How to propose a phased digital investment so the board can decide on phase one

Phasing should improve a decision, not make a large programme look small enough to slip through approval. A responsible phase delivers a coherent outcome or tests a material uncertainty, with a genuine option to stop.

That can make an investment easier for a board to assess because commitment grows with evidence. It cannot guarantee a yes, and the board should still see the plausible total cost and destination.

Decide whether the work can be phased

Some initiatives have natural boundaries. A journey can be improved for one audience, a platform assumption can be proved, or one workflow can be piloted safely.

Other work has indivisible foundations. Security remediation, statutory compliance, data migration or a vendor exit may require a minimum complete scope. Artificially separating them can increase risk and total cost.

Ask:

• Can phase one produce value or decision evidence on its own?
• Can it operate safely if later phases never happen?
• Will it test an assumption that materially affects the programme?
• Are shared foundations funded early enough?
• Does phasing create excessive dual running, rework or technical debt?

If the answers are weak, present a different procurement or funding structure instead of forcing a pilot.

Choose a coherent first commitment

The best first phase balances relevance, learning, feasibility and consequence. It is not automatically the highest-impact, lowest-risk item; a trivial low-risk task may reveal nothing about the difficult programme.

Select a slice with:

• a specific client or operational problem
• a named outcome and owner
• representative users, data and integrations
• controlled failure and recovery
• enough frequency to observe within the phase
• a clear relationship to the longer strategy

For a website programme, rebuilding three pages may be coherent only if it tests proposition, content operations and enquiry handling. For an AI programme, one workflow may be useful if quality and oversight can be evaluated with representative work.

Document what remains outside scope. “While we are here” additions undermine the evidence and forecast.

Set success and stop conditions before approval

Agree a baseline and a small balanced set of measures. Include the intended business or service outcome, quality, risk, cost and adoption.

Targets should come from current evidence. Avoid invented percentages or a universal number of measures. State sample, timeframe and attribution limits.

Define conditions for stopping or pausing, such as a serious confidentiality failure, unavailable data, excessive review effort or a cost forecast outside tolerance. Also define a minimum evidence threshold; inconclusive results may require more observation rather than a convenient success narrative.

The board should know which outcomes the delivery team controls and which need benefit owners elsewhere in the firm.

Set duration from the decision

Ninety days maps neatly to many board cycles and may suit some phases. It is not a universal sweet spot.

Duration depends on task frequency, procurement, research, technical uncertainty, assurance, seasonality and the time needed to observe an outcome. A rushed phase can produce false confidence; an open-ended pilot can avoid decision indefinitely.

Set a fixed review date and define the evidence expected by then. If the outcome needs longer observation, separate build acceptance from the later benefit review.

Name accountability at three levels

One accountable phase owner prevents committee diffusion, but delivery rarely belongs to one person alone.

Name:

• the sponsor accountable for the investment and organisational decisions
• the delivery owner accountable for scope, forecast and evidence
• the service or benefit owner accountable for operation and outcome after release

Add named risk and assurance owners where appropriate. Give each enough authority, capacity and escalation.

An external supplier may deliver the work; it cannot accept the client's strategic or regulatory accountability.

Show the whole horizon transparently

Present later phases as options with indicative scope, range, dependencies and timing. State which assumptions make those estimates uncertain.

Use wording such as:

Today's decision concerns phase one only. If its evidence meets the gate criteria, we expect to recommend [next scope] within an indicative range of [range]. The board will decide that separately. If phase one does not support continuation, we will stop or redesign the approach.

Show continuing licence, support, product ownership and decommissioning costs. Do not present a low-cost pilot while concealing the commercial commitment needed to retain its result.

This transparency distinguishes staged governance from salami-slicing a budget.

Design a real gate review

Book the gate when phase one is approved. Specify decision makers and pre-read evidence.

The review should compare:

• planned and delivered scope
• baseline and observed outcomes
• quality and assurance evidence
• cost, forecast and internal effort
• incidents, exceptions and adoption
• assumptions supported or contradicted
• operating readiness
• changes to the longer business case

Offer at least four choices: continue, change and retest, retain the result within its current boundary, or stop. A recommendation to continue should include the next phase's scope, cost range, measures and risks.

Avoid sunk-cost arguments. Spending on phase one bought evidence and perhaps a useful service; it does not make phase two valuable.

Preserve architecture and service coherence

Phases can create a patchwork when each optimises a local result. Maintain target principles for content, data, identity, integration, accessibility, security and operations.

Review technical debt created intentionally for speed and decide when it is repaid. Do not postpone essential support, monitoring or governance to make phase-one economics look attractive.

Likewise, plan the client experience across phases. A partial journey should have clear boundaries and hand-offs. Users should not have to understand the delivery roadmap.

Use WHNN® without pre-committing Next

WHNN® separates the What and How for the Now and Next. “Now” is the authorised phase and evidence. “Next” keeps the strategic direction visible and names the conditions for another decision.

The quarterly rhythm may be useful when it fits the work. The governing principle is more important: later commitment follows observed evidence, rather than a forecast made before delivery.

Present the phase as a decision

The board paper should state:

• problem and evidence
• why phasing is appropriate
• phase-one outcome, scope and exclusions
• cost range, cash timing and internal capacity
• measures, assurance and stop conditions
• accountable owners
• gate date and choices
• indicative later horizon and total exposure

The companion article on what boards should know before approving a digital transformation budget covers wider governance. The article on the cost of inaction can help structure the credible defer option.

The phase-one proposal template available with this article includes scope, evidence, ownership and gate fields. Use it to make the first commitment genuinely decision-worthy, not merely easier to approve.