Strategic mandate and decision boundaries · Lesson 1

Translate strategy into a portfolio thesis

6 min read

Lesson objectives

  • Build an outcome chain, outcome measure and guardrail to test an investment thesis.
Reading tools

An investment board receives six requests that all appear useful, but strategy does not instruct it to fund everything attractive. Suppose the organisation seeks to shorten delays in a core service while preserving financial trust over two years. The first portfolio governance task is to turn that statement into a decision thesis. What observable change in customer experience or operations matters? Which work could plausibly cause it? What evidence would change the board's view? A portfolio is a collection of initiatives managed together in service of strategic objectives, rather than a list of projects or a traffic light dashboard. PMI's public overview of its fourth portfolio standard connects portfolio management to strategic planning. It does not prescribe this organisation's weights or legal authority. The thesis must be agreed locally with sponsors, operations and finance, distinguishing assumptions from approved choices.

Name an end outcome, then trace a causal route from investment to that outcome. If the aim is shorter waiting time, a new booking system is an output. A higher share of appointments completed without rebooking is an intermediate change. A shorter wait at stable quality is the desired outcome. Screens delivered, users enrolled and days waited answer different questions. Completion of the system may not deliver the outcome if operating teams keep the old roster. State a necessary condition and a collectable observation for each link. This reveals where the project manager's delivery accountability ends and the operational benefit owner's responsibility begins. A proposal to stop or reshape the work can then point to a broken causal link rather than an impression that the team looks busy.

Choose measures that do not reward the wrong behaviour. An average wait can fall because complex cases disappeared from the sample. Define the denominator, period, customer segments and a balancing measure of quality. The median days from complete request to completed service and the ninetieth percentile may expose different experiences; the proportion of requests returned for correction protects quality. Establish a baseline from a comparable period and separate request types. Comparing a peak season only with the preceding quiet month is weak evidence. Name the data steward who validates the indicator, the operational owner accountable for the outcome, and a review date separate from project completion. The decision does not need an academic causal model, but it needs a clear account of what is measured and why the number could move for reasons other than the investment.

A usable thesis also states what the portfolio will not fund during this decision window. Mandatory compliance maintenance may have to proceed even if it does not improve the service metric. Treat it as a separate eligibility obligation or constraint, and compare discretionary proposals within the remaining capacity. A visually appealing website redesign may still leave the processing bottleneck untouched. Deferral does not mean that redesign has no value; it recognises opportunity cost when capacity is scarce. Keep a short record for each proposal: strategic objective, expected causal contribution, critical dependency, minimum evidence and decision owner. Record conditions for reopening the choice, such as changed demand, lost specialist capacity or a failed adoption test. Without explicit reopening conditions, strategy becomes an after the fact justification for whichever project has the strongest advocate.

Test the thesis in a decision meeting, not only in a lengthy document. Ask the sponsor why the same initiative would be preferred if the proposing team changed or the most attractive technology promise disappeared. Ask operations what daily action will change when the output is delivered. Ask finance where costs will appear and where released staff time will not become a cash saving. Divergent answers expose an investment hypothesis gap, not merely awkward wording. Record the choice, rejected alternatives, unproven assumptions, the owner of each test and its date. At a later review, compare events with conditions recorded before the outcome was known. A sound decision may face an external shock; a weak decision may succeed through luck. Governance quality therefore includes the information, tradeoff and authority available at decision time, followed by honest learning from observed results instead of rewriting the history of the choice.

Lesson summary

A portfolio thesis specifies desired change, causal links, measures, boundaries and evidence that would change the choice.