Stakeholder Analysis for Software Projects: A Template

ยท 8 min read

A stakeholder analysis for a software project is a short table of the people who can shape, block, use or pay for the product, with what each of them needs and how much say they have. You build it in the first days of discovery, use it to plan interviews and sign-off, and keep it up to date until handover. This article gives you the columns, a power and interest grid, and an invented example you can copy.

Why discovery starts with people

Most of what a discovery phase has to find out lives in people's heads, not in documents. The sponsor knows the budget limit. The operations lead knows the workaround nobody wrote down. The security reviewer knows the rule that will stop a launch. If you do not know who these people are in week one, you find out about them in month four, when their objection costs a redesign.

A stakeholder list does three jobs in discovery:

If you are new to the wider process, the overview of the discovery phase in software development explains where this step sits among the others.

Who counts as a stakeholder

Anyone who can change the outcome, or whose work the outcome changes. On a software project that usually means more people than the client's contact list suggests. Go through these groups and write down names, not departments:

  1. Sponsor and budget holder. Who pays, and who can cancel the project.
  2. Product owner or day-to-day decision maker. Often not the sponsor.
  3. End users. Split them by role: the field technician and the dispatcher use the same system very differently.
  4. People who run it afterwards. Support, IT operations, whoever answers the phone when it breaks.
  5. Owners of systems you integrate with. The CRM admin, the finance system owner, the vendor's account manager.
  6. Gatekeepers. Security, legal, data protection, procurement, accessibility. They rarely ask for features; they say no late if nobody asked them early.
  7. The delivery team. The tech lead and designer on your side are stakeholders too: they own the estimate.
  8. Outside parties. Regulators, partners, the client's own customers, where they are affected.

A useful test for each candidate: "If this person saw the finished product for the first time at launch, could they stop it or make it fail?" If yes, they belong on the list.

What to record for each person

Keep the columns few enough that the table gets filled in. These are the ones that earn their place in software work:

Column What goes in it Why it matters
Name and title The person, not the team You interview people
Organization Client, vendor, partner, your team Shows who you can reach directly
Role in the project Decides, advises, uses, runs, gatekeeps Sets how you involve them
Power High or low: can they change scope, budget or the go-live date? Drives the grid below
Interest High or low: how much does the outcome affect their work? Drives the grid below
What they need One sentence in their words Becomes goals and requirements
What they decide or sign off Budget, scope, security review, data model Your approval path
Concerns or position Supportive, neutral, sceptical, and why Shows conflicts early
Contact and best channel Email, chat, a weekly call Saves chasing

If you want a shorter starting point, the stakeholders section of our discovery phase template asks for the essentials and says when the section is done:

Stakeholders section of the discovery phase template: name, role and organization, what they need, influence and how to reach them, open disagreements, and a done-when line

Two cautions. First, power and interest are your judgement, not a fact about the person, so keep the table among the team and word the notes as you would want them read aloud. Second, a stakeholder list holds personal data (names, roles, contact details, opinions). Store it where the project's other confidential material lives and share it on the same terms.

Power and interest: a grid you can use

The most common way to sort the list is a two-by-two grid of power against interest, usually credited to Aubrey Mendelow's 1991 work on stakeholder mapping. The Nielsen Norman Group's guide to stakeholder analysis describes the same four groups. In discovery terms:

Low interest High interest
High power Keep satisfied. Short updates, ask for their decision only where it is theirs to make. Manage closely. Interview early, review drafts, sign off.
Low power Monitor. Inform at milestones. Keep informed. Interview for detail; they often know how the work is really done.

The grid is a starting point, not a verdict. Two things change it during discovery:

Using the list: interviews, sign-off, communication

Once the table exists, turn it into three plans.

Interviews. Book everyone in "manage closely" and a sample of "keep informed" first, usually the sponsor, the decision maker, two or three end users per role, and each integration owner. Send each person the question you most need them to answer, so the session is not spent on background. Record which answers were confirmed by more than one person and which rest on a single voice.

Sign-off. For each deliverable, name the one person who approves it. Common splits: the sponsor approves goals and the budget range, the product owner approves requirements and priorities, the tech lead and the systems owners approve the architecture, and gatekeepers approve what touches their area. Write it down before the work starts; an approval path invented at the end invites a second round.

Communication. Match the channel to the box in the grid: a weekly call for "manage closely", a short written update at each milestone for "keep satisfied" and "keep informed", nothing more than the final report for "monitor".

Disagreements found along the way go to the open questions and risks log with an owner and a date, not into the requirements as a compromise nobody agreed to.

A filled-in example (invented)

The project below is made up: a mid-sized utilities company replacing its paper-based field inspection process with a mobile app.

Name and title Role Power / interest What they need Signs off Position
Dana Ortiz, Head of Field Operations Sponsor High / high Fewer repeat visits, faster reports Budget range, go-live Supportive
Sam Patel, Operations Manager Decision maker High / high One tool that works offline Requirements, priorities Supportive, worried about training time
Field technicians (3 interviewed) End users Low / high Forms that fit a phone in gloves Usability test results Sceptical after a failed pilot
Lee Chen, IT Security Lead Gatekeeper High / low Device management, data stays in region Security review Neutral, needs early sight of the data flow
Morgan Reyes, Finance Systems Owner Integration owner Low / high Inspection results feed billing without retyping Interface spec Neutral
Alex Kim, Tech Lead (delivery team) Delivery High / high Clear integrations before estimating Architecture, estimate Supportive

Read across a row and the next steps write themselves: the security lead needs the data-flow diagram before the architecture is final, the technicians' scepticism is a risk to record, and the finance owner's need becomes a requirement with an acceptance criterion.

In a finished report the list gets shorter: a name, a title and one line on what each person owns or decides. This is how it looks in our example report, written for an invented project called Fieldwise (all names are made up):

Stakeholders section of the example discovery report for the invented Fieldwise project: five people, each with a title and one line on what they own or decide

In Discovery Phase AI, the Stakeholders stage keeps each person's name, title, contact details and a free-text description for influence and involvement, and the AI can draft that list from uploaded meeting notes or decks for you to review before anything is saved.