Who can change what the job is for? Drivers, unwritten expectations and involvement that fits the phase

Ask the person running a significant job a simple question: who is allowed to change what this job is for? The first answer is usually the customer, or the owner. Then comes a qualification: certain other people would obviously need to be consulted. Then a longer pause, and a list that grows as they think through everyone whose views have actually changed the job’s direction over the past few months. By the end, the honest answer is much longer than the contract or plan suggests, and most of the people on it have never been told they hold that power.

That is the real mechanism behind what businesses call scope creep, a phrase that conveniently suggests scope moves by itself. It does not. People move it, one reasonable request at a time. And alongside the formal requests sit expectations that nobody wrote down: things people came to assume because of what was said in a meeting, what a prototype looked like or what the team did once as a favour. Those expectations become the real test of whether the job is judged a success.

This article explains how to decide who may redefine a job, how to surface and settle unwritten expectations before they harden into obligations, and why the same person can be essential at one stage and disruptive at another. It is general information for businesses delivering work for customers and for their own internal projects.

Drivers, supporters and observers

Project management writer Stanley Portny sorted the people around a project into three groups:

• Drivers have some say over what results the work is meant to deliver.
• Supporters enable or carry out the work.
• Observers take an interest without doing either.

Portny’s wording was careful: drivers define the results only to some extent. He was describing influence, not formal authority. But the distinction is a useful foundation for a decision that many businesses never make explicitly: who is actually permitted to redefine the outcome?

Four misreadings get in the way:

• Interest and influence are what matter. Those describe how much attention someone needs. Two people can be identical on both and differ entirely on whether their objection can change what the job is for.
• Being consulted and being a driver are different. In practice they converge. When someone’s concerns have repeatedly reshaped a job’s objectives, they are a driver, whatever the paperwork says.
• More drivers means better alignment. It means the opposite. Every extra person who can redefine the result adds another definition of success the job must satisfy. Beyond a few, the job stops having a purpose and becomes a negotiated position that cannot be used to say no to anything.
• Observers are harmless. They become costly when treated as drivers out of courtesy. A senior person with a genuine interest but no accountability for the outcome will naturally offer views. Whether those views change the job depends on whether anyone has said they should not.

Make driver status a decision

Naming drivers is an allocation of authority, and it deserves the same care as any other. A practical method:

1. List everyone whose views have changed the job’s direction, based on what actually happened, not the organisation chart.
2. Sort each person into driver, supporter or observer.
3. Reduce drivers to the smallest defensible number, ideally fewer than five.
4. Test accountability. Does each driver carry consequences if the result is not achieved? If not, they are an observer with influence.
5. Tell people their role. This is the uncomfortable step, and skipping it voids the exercise.
6. Name where disagreements between drivers are settled, by someone with authority to decide, not just discuss.
7. Re-check at each stage.

Driver status should follow accountability for the outcome, not seniority. Naming drivers also means naming, explicitly, that some senior people are not drivers. That is awkward, which is why it is often avoided. But avoiding it does not remove the decision; it hands it to whoever is most persistent.

Expectations become requirements

Jobs rarely fail only against what was written down. They also fail against what people expected. The customer expects the first stage to show something visible. The operations team expects no disruption. Finance assumes a saving will recur. Users expect the new system to be easier. None of this may be in the contract or plan, yet each can become a practical acceptance test.

Expectations come from many places: presentations, early prototypes, informal conversations, past behaviour and, importantly, what you do. If a contractor accommodates small changes over the phone without charge, the customer learns that small changes are free. If a pilot is supported generously by hand, users assume the same support will continue. Ask not only “what have we promised?” but also “what are people learning to expect from how we behave?”

The answer is neither to ignore expectations nor to absorb them all. Treat each material expectation as a claim on the job with one of four outcomes:

• Accept it as part of the agreed result.
• Clarify it, because it is ambiguous.
• Negotiate a different outcome or condition.
• Decline it explicitly, because it is outside the job’s purpose.

Silence is a fifth outcome and usually the most dangerous, because different people go on working to different definitions of success.

A useful warning sign is a shift in language. “We discussed” becomes “we agreed”. “It would be nice if” becomes “the job will”. “The trial included” becomes “the service requires”. These shifts show an informal expectation turning into a perceived commitment. The when a conversation becomes a commitment article covers how informal exchanges can create real obligations.

Keep a short expectation register

A simple register keeps expectations visible. For each material expectation, record:

Review it at each stage meeting. Most entries settle quickly once they are written down; the value lies in the few that would otherwise have surfaced at handover as a dispute.

The same applies inside your own business

Internal projects have drivers, observers and unwritten expectations too. A new system, a new site or a reorganisation will attract views from people who are interested but not accountable, and expectations will form from what managers say in passing. Name the drivers for internal work as deliberately as you would for a customer job, and give staff a clear route to raise what they expect, so it can be accepted or declined while there is still time to plan for it.

The right person at the wrong time

Research by Christina Beringer, Daniel Jonas and Alexander Kock, published in 2013 and covering nearly two hundred project portfolios, found that the effect of stakeholder involvement depended on the phase. Senior management involvement in strategic planning and selection was valuable; heavy senior involvement in day-to-day steering was associated with worse results in established portfolios. Role clarity also mattered: where roles were unclear, managers who controlled resources could protect their own priorities at the expense of the whole.

For a small business, the lesson is practical. An owner or customer who is deeply involved in deciding what a job should achieve is doing exactly the right thing. The same person stepping into daily site or delivery decisions can slow everything, teach the team to escalate instead of resolve and pull attention away from the decisions only they can make.

For each important person, ask four questions:

• Which phase do they need to be involved in?
• Which decisions do they own?
• What evidence should they see before stepping in?
• What authority do they hold, and what is outside it?

Influence moves

A list of stakeholders is a snapshot. Influence changes as the job moves. A supplier becomes critical when it is the only source of a key part. A landlord or council becomes central when an approval is needed. Operations gains an effective veto as the change approaches go-live. Customers gain influence when their reaction determines whether the benefits arrive.

So review the picture at each stage, asking:

• Who now has more or less ability to affect the outcome?
• Whose interests or exposure have changed?
• Which decisions are coming up, and who holds the authority or knowledge for them?
• Who has information we need but do not have?

The stakeholders you do not control article covers weighing exposure and knowledge, not just power, and treating resistance as information.

A worked example

This is an illustration. A small design-and-construct firm is fitting out a new dental practice owned by three partner dentists, with a practice manager running operations. The contract names the managing partner as the client’s representative.

Three months in, the firm counts 11 change requests from five different people, four of which contradict each other. The other two partners have been visiting the site and asking the site supervisor directly for changes to cabinetry and lighting. The supervisor, keen to keep the customer happy, made several small changes without recording them as variations.

The firm’s director meets the managing partner and practice manager to sort roles:

• Drivers: the managing partner, for scope and budget; the practice manager, for operational requirements such as the sterilisation room workflow.
• Observers: the other two partners and the practice staff, whose views are welcome through the practice manager.
• Supporters: the firm’s team and its suppliers.

All changes now go through the practice manager to the firm’s project lead, and the managing partner approves anything affecting cost or time. The firm also recognises its own role in creating an expectation: by making small changes for free, it had taught the client that changes cost nothing. It explains the new process politely and backs it with a simple variation form.

The firm then lists unwritten expectations. One stands out: the practice manager had told the IT installer the reception area would be ready on 1 May, a date the firm had never agreed. The firm clarifies the real date in writing and agrees a partial handover of the reception area so the IT work can start a week later.

Later, a fire services modification needs the landlord’s approval. The landlord, previously a background party, becomes critical. The firm adds the landlord’s building manager to its weekly update for the following month.

How this applies to a small Australian business

• Ask who can change what the job is for, and write down the answer.
• Keep drivers few and tie their role to accountability.
• Tell people their role, kindly and clearly.
• Route changes through drivers, with a simple variation process.
• List unwritten expectations and accept, clarify, negotiate or decline each.
• Watch your own behaviour for expectations it creates.
• Match involvement to the phase.
• Review who matters at each stage.

Signals worth watching

• Change requests arriving from many directions.
• Contradictory requests from people on the same side.
• “We agreed” used about things that were only discussed.
• Favours that have quietly become expectations.
• Senior people stepping into routine delivery decisions.
• A party that suddenly matters because an approval is due.

Common mistakes

• Treating every interested person as a driver.
• Avoiding the conversation about who decides.
• Absorbing expectations without deciding on them.
• Making free changes that teach customers changes are free.
• Keeping the same stakeholder picture for the whole job.
• Confusing executive accountability with constant involvement.

Frequently asked questions

Won’t telling people they are observers offend them? It can be done respectfully: their views are welcome, and here is how to make sure they are heard. Most people prefer clarity to being consulted and then ignored.

What if the customer has several decision-makers? Ask them to name one representative for scope and cost, and agree how they resolve disagreements among themselves.

How do we decline an expectation without damaging the relationship? Explain the job’s purpose and what accepting it would cost in time or money, and offer an alternative where possible.

Should every expectation be written down? Only material ones: those that could affect cost, time, quality, risk or whether the customer judges the job a success.

How often should we review who matters? At each stage of the job, and whenever an approval, supplier change or major decision is coming.

Questions to ask

• Who has actually changed the direction of this job so far?
• Who is permitted to, and do they know?
• What do people expect that we have never agreed?
• What have our own actions taught customers to expect?
• Is anyone involved in decisions at the wrong stage?
• Who will matter more at the next stage than they do now?

Bringing it together

Scope does not creep by itself; people move it. Decide who may redefine the job, keep that group small and tied to accountability, and tell everyone their role. Surface the expectations that formal documents do not capture, including those your own behaviour creates, and accept, clarify, negotiate or decline each one rather than leaving them to silence. Involve people in the phases and decisions where they add value, and refresh your picture of who matters as the job moves. Clear roles and named expectations protect the purpose of the work and the relationship with the people it serves.

---

Source: KEVOS notes, drawing on S. Portny’s classification of project drivers, supporters and observers, E. Baker (2012) on stakeholder expectations, teaching material on program and portfolio stakeholder engagement, and C. Beringer, D. Jonas and A. Kock, “Behavior of internal stakeholders in project portfolio management and its impact on success”, International Journal of Project Management (2013). Examples in this article are illustrations. This article is general information.