A planning phase becomes a stall when it continues to collect information without changing a decision.
Workshops are held, options expand and stakeholder lists grow. The programme appears active. Nobody has committed to a result, rejected an option or released a change that produces evidence.
The source calls this “strategy coma”. The phrase captures activity without consequence. This article focuses on how a sponsor ends that condition without using arbitrary speed as a substitute for judgement.
Planning needs a decision contract
At the beginning of a planning phase, state:
- the decision to be made;
- who makes it;
- evidence required;
- questions outside scope;
- options that will be considered;
- output and acceptance;
- budget and decision date;
- conditions that justify extension;
- the smallest action possible afterwards.
Without this contract, every new question can appear relevant and every stakeholder can delay closure.
An extension should name the missing evidence and how it could change the choice. “More confidence” is not enough.
Distinguish necessary uncertainty from avoidable indecision
Some programmes face uncertainty that cannot be resolved cheaply before delivery. Legacy systems reveal hidden dependencies, user behaviour differs from documentation and suppliers change.
Other uncertainty reflects missing ownership or an unwillingness to choose between interests. More research cannot decide which practice receives priority or whether the firm will accept a temporary service compromise.
Classify open issues:
Evidence gap: investigate through research, data or a prototype.
Trade-off: put options and consequences to the authorised leader.
Dependency: assign an owner and date, with a fallback.
Risk acceptance: obtain the appropriate accountable decision.
Idea: place it outside the current decision unless it changes value materially.
This prevents the research team from carrying questions only leadership can answer.
Fear after failure needs a targeted response
A previous project can make caution rational. Review that project before designing another year of assurance. The companion article on why the last digital project did not deliver provides a fuller recovery view.
Identify which assumptions failed, which warnings were missed, how contracts and governance behaved, what evidence was unavailable and whether the organisation can now control those causes.
If data migration caused the failure, prototype representative data and reconciliation. If ownership was absent, appoint an empowered service owner. If adoption was assumed, observe tasks and include users. If a supplier concealed drift, change reporting, milestones and commercial controls.
General additional planning may leave the original failure mechanism untouched.
Consensus is not a universal requirement
A transformation can affect many stakeholders without giving each a veto over every decision.
Map decision rights. Consult people who hold evidence or carry impact. Give an accountable owner authority within risk, budget and scope tolerances. Escalate material trade-offs through a defined route.
Record dissent and accepted consequences. Forced agreement can produce vague scope; transparent disagreement lets delivery proceed with accountable judgement.
Partnerships may need formal consent for some commitments. Identify those early and separate them from normal product or project decisions.
One more benchmark rarely creates certainty
Competitor comparisons and vendor evaluations can inform a decision. They can also expand indefinitely because the external target keeps moving.
Ask what decision each benchmark affects. A competitor feature may reveal a user expectation and does not prove your clients need the same solution. An analyst grid may narrow products and cannot establish implementation fit.
Time-box market research according to materiality and refresh it only when a defined condition changes. Preserve sources and dates so leaders know what the evidence represents.
The first move should produce evidence
Starting small is useful when the increment tests an important assumption or removes a meaningful constraint.
Choose work with:
- connection to the intended outcome;
- enough value to matter;
- manageable risk and dependency;
- a clear owner;
- representative users or data;
- observable result;
- a route to integrate or stop.
A cosmetic quick win can build activity and no strategic learning. A 90-day sprint is not inherently right. The duration should match the task and evidence.
Examples include testing one onboarding hand-off, prototyping an integration against messy data, improving one high-volume client task or resolving content ownership for a priority journey.
Discovery should narrow commitment
A diagnostic is valuable when it reduces consequential uncertainty. It is another layer of planning when its questions, output and decision are undefined.
Use existing evidence before commissioning more. Challenge whether a full organisation assessment is needed for one contained choice. Require the discovery team to state which recommendations are independent of its own delivery services.
The source contained several two-week assessments, costs and conversion outcomes repeated elsewhere in the collection. Those cases need verification and should not be reused as universal proof that a diagnostic leads to a sprint.
Make the cost of waiting visible
Planning is not free. Include external fees, internal time, expiring contracts, duplicated operations, unaddressed risk and opportunity cost.
Do not invent lost revenue from anonymous visitors or assume competitors are “shipping” valuable work. Use firm-specific evidence and ranges.
Waiting can create value when it aligns with a contract expiry, completes critical research or avoids a dangerous commitment. Put that value beside the cost.
“Do nothing for now” can be a managed option with controls, triggers and a review date.
Use a decision meeting, not another workshop
Circulate a short paper containing the decision, evidence, options, uncertainties, recommendation and consequences. Invite the people needed to decide or supply material challenge.
During the meeting:
- Confirm the decision and authority.
- Resolve factual disputes or label uncertainty.
- Compare options against agreed criteria.
- Decide, or specify the single blocking evidence.
- Name the owner and next commitment.
- Record dissent, risks and review trigger.
Do not end with a general action to refine the strategy. If a decision cannot be made, its blocking condition must be explicit.
Govern the first release and the next choice
The first increment needs budget, quality boundaries, security, accessibility, data controls and acceptance. Moving quickly does not suspend normal duties.
Book the evidence review before work begins. Ask what happened, what changed in the original assumptions and whether to continue, adapt, stop or scale.
This is where strategy and delivery connect. The roadmap becomes a record of decisions informed by operation, instead of a document completed before operation starts.
WHNN can provide a recurring structure; read what WHNN looks like in practice. A fixed quarterly review should complement the cadence required by the work.
Know when more planning is right
Continue planning when a material, answerable question could reverse the commitment; affected people have not been represented; a legal, security or financial exposure is unknown; or the organisation cannot yet operate the service safely.
End or pause the proposal when nobody owns the outcome, the commercial rationale has disappeared, required capacity will not be provided or the preferred solution survives only by ignoring contrary evidence.
Begin when the next bounded step has a worthwhile outcome, sufficient controls and a decision-maker prepared to learn from it.
The firms that win at this aren't the ones with the best plans. They're the ones who started.
If you're not sure where that first step should be, the practical companion piece on where to start goes into the mechanics - what a good first 90 days actually looks like, and how to pick the right problem to fix first.
