). The automation build and guardrail resolution run concurrently — gate must clear before go-live, not before build start.
3 Resolve blockers for enablers If a Q2 process carries the Enabler Bonus (+10) and has a resolvable dependency, prioritize clearing that dependency so the enabler can unlock downstream work. The dependency itself may not be high-value — do the minimum to unblock.
4 Q2 PASS — once governance rhythm is proven After 1–2 successful Q1 deliveries, the team has earned the right to take on a Strategic Bet. Sequence the highest-priority Q2 PASS process. Limit: one Q2 at a time until team capacity is well-understood.
5 Q2 CONDITIONAL — guardrails resolved Strategic Bets that were CONDITIONAL enter the queue once their guardrails are cleared. If guardrails aren't resolved within one planning cycle, demote to the next cycle's intake review.
6 Q3 Fillers — capacity backfill When the team has 1–2 weeks of slack between major items, pull a Q3 Filler. These are also ideal for onboarding new team members or testing new tooling in a low-risk context.
7 Deterministic track — separate queue Processes that FAILed the AI gate and were rerouted to deterministic/RPA are ranked in their own queue using the same formula (without the −30 modifier, since they're not competing with AI-track items). Sequence deterministic Quick Wins first, then Strategic Bets.
8 Q4 Hold — revisit quarterly No Q4 process enters the active queue. During the quarterly pipeline review (Step 6), re-evaluate: has strategy shifted? Has technology changed? Has a dependency resolved? If conditions change, re-score and re-plot.

Tiebreaker Rule	When Scores Are Equal
1. Strategic deadline	Process with a hard deadline wins
2. Enabler status	Process that unblocks others wins
3. Risk Safety (Step 2)	Lower-risk process wins (fewer surprises = faster delivery)
4. AI gate result	PASS beats CONDITIONAL; CONDITIONAL with fewer yellow dimensions beats more
5. Submission date	Earlier submission wins (tiebreak of last resort)

4. Sequenced Queue — Sample Initial Batch

Ten representative processes scored and sequenced. This demonstrates the full logic end-to-end. Processes marked with DET are in the deterministic track and sequenced separately.

#	Process	Fin	Str	Ease	Risk	Composite	AI Gate	Modifiers	Priority	Q	Seq
1	Expense Report Approval	4	3	5	4	80	PASS	+15 deadline	95	Q1	1st
2	New Hire Onboarding Tickets	3	3	5	5	77	PASS	—	77	Q1	2nd
3	Customer Sentiment Classification	3	3	4	3	65	COND	−10 cond	55	Q1	3rd
4	Invoice Exception Handling	5	4	3	3	78	PASS	+10 enabler	88	Q2	4th
5	Contract Clause Extraction	4	4	3	2	65	COND	−10 cond	55	Q2	5th
6	Email Triage & Routing	2	2	4	5	62	PASS	—	62	Q3	6th
7	Meeting Notes Summarization	1	2	5	5	60	PASS	—	60	Q3	7th
8	Vendor Risk Assessment	4	4	2	2	62	COND	−10 cond−15 dep	37	Q2	8th†
Deterministic Track (AI Gate FAIL — separate queue)
D1	Regulatory Report Generation	5	5	2	1	67	FAIL	+15 deadline det	82	Q2	D1
D2	Legacy System Data Migration	4	3	1	3	57	FAIL	−15 dep det	42	Q4	Hold

† Sequence position 8 (Vendor Risk Assessment) carries both a CONDITIONAL AI gate and a blocking dependency. It appears in the queue for visibility but cannot start until the dependency resolves AND guardrails are cleared. If neither resolves by the next planning cycle, it's demoted to Hold.

Deterministic track items (D1, D2) are not interleaved with AI-track items — they have their own capacity allocation and delivery cadence. D1 (Regulatory Report) is sequenced first in its track due to the strategic deadline modifier.

Q4 items (Legacy Migration) don't receive a sequence position — they're reviewed quarterly. The table includes them so nothing vanishes.

5. Integration: How the Three Steps Connect

Step	Output	Feeds Into
Step 1 — Intake	Submission template, triage workflow, intake register	Produces the raw candidate list. Only processes passing Gate 1 (documentable) and Gate 2 (automation-suitable) reach evaluation.
Step 2 — Evaluation	Composite score (0–100) across 4 weighted dimensions. Individual dimension scores (Fin, Strat, Ease, Risk) on 1–5 scale.	Composite score becomes the base for the Priority Score formula above. Individual Ease and Impact scores plot the quadrant position.
Step 3 — AI Gate	Gate outcome: PASS / CONDITIONAL / FAIL. Per-dimension R/Y/G indicators. Named guardrail conditions for CONDITIONAL processes.	Gate outcome determines dot color on the quadrant, applies the modifier penalty (−10 or −30), and routes FAIL to the deterministic track.
Step 4 — This Step	Priority score, quadrant position, sequence position, and a single ordered queue across both tracks.	Feeds into the rollout roadmap (Step 5) and the governance loop (Step 6), which monitor actual delivery against this sequence.

Recalibration cadence: Re-run the priority formula and quadrant plot after every 5 completed deliveries. If the top of the queue consistently under-delivers vs. plan, the Ease scores in Step 2 are probably too optimistic — recalibrate the Implementation Ease anchors. If Q1 is perpetually empty, the intake pipeline isn't producing enough low-effort candidates — the triage gates may be screening too aggressively, or process owners aren't submitting the right processes.