MVP scope: how to define what belongs in your first release

By Dmitry Volopianskii — product entrepreneur, consultant and digital product builder. Prepared September 29, 2026.

MVP scope is the boundary of a first product experiment: the customer it serves, the workflow it supports, the conditions it must meet, and the work deliberately deferred. Define the decision you want to make after the pilot, then include the capabilities needed to deliver and observe that workflow. Record exclusions and manual operations alongside the software.

This guide is for a founder preparing a business-to-business software pilot. A minimum viable product, or MVP, is an early product used to test a meaningful assumption with users. Here, the scope includes working software and any human assistance required to make it useful. If the customer problem is still uncertain, first validate the startup idea before building.

On this page: pilot decision · scope example · prioritization · manual work · scope template · change requests · FAQ.

How do you define MVP scope around a pilot decision?

Start with the decision that the first release should inform. “Build a project management platform” is too open to guide trade-offs. “Find out whether small design studios will use a shared approval record for recurring client deliverables” gives the team a specific workflow to investigate.

The next decision might be whether to invest in self-service onboarding, expand to another type of customer, or stop. Write down the evidence you would need for that decision before listing features. Usage alone may be insufficient if the pilot depends on intensive founder support or a buyer who has no intention of continuing.

Describe the workflow from a real trigger to an observable outcome. For an approval tool, that could mean a studio uploads a deliverable, an invited client reviews it, and both parties can see the resulting decision. It includes two roles because the handoff between them is the subject of the experiment.

Keep the pilot boundary explicit: which customers, which file types, which decision states, and which support arrangement are included? A first release can be narrow while still serving everyone needed to complete its core task.

What should an MVP include? A worked B2B scope example

Consider a hypothetical approval tool for small design studios. This is a teaching example, not a customer story or a claim about achieved results. Assume the pilot tests whether a shared decision record helps a studio and its client complete approval without searching through separate message threads.

The product needs to preserve the relationship between the version reviewed and the decision made. Otherwise a client could approve one file while the studio believes another was approved. That dependency is more important to this experiment than a broad collection of collaboration features.

CapabilityFirst-release decisionReason in this pilot
Create a review request with a specific deliverable versionBuildEstablishes what the client is reviewing
Restrict access to intended participantsBuildReview material must not be exposed to unrelated clients
Approve or request changes and retain the decisionBuildCompletes the workflow being tested
Record completion and assistance neededBuild a simple recordProvides evidence for the next investment decision
Set up each studio workspaceFounder-assistedSelf-service acquisition is outside this pilot
Chase an unanswered requestManual, with a recorded interventionAllows operation while revealing support effort
Full annotation canvasDeferA short comment can explain a change request in this example
Accounting integrations and a template marketplaceDeferServe separate workflows that the pilot does not test

This is not a standard checklist for approval software. If precise visual annotation is the customer problem, the canvas moves into scope. If the pilot tests unattended approvals at scale, manual chasing changes the experiment and may no longer be an acceptable substitute.

Write acceptance criteria for the handoff

A feature name is not an acceptance criterion. “Approval” leaves questions about versions, participants, errors and visibility unresolved. For this example, a useful first criterion is: an authorized client can record a decision against the version shown, and the studio can retrieve that same version and decision later.

Then address failure states that matter to the pilot. What happens if a user opens an old link, loses access, or submits a decision while saving fails? Agree on these behaviors before declaring the workflow complete. A postponed convenience and a broken record of the customer's decision have different consequences.

Which MVP feature prioritization method should you use?

Use a method to make assumptions visible. First classify the requirements needed to complete the experiment; then compare optional improvements. Ranking every idea together can obscure dependencies and make an essential capability look dispensable.

Agile Business Consortium's MoSCoW guidance distinguishes Must Have, Should Have, Could Have and Won’t Have this time. A viable workaround can move a requirement out of the Must Have category. Recording what is excluded from the current timeframe also makes the boundary explicit.

For each candidate in this pilot, ask:

  1. Which step or operating constraint does it support?
  2. What evidence would we lose without it?
  3. Can an agreed workaround preserve the experiment?
  4. What other work does this choice depend on?

These questions produce a reason for inclusion. A priority label alone does not.

Where does RICE fit?

Intercom's RICE framework combines reach, impact, confidence and effort. Its confidence factor makes uncertainty in estimates explicit, and its reach guidance favors measured product data when available.

For a new product, you may not yet know how many customers will encounter a feature or how much it will change their behavior. Label those inputs as assumptions. Use the exercise to expose the missing evidence, rather than allowing a precise-looking score to settle the scope. Return to it as observations accumulate.

For the studio tool, version identity remains essential even if only a small pilot group needs it. A popularity ranking would not capture the fact that approvals become ambiguous without it.

Can manual operations be part of an MVP?

Yes, provided the operation is feasible and you understand what it lets you learn. Human assistance can help deliver a service while some automation is deferred. It also limits the conclusions you can draw about independent use, operating effort and scale.

Give each manual task an owner, a record and a review condition. For founder-assisted onboarding, document what information the founder collects, how access is granted and how much help the customer needs. Tell pilot participants what service they are receiving; do not present human work as a proven automation capability.

In the approval example, record whether a decision happened after a founder reminder. An approval achieved with help is useful evidence, but it answers a different question from an approval completed without intervention.

Use a small measurement plan:

ObservationWhat it helps you understandQualification
Review requests started and completedWhether the workflow reaches its outcomeExclude internal test requests and record incomplete ones
Assistance per requestHow much work sits outside the softwareRecord the type of help as well as time spent
Use on the next eligible deliverableWhether customers return when the need recursA customer without another deliverable is not an immediate rejection
Reasons for abandoning or reverting to emailWhat prevents adoptionSeparate access issues, workflow mismatch and missing capabilities
Buyer's decision after the pilotWhether the service warrants continued commitmentA verbal compliment is different from a concrete next step

There is no universal conversion threshold that makes this pilot successful. Agree on the evidence required for the next decision in the context of the business. A small pilot can guide an investment while leaving broader market demand uncertain.

What belongs in an MVP scope template?

A useful MVP scope template records the experiment, the delivery boundary and the decision process. Keep the first page readable by both the founder and the builder; attach technical details where they are needed. Brevity should not hide an unresolved dependency.

FieldWhat to write
Decision after the pilotThe investment or change this release should inform
First customer and user rolesWho participates and who is outside this release
Workflow start and finishTrigger, essential actions and observable outcome
In scopeCapabilities and the dependencies required to deliver them
Acceptance criteriaHow the team will demonstrate the promised behavior
Manual operationsOwner, participant expectations, effort log and fallback
Out of scopeDeferred capability, rationale and evidence needed to revisit it
Operating constraintsAccess boundaries, data, supported environments and recovery needs
Pilot observationsEvents or records, sources, owners and interpretation limits
Decision owner and review pointWho decides, when, and how scope changes are handled

For every unknown dependency, add an action to resolve it. “Connect to the customer's storage” might require a permission model, access approval and a reliable way to identify file versions. A small feasibility test can resolve the uncertainty before the team commits to that integration.

Time and budget belong in this conversation, but a preferred launch date is not evidence that the proposed scope fits. Ask the builder to expose unknowns and trade-offs. For a planning range and an example calculation, see what drives MVP development cost.

To reduce the estimate, remove or simplify a named responsibility and have the builder price the revised scope. In this example, founder-assisted onboarding can defer a self-service setup flow. It does not remove the need to restrict access or preserve the approved version. Record the founder’s recurring effort separately, so a lower build price does not hide an operating cost.

How do you prevent MVP scope creep without ignoring learning?

Assign a decision owner and keep a brief change log. New information can legitimately change the first release; the important question is how the change affects the experiment and the agreed constraints.

Basecamp's Shape Up discussion of when to stop describes scope as something teams must actively adjust as details emerge. It also distinguishes choosing what to build from sacrificing the quality of the work that remains. The application here is to review each addition before treating it as a commitment.

For the approval pilot, a request to support a new file type deserves several checks. Is it necessary for the selected studios' actual deliverables? Does the current format prevent the pilot from operating? Could you recruit a narrower group without changing the question you want to answer? If support is necessary, what does it displace or change in the delivery plan?

Classify the outcome as a required correction, a substitution within the boundary, or a deferred experiment. Keep the reason visible. “Later” should have a revisit condition, such as evidence that the excluded format prevents otherwise suitable customers from completing the core workflow.

Frequently asked questions about MVP scope

How many features should an MVP have?

There is no reliable universal count. Include enough capability to complete the chosen workflow, satisfy its operating constraints and observe the outcome. Count dependencies as part of the work. One feature with a difficult integration can be a larger commitment than several simple actions in a contained workflow.

Is an MVP scope document the same as a product requirements document?

They can overlap. The scope document establishes the boundary of the release and its purpose. A product requirements document may describe behavior and constraints in greater detail. Use as much detail as the team needs to make shared decisions, and avoid maintaining two conflicting descriptions of the same release.

Should authentication and analytics always be included?

Their implementation depends on the pilot. You need an appropriate way to control access where the workflow requires it, and a way to obtain the observations needed for your decision. Those requirements do not automatically imply a custom identity system or a full analytics dashboard. Resolve the actual need first.

When should a deferred feature return to scope?

Revisit it when new evidence connects it to the current decision or the next agreed experiment. A repeated obstacle in the selected workflow may justify inclusion; a request from a different customer segment may justify a separate pilot. Record the reason and the trade-off before adding the work.

Make the first release understandable to everyone building it

A useful MVP scope names the customer, the completed workflow, the operating conditions, the manual work and the next decision. These boundaries give a team a basis for estimating work and responding to new information.

My product work spans EdTech, SaaS, AI and internal tools; the Lectoroom product case study provides background on that work. The approval-tool scenario above is an independent teaching example, not an account of a Lectoroom release or a client outcome.

Sources and editorial approach

Methodological sources checked September 29, 2026: Agile Business Consortium on MoSCoW, Intercom on RICE, and Basecamp on scope trade-offs. The worked scenario, measurement table and worksheet are editorial applications, not empirical findings. No client performance figures are claimed. Future factual corrections should record the change date and supporting evidence.