Tech Prioritisation Toolkit

The approach uses a gate first, then shared scoring, then a roadmap. No single framework covers all of this.

Flow of work through the templates

1. Everything found (systems, projects, ideas) goes through the Compliance gate first.
2. Systems are also classified with TIME. “Invest” and “Migrate” systems generate candidate work items.
3. Mandatory items go straight into the roadmap’s Now column and take capacity first.
4. Discretionary items are scored with WSJF. Low-confidence items become discovery work, not delivery.
5. New ideas go through the Opportunity Solution Tree before they are scored.
6. The ranked list becomes the Now / Next / Later roadmap. “Next” is explicitly dependent on funding delivery capacity.

Separate work the function must do from work it could do, before any scoring. This stops a compliance deadline being argued against a nice-to-have.

When Days 0–30, and again for every new item that arrives.

Gate questions

Triage log

Next step: mandatory items go to the roadmap’s Now column with their deadline. Discretionary items go to WSJF scoring (Template 4). G4 items also flag the system for review in the TIME model (Template 2).

Turn a messy system inventory into decisions. Each system is rated on business value and technical health, and that places it in one of four quadrants.

When Days 0–30. Revisit quarterly.

The quadrants

Scoring guide (1–5)

• Business value: 1 = rarely used, easily replaced · 3 = supports a regular process · 5 = critical to compliance or reporting
• Technical health: 1 = unsupported, undocumented, single owner · 3 = works but manual or brittle · 5 = supported, documented, integrated
• Scores of 3 or more count as “high”.

Inventory

Next step: Invest and Migrate systems generate candidate work items for WSJF. Eliminate systems generate decommissioning items, which are often quick wins. Add a simple data-flow sketch of how the systems connect once the inventory is stable.

One list of every ongoing project and initiative, formal or informal, with its real status. Informal “someone is building a spreadsheet” work counts too.

When Days 0–30, built from stakeholder interviews.

Interview prompts for each stakeholder

• What are you working on, or waiting for, that involves systems or data?
• What problem does it solve, and what happens if it stops?
• Who is funding or resourcing it today?
• What’s blocking it?
• What would you do next if you had a dev team?

Inventory

Look for: overlapping projects solving the same problem, stalled work that only needs a decision, and projects that depend on a Migrate or Eliminate system.

Rank discretionary work by what it costs to wait, divided by effort. It moves the stakeholder debate from “whose item” to “why now”.

When Days 30–60, scored jointly in a workshop. Re-score quarterly.

Formula

Scoring scales (1–5)

Confidence rating (for mixed data)

• High: backed by data (volumes, hours, error rates)
• Medium: consistent stakeholder evidence, some data
• Low: opinion or assumption only

Confidence is not multiplied into the score, because that hides uncertainty. Instead it decides what kind of work the item becomes:

Scoring sheet

Worked example (illustrative)

The reconciliation workbook ranks first despite lower value, because it’s cheap and removes a key risk. The dashboard is low-WSJF and low-confidence, so it gets parked.

Explore new ideas without jumping to solutions. Ideas are mapped to the outcome they serve and the problem (opportunity) behind them. Then problems are compared before solutions are.

When Days 30–60, alongside interviews. Keep it separate from the existing project list until ideas are ready to score.

Structure

OUTCOME (one measurable goal for the function)
├── Opportunity A (a user problem or need, in their words)
│   ├── Solution idea A1
│   │   └── Test / experiment
│   └── Solution idea A2
└── Opportunity B
    └── Solution idea B1
        └── Test / experiment

Example outcomes

• Reduce manual effort in the reporting cycle
• Improve accuracy and auditability of data
• Shorten the time to answer business queries

Opportunity log

Show direction and sequence without committing to dates you can’t yet support. It suits a function with no dev team yet.

When Days 60–90. Review monthly; re-plan quarterly.

Column rules

• Now (0–3 months): mandatory items from the Compliance gate, plus top-ranked, high-confidence WSJF items that fit current capacity (vendors, central IT, config changes).
• Next (3–6 months): top-ranked items that depend on the proposed dev team, plus discovery spikes that have validated. Mark each one with its dependency.
• Later (6+ months): strategic Invest and Migrate work, larger ideas from the Opportunity Solution Tree, and anything still low-confidence.

Roadmap

For each roadmap item, record: ID · outcome · owner · dependency (e.g. “needs dev team”, “needs vendor”) · confidence · mandatory (Y/N).

Score WSJF jointly, so competing stakeholders own the ranking rather than contest yours.

Before (1 week out)

• Share the scoring scales and the pre-filled item list (Templates 3–5).
• Ask each sponsor to bring evidence for their items: volumes, hours, error rates, deadlines.
• Agree on a decision owner for tie-breaks, such as the head of the function.

Agenda (about 2 hours)

Pitfalls to avoid