Delivery & Governance · 10 min read
Project Governance for Small Teams: Just Enough Structure
Governance sounds heavy, but a small team needs five things: clear roles, a decision rhythm, decision rights, simple gates and a record of choices.
Dennis AlejoBusiness consultant · Project & operations manager · · 2,228 words
The word "governance" makes many people picture committees, forms and slow approvals. It sounds like something that belongs in banks and government agencies, and that a five-person company should avoid. But every team already has governance, whether it admits it or not. Someone decides what gets done first. Someone approves the spending. Someone says when a piece of work is good enough. The question is only whether those decisions are made on purpose, in a way everyone understands, or by accident, in a way everyone argues about.
This article is about designing just enough governance for small teams and modest projects. The goal is not to add process. It is to remove the arguments, delays and surprises that come from unclear decisions. You will find five essentials, a plain-English explanation of RACI, a way to design gates that include the courage to stop and a short check to see where your own project stands.
What governance is, and what it is not
Governance answers four questions. Who is accountable for the result? Who decides what, and with what information? When do we check whether we should continue? And how do we keep a record so that decisions are transparent and can be learned from?
Notice what is not in that list: reports for their own sake, meetings to prove that people are busy or approval layers added out of fear. Those are symptoms of governance gone wrong. Good governance makes work faster, because people know where decisions live and do not wait for permission that they already have. It also protects the people doing the work, because the rules are known beforehand and are not invented after something goes wrong.
The cost of too little, and too much
Picture a dial. At one end is no governance. Symptoms include decisions that are repeated, because nobody remembers what was agreed; work that stalls waiting for an unnamed approver; scope that changes by whoever shouted last; and projects that nobody dares to cancel. At the other end is heavy governance: every spending decision needs three signatures, every change goes to a committee and reports take longer to write than the work they describe. Symptoms include slow progress, hidden work done to avoid the process and talented people who stop proposing ideas.
The right setting depends on the size, the risk and the novelty of the project. A one-week internal experiment needs almost none. A multimillion-dollar infrastructure program needs a great deal. Most small teams belong near the lighter end, with a few well-chosen controls where the stakes are highest.
The five essentials
A small team can cover almost everything with five elements.
- Roles: who is the sponsor, the owner, the project manager and the contributors.
- Rhythm: when the team reviews, when the sponsor reviews and when the project is formally checked.
- Decision rights: who decides what, and up to what threshold.
- Gates: a few checkpoints with criteria to continue, change or stop.
- Records: a short log of decisions, risks and changes.
Let me take them one at a time, because each is simpler than it sounds.
Roles, and RACI in plain English
Clear roles remove most governance friction. Four are common in small projects. The sponsor owns the business case and removes obstacles; I describe the role in keeping your executive sponsor engaged. The project owner or manager runs the work day to day. The contributors do the work. And the stakeholders are everyone else affected, as mapped in stakeholder mapping.
A well-known tool for clarifying who does what is the RACI chart. For each task or decision, assign each person one of four letters. R means Responsible: they do the work. A means Accountable: they own the outcome and have the final say. C means Consulted: they are asked for input before a decision. I means Informed: they are told afterward. The golden rule is that every task or decision has exactly one A. If there are two accountable people, there are none. If there is no A, the decision will drift.
You do not need a large chart. Take the five or six most important recurring decisions in your project and write the letters next to the names. Five minutes of clarity at the beginning saves hours of confusion later.
Decision rights
The second source of friction is not knowing how far your authority goes. A simple table can fix that. List the types of decisions and who can make them, with thresholds where it matters.
- Spending within the approved plan: the project manager, up to an agreed amount.
- Spending above that amount, or any use of contingency: the sponsor.
- Changes to scope, time or budget: the sponsor, following the change process in how to stop scope creep.
- Day-to-day task assignment: the project manager and the team.
- Hiring, firing and contracts: according to company policy, with the sponsor informed.
- Stopping or restarting the project: the sponsor or a small steering group.
The exact numbers are yours to choose, and they should reflect your level of trust and risk. The point is that everyone knows them in advance, and nobody needs to ask permission for what they already own.
Gates: decisions with criteria, including stopping
A gate is a planned checkpoint where the project pauses long enough to decide: continue, change course or stop. Gates are powerful because they turn a vague "how is it going?" into a structured question with agreed criteria. They also give you the opportunity to cancel a project that has stopped making sense, which is one of the most valuable and least used options in business.
A useful gate has three parts: the question being decided, the criteria for a yes and the evidence that will be reviewed. For example, at the end of a design phase, the question might be "do we begin construction?" The criteria could include an approved budget, required permits in hand and a risk review completed. If the criteria are not met, the answer is "not yet," and the team knows exactly what is missing.
Gates look different in different projects. In Trade Arena, a trading strategy has to pass several increasingly demanding tests before any real money is used, and each gate has a clear pass or fail. In an exploration project such as Llama Purisima, the work is staged so that each phase ends with a decision based on the evidence gathered, and the next investment is made only if the results justify it. In a construction project, gates follow design, permits, structure and handover. In software, they follow prototype, pilot and release. The pattern is universal: spend a little, learn, decide, then spend more.
Do not forget kill criteria. When you start a project, write down what would make you stop. "If after the pilot fewer than 20 qualified inquiries come in, we will stop and rethink." It is much easier to agree on stop conditions at the start, when emotions are low, than in the middle, when everyone is invested.
Rhythm: the heartbeat of governance
A project's rhythm is a small set of regular meetings with clear purposes. For a small team, three levels are usually enough.
The weekly team review takes thirty minutes and looks at a few numbers, exceptions and decisions, as I describe in data-driven project management. The monthly sponsor or steering review takes up to an hour and covers progress against the plan, the top risks, changes requested and decisions needed. The gate review happens at the end of each phase and is a more formal decision.
Keep the agendas fixed and short. Send a one-page summary beforehand. Close each meeting by reading the decisions aloud and writing them in the log. If you adopt only one habit from this article, make it that last one.
The decision log
A decision log is a simple list that records what was decided, when, by whom and why. A table with five columns is enough: date, decision, owner, reasoning and review date. It takes a minute to update and saves hours of rehashing. When a new person joins, they can see why things are the way they are. When someone says "I thought we agreed otherwise," you can check. And when a decision turns out to have been wrong, you can learn from the reasoning instead of the blame.
Try the quick check below to see how complete your governance is. Tick only what is already true.
Scale the structure to the project
One size does not fit all. A helpful rule is to scale governance with three factors: how much money is at stake, how uncertain the work is and how many people it affects.
For a small project, such as an internal improvement of a few weeks, a one-page plan, a named owner, a weekly check and a decision log are plenty. For a medium project, such as a new location or a product launch, add a monthly sponsor review, a risk register as in risk management for small business, a change process and two or three gates. For a large project, such as a major construction program, add formal change control, independent verification of progress, more detailed reporting and stricter approval limits.
The mistake is to use the large-project template on small work, or the small-project habit on large work. Ask at the start: what is the lightest structure that still lets us sleep at night?
Governance across several projects
When you run several projects at once, a different question appears: which ones deserve your limited time and people? A one-page portfolio list can help. For each project, list the owner, the sponsor, the status, the three numbers that matter and the next decision date. Review it monthly, and be willing to pause or stop projects to make room for those that matter more. This is the logic behind choosing only three projects in the 90-day sprint, and it connects to the question of limited capacity that I discuss in capacity planning and bottlenecks. It is also the heart of Project Focus: fewer projects, owned and finished.
Governance when several parties are involved
Projects with partners, investors or joint ventures add a layer: decisions are shared across organizations. Here governance should be written into the agreement. Who sits on the joint committee? What decisions need unanimous consent and which need a majority? How are disputes escalated? How often are results reported? A one-page governance schedule attached to the agreement avoids many painful conversations. I treat partnerships as projects in my business partnerships service, for exactly this reason.
A starter kit you can set up in one hour
If you want to move from reading to doing, here is a one-hour agenda for a team that has never written down its governance. It is deliberately modest, and you can run it with just the people who lead the project.
In the first ten minutes, agree on the purpose of the project in one sentence and name the sponsor, the owner and the contributors. In the next fifteen, write the six most important recurring decisions on a whiteboard and place the RACI letters next to the names. Where you find two accountable people or none, stop and settle it. In the following ten minutes, set the decision rights: the spending limit of the project manager, what needs the sponsor and what counts as a change.
Then spend fifteen minutes defining two or three gates. For each, write the question, the criteria and the date. Add one kill criterion: the result that would make you stop. Finally, use the last ten minutes to fix the rhythm: the day and time of the weekly review, the date of the first monthly sponsor review and where the decision log will live.
Afterward, write all of it on one page, share it and put the first review in everyone's calendar. You will leave with something most teams lack: a short, shared document that answers "who decides this?" before anyone has to ask. Revisit it after the first month and prune what nobody used. Governance should shrink to the essentials and grow only where the project demands.
Common mistakes
Watch for these patterns. Governance by committee: too many people with a vote, so decisions drift. Gates without criteria: a meeting that always ends with "looks good, continue." Missing stop conditions: nobody remembers how to end a project. Unwritten decisions: agreements that live in memory and are remembered differently. Process for its own sake: forms that nobody reads. And governance only at the start: a charter written in week one and never revisited as the project changes.
Signs that it is working
Good governance is quiet. Decisions are made quickly and by the right people. Surprises are rare, and when they come, they are handled calmly. New team members can find the reasons behind past choices. Sponsors do not need to chase updates. And projects that should stop do stop, while those that continue do so with open eyes.
Governance for a small team is not a document. It is a handful of habits that make decisions visible, fast and fair. If you would like help designing a light structure for your team or your portfolio of projects, it is a core part of my scaling strategy and startup advisory work.
Which decision in your team is always unclear: who actually makes it? Tell me in the comments.
Frequently asked questions
What is project governance?
It is the way decisions are made on a project: who is accountable, who decides what, when the project is checked and how decisions are recorded. Good governance speeds work up by removing confusion, and it is not the same as bureaucracy.
What is a RACI chart?
RACI assigns each person one of four roles for a task or decision: Responsible (does the work), Accountable (owns the outcome and decides), Consulted (gives input beforehand) and Informed (told afterward). Each decision should have exactly one accountable person.
What is a project gate?
A gate is a planned checkpoint at the end of a phase where the project is formally reviewed against written criteria and a decision is made to continue, change course or stop.
How much governance does a small project need?
Usually a named owner, a one-page plan, a weekly review, a short decision log and a sponsor who knows the thresholds for approvals. Add risk, change control and gates as the size, risk and uncertainty grow.
Join the conversation
Did this help? A quick reaction tells me what to write next.
Be the first to react.
Loading comments…