A grid that sorts people by how much they care and how much they can hurt you is a communications budget. It is not a reading of the situation.
A government agency is two years into digitising a citizen-facing service. Its stakeholder analysis is thorough by any reasonable standard: a four-quadrant grid, twenty-eight named parties, each placed by power and interest, colour-coded for likely support. The minister's office and the departmental secretary sit in the top right, marked manage closely. A regional caseworker network sits bottom right — high interest, low power — marked keep informed.
The program is delivered on time. Adoption then collapses, because the caseworkers who process the majority of complex cases will not use it, and the agency discovers that its low-power, high-interest group holds something no quadrant recorded: the ability to withhold the practical cooperation on which the whole benefit case depended.
The map was not wrong about power or interest. It was answering a different question from the one the agency needed answered.
The Strategic Context
The power and interest grid is one of the most widely reproduced instruments in management practice. Four quadrants: manage closely, keep satisfied, keep informed, monitor. Place each stakeholder by how much power they hold over your work and how much interest they take in it, then allocate effort accordingly. The device appears in practitioner guides, project methodologies and consulting decks in almost identical form. Its origin is contested and usually unattributed. [FACT CHECK REQUIRED] — the grid is commonly credited to a specific author in the strategic management literature, and the attribution should be verified before crediting anyone.
Its usefulness is real and should not be dismissed. Attention is finite; a delivery leader cannot treat everyone equally; some ordering is better than none. The instrument does what it was built to do.
The difficulty is what organisations then use it for. A device designed to ration communication effort has been quietly promoted into a governance instrument — the thing consulted when a program asks who matters, whose objection counts, and what the pattern of support tells us about whether the investment is sound. For those questions it is the wrong tool, and its very familiarity conceals the substitution.
What Leaders Commonly Misread
The first misreading is that interest is a proxy for exposure. It is not. Interest measures how much attention a party is currently paying. Exposure measures how much of the consequence they will carry. These diverge constantly, and the divergence is where programs fail.
A group with high exposure and low current interest is the most dangerous entry on any map — people whose working lives will be materially changed, who have not yet engaged because the change is abstract, and whose interest will arrive suddenly and at full strength once it is not. The grid records them as monitor, minimum effort. That is precisely backwards.
The second misreading is that power means formal authority. The grid's power axis is usually populated with organisational seniority: ministers, executives, boards. But the power that determines whether a program realises its benefits is frequently the capacity to withhold something the program needs — cooperation, data, access, discretionary effort, a licence, local consent. That capacity is often held by parties with no seniority at all. The caseworkers in the opening example had no power in the grid's sense and complete power over the outcome.
The third misreading is the most consequential, and it is embedded in the instrument's own guidance. Practitioner advice commonly recommends colour-coding stakeholders by disposition — advocates in green, neutrals in orange, critics in red — and giving especially careful management to opponents who hold power.
The framing does something the user rarely notices. It classifies opposition as a hazard to be handled rather than as information to be read. Once a critic is red, the operative question becomes how to win them round or manage their resistance. The question that is never asked is the one that matters: why does the person best placed to obstruct this believe it should not proceed?
Concentrated, well-informed opposition among parties with high exposure is not a communications problem. It is evidence about the investment case. An organisation that has trained itself to see it as the former will systematically discard its best early warning.
Reframing the Issue
The reframing is not to discard the grid but to recognise what it does and to add the axis it is missing.
Stakeholder analysis in practice answers three separable questions that organisations routinely collapse into one:
- Who should receive how much attention? The power and interest grid answers this well. Keep it for this.
- Who is allowed to change what the program is for? A decision-rights question, with its own logic, treated in [Related article: Drivers, Supporters and Observers: Who Is Allowed to Change What Your Program Is For]. That classification captures influence over the definition of the outcome, which a power and interest grid does not attempt.
- Who carries the consequences, and who can withhold what we need? This is the question the opening example turned on, and neither of the other two instruments answers it.
A fourth question is often confused with all three and belongs elsewhere entirely: who holds the authority to stop this? That is a property of an office rather than of a map, and is treated in [Related article: Sponsorship Is an Office, Not an Endorsement].
The third question needs its own reading. Not a replacement for the others — an addition, and the one most often missing.
Strategic Analysis
The exposure and leverage reading
Map the same stakeholder set on two different axes.
Exposure to consequence: how much of the outcome — good or bad — lands on this party. Whose work changes, whose costs move, whose risk profile shifts, whose service is affected.
Capacity to withhold: what does this party control that the program cannot proceed or realise benefits without? Approval, data, access, expertise, discretionary cooperation, social licence, a workforce agreement.
Four positions follow, and each carries a different strategic instruction.
High exposure, high withholding capacity. These parties determine whether the program works. They are not an audience; they are effectively co-owners of the outcome, and should be involved in design rather than informed of it.
High exposure, low withholding capacity. The ethically and commercially most dangerous quadrant. These parties will bear consequences and cannot stop them. They generate the reputational failures, the industrial disputes, the political interventions and the adoption collapses that appear from nowhere eighteen months in. They are also, frequently, the people whose objection would have been most useful to hear.
Low exposure, high withholding capacity. Gatekeepers — regulators, central functions, technical authorities. Their requirements are usually procedural and knowable in advance. Manage these as dependencies with dates, not as relationships.
Low exposure, low withholding capacity. Genuinely low priority, and the only quadrant where minimum effort is the right answer.
The same principle in a different sector
The pattern is not a public-sector artefact. A mining operation seeking to expand a processing facility can hold every regulatory approval and every board endorsement, and still find that a small, highly exposed adjacent community holds an effective veto through legal challenge and political attention. Formal power sits with the regulator. Withholding capacity sits with the community. A grid built on power and interest places the regulator top right and the community bottom right, and the project is surprised by something entirely foreseeable.
Where opposition and exposure coincide, the finding belongs in the risk register and the investment case, not in the communications plan. Where it coincides with a physical or regulatory boundary as well, it is likely to be the binding constraint on the whole initiative — see [Related article: Read the Constraints, Not the Deliverables].
Communication as design, not volume
None of this makes communication planning unimportant. It makes it a design problem rather than a distribution problem.
The practitioner literature sets out eight variables for a communications plan: the stakeholders, the objective of each communication, the message, the information conveyed, the channel, the feedback mechanism, the level of detail, and the timing. Six of these are ordinary logistics. Two are governance decisions in disguise.
Level — how much detail — determines whether a recipient can form an independent judgement or only receive a conclusion. A board given summary status can approve or query. A board given the underlying assumption can test it. The choice of level determines which of those is possible.
Timing decides whether a problem is disclosed early enough to be acted on. The same guidance advises flagging problems as early as possible, and it is right, though the reason given — preserving the manager's reputation for reliability — undersells it. Early disclosure preserves the organisation's options. A problem raised at month four can be solved several ways. The same problem raised at month eleven can be solved one way, expensively.
The instrument also warns against the opposite failure — overwhelming senior stakeholders with detail they cannot use. That is a real risk, and the resolution is not a volume setting but a rule: give each party the level of detail that lets them do the specific job they hold. A sponsor deciding continuation needs the assumption. An interested observer does not.
Decision Framework
A five-step review, run once at initiation and repeated at each major gate.
1. Keep the existing grid. It answers the effort question. Do not replace it.
2. Re-map on exposure and withholding capacity. Same stakeholder list, two new axes. This takes about ninety minutes with the right people and typically relocates a third of the entries.
3. Identify the high-exposure, low-power quadrant explicitly and name it. These parties will not escalate through governance because they have no route into it. Establishing a deliberate channel to them is the single highest-return action in the whole exercise.
4. Apply the opposition test. For every well-informed opponent with material exposure, write down their objection in their own terms, and answer this: if they are right, what changes about the investment case? Where the honest answer is "quite a lot", the finding belongs in front of the sponsor, not in the engagement plan.
5. Set level and timing per party as an explicit decision. For each significant stakeholder, record what decision they hold and therefore what depth of information they require. Where a party holds no decision, they need information sufficient to raise a concern, and no more.
From Strategy to Execution
Immediate. Take the current stakeholder map for the program with the greatest benefit risk and complete step two. The relocations are usually visible within the first twenty minutes, and the exercise costs nothing.
Medium term. Change what governance requires at initiation. A stakeholder map should not be accepted as complete without an exposure reading and a named channel to the high-exposure, low-power group. This is a small addition to a standard artefact and it changes what programs notice.
There is an internal analogue to all of this. The boundaries between the organisation's own functions fail in a comparable way and for comparable reasons, and they are examined in [Related article: The Work Happens Inside Functions. The Value Is Lost Between Them.].
Long term. Build the discipline of reading opposition as evidence. This is cultural rather than procedural, and it is established mainly by example: what happens the first time a senior leader responds to a well-argued objection by reopening the business case instead of commissioning an engagement strategy. The behaviour is also closely connected to whether the organisation can hear anything from below at all, which is examined in [Related article: Whose Knowledge Does Your Governance System Actually Hear?].
Signals to Monitor
- Objections that arrive through a channel other than governance. A concern reaching an executive through a union, a member of parliament, a journalist or a customer complaint is a sign that the high-exposure, low-power group has no legitimate route in.
- Adoption forecasts unsupported by the affected group's own statements. Where a benefit case assumes behaviour change from people who have not been asked whether they will change, the forecast is an assumption wearing a number.
- A stakeholder map unchanged for more than two quarters. Exposure moves as scope and sequencing move. A static map means nobody is reading it.
- Engagement effort concentrated on parties who already agree. Easily measured from meeting records, and a reliable indicator that the map is being used to allocate comfort rather than attention.
- Opposition characterised in terms of the opponent's motives. When the language shifts from what they object to toward why they are being difficult, the organisation has stopped receiving the information.
Questions for the Leadership Team
- On our most exposed program, who carries the greatest consequence — and how much attention are we giving them?
- Who could withhold something this program cannot proceed without, and are they anywhere on our stakeholder map?
- What is our best-informed opponent's actual argument, and has anyone in this room heard it in their words rather than summarised?
- If that argument is correct, what changes in the investment case — and have we tested it?
- For each of our major stakeholders, what decision do they hold, and does the information we send them support that decision or merely reassure them?
- Through what route would someone with high exposure and no organisational power raise a concern with us — and has anyone ever used it?
Closing Perspective
Stakeholder analysis has become a familiar ritual, and familiarity is the reason it is rarely examined. The grid gets drawn, the quadrants get populated, the engagement plan follows, and the organisation believes it has understood its environment when what it has actually done is allocate a communications budget.
The information that determines whether a program succeeds sits in a different reading: who bears the consequence, who can withhold what the program needs, and what the best-informed objection is actually saying. None of that appears on the standard instrument, and none of it is difficult to establish.
The organisations that get surprised late are seldom the ones that failed to communicate. They are the ones that mapped attention, called it understanding, and never asked who would be living with the result.
About the author
Kevin Jogin is Founder & Principal Advisor at EraNorth. Meet the Founder.