Illustrative workspace or planning imagery; not a delivered client system or a client-results claim.
You have an idea for a customer portal, an internal business tool or a new digital service. Development proposals are arriving, but the team has not agreed which customer problem matters most or what the first release should leave out. A product discovery workshop can help organise those decisions before they become expensive assumptions inside a build.
The useful purchase is a defined piece of decision-making work. My product discovery and MVP planning service can be scoped around your questions, existing evidence and technical context. Use this guide to prepare the brief and assess what a proposed engagement should produce.
1. Name the decision the workshop must support
Start with a decision that someone can make after the work. For example: which customer group should the first version serve, whether a workflow needs custom software, or which uncertain part needs investigation before development is commissioned. A request to discuss every possible feature is too broad to evaluate against a useful finish line.
Write the decision, its owner and the consequences of postponing it. Keep a proposed solution separate from the problem. In an illustrative supplier-portal project, the problem might be repeated status enquiries; a portal is only one possible response. Better notifications or changes to the existing process might also deserve consideration.
The GOV.UK discovery guidance separates understanding a problem from committing to build a service. It is a useful reference for framing a commercial brief, rather than a timetable your project must copy.
2. Bring evidence and make missing information visible
Prepare a short evidence pack before choosing workshop activities. Include the current user journey, support questions, available analytics, examples of the existing process and known business constraints. Identify where each item came from and how recently it was collected. A sales request, a customer interview and a measured usage pattern answer different questions.
Do not turn preparation into a document collection exercise. Summarise the relevant pattern and link to the underlying material where access is appropriate. Mark gaps explicitly: perhaps nobody has spoken to occasional users, or the available report cannot distinguish a failed task from an abandoned visit.
- What do you already know, and what is still an interpretation?
- Which people are missing from the available evidence?
- Which systems, permissions or agreements constrain possible changes?
- What information can the discovery team access safely?
3. Agree participants and decision rights
Invite people because of the knowledge or authority they contribute. Depending on the project, that may include a business sponsor, a customer-facing colleague, operations, design and the technical owner. Someone must be able to explain how the current process actually works, including the steps that happen outside the website.
Also identify who can approve the proposed scope and who must review specialist constraints. Attendance does not automatically give a participant authority to commit a development budget, change a security requirement or promise another department’s capacity.
I recommend recording three simple responsibilities in the proposal: who supplies evidence, who advises on feasibility and who makes the final product decision. If an essential person cannot attend, agree how their input will be gathered and when their review is required. This prevents a workshop agreement from being overturned later by an absent owner.
4. Separate workshop discussion from user research
An internal discussion can expose disagreement and organise research questions. It cannot establish what customers need merely because colleagues reach a consensus. Ask whether the engagement includes interviews, observation, review of existing research, or only planning for those activities. Confirm recruitment responsibilities and access requirements before assuming research will happen.
GOV.UK’s discovery research guidance focuses on understanding users and the context of their work. For your proposal, translate that into practical questions: who needs to be heard, what you need to learn and which method can supply relevant evidence.
Use public, synthetic or appropriately anonymised examples during initial scoping. Do not attach private customer records to an enquiry. Any later access to research material should be agreed through the company’s normal permissions and information-handling process.
5. Turn assumptions into a prioritised learning plan
A useful discovery record distinguishes facts, assumptions and decisions. For each important assumption, write what would change if it proved wrong. That makes it easier to decide which uncertainty deserves attention before choosing a feature set.
Consider an illustrative assumption that customers will use a self-service status page. Relevant questions include whether customers can identify their reference, whether the underlying status is reliable and whether someone still needs to explain exceptions. Drawing a screen does not resolve all three.
I recommend a small working register containing the assumption, its consequence, the available evidence, a proposed investigation and a named owner. Select the next investigation according to the decision it can inform. Atlassian’s product discovery overview also treats customer understanding and evaluating ideas as part of product work. No particular software tool is required to keep your own register understandable.
6. Define the MVP boundary before compiling the backlog
A minimum viable product brief needs a clear user, a task and a reason to try the first version. A long list of features with priority labels can still conceal disagreement about what the release is meant to demonstrate.
Describe an end-to-end scenario and the minimum support it needs. For the fictional supplier portal, the first scenario might cover checking a known request and understanding its current status. Account administration, historical reporting and multiple integrations might remain separate decisions. Those exclusions need review; they are not automatic recommendations for every portal.
Record included journeys, excluded journeys, dependencies and operational ownership. Explain any manual steps that remain and who will perform them. Ask engineering to assess feasibility and identify unresolved technical work. A workshop brief should not represent an untested estimate as a fixed delivery promise.
7. Agree deliverables and how they will be accepted
Define the outputs in the proposal so you can review the engagement against them. A scoped package might include a problem statement, evidence summary, current journey, assumptions register, proposed MVP boundary and a next-step recommendation. Wireframes or technical investigation may be appropriate, but confirm them explicitly instead of assuming they are included.
Useful review criteria are specific: can the sponsor identify the decisions that were made, can a developer see what remains unresolved, and can the team distinguish approved scope from suggestions? Ask for editable documents and a clear route for correcting factual errors.
It can be reasonable to investigate further, narrow the idea or stop. Discovery does not guarantee a profitable product or prove market demand by itself. The completion discussion should explain the recommendation, its evidence and what still needs validation.
8. Send a brief that can become a scoped proposal
Include your product idea or existing service, the decision you need to make, the intended users, current evidence and the people available to contribute. Mention deadlines that have a real business reason, known integrations and any access limitations. Describe the uncertainty without sending passwords or private research records.
I can use that information to discuss a bounded discovery engagement, its activities, deliverables and review responsibilities. Explore product discovery and MVP planning, then request a proposal around your decision.
If you already have an agreed direction and need continuing ownership of research, priorities and roadmap decisions, an ongoing fractional product management engagement may fit better. Development, ongoing product management and the discovery workshop should each have a clear scope.
