
A minimum viable product is a limited offering designed to produce useful learning through customer experience. Its scope follows a specific business question: whether a defined customer will use a proposed solution, obtain the intended result, or make a commercial commitment. The minimum refers to the work required to examine that question meaningfully.
An MVP can be software, a physical product, or a service. It can contain manual work behind the scenes and serve a small customer group. Its distinguishing feature is the relationship between the offering and the evidence it is intended to produce. A small product without a defined learning purpose is simply a small product.
Eric Ries’s Lean Startup methodology connects the MVP to a build-measure-learn cycle. The offering creates an experience, measurement records what happens, and the results inform the next development decision. This makes customer behavior part of product development rather than an activity postponed until launch.
Starting With the Assumption Being Tested
A business idea contains several assumptions. Customers may experience a particular problem, understand a proposed solution, use it repeatedly, pay a stated price, and accept the delivery method. Different experiments examine different assumptions. An MVP’s design becomes clearer when one of those questions has priority.
If the question concerns whether customers can complete a task, a working interaction may be essential. If the question concerns the value of a report, a manually produced report may be sufficient. If the question concerns purchase behavior, the experiment includes a specific offer and commercial terms.
The result can be defined before the offering is built. A team might examine whether a target customer completes an appointment-booking task without assistance or returns to a service for another reporting cycle. These are observable events tied to the product’s intended function.
A learning objective also defines what can be left outside the first version. Features unrelated to the test can remain deferred. The team still preserves the conditions necessary for a meaningful experience, including clear communication and reliable delivery of the limited function being offered.
Proof of Concept, Prototype, MVP, and Pilot
A proof of concept examines whether a mechanism or technical approach can work. It may operate entirely within a laboratory or development environment. Its audience can be engineers rather than prospective customers, and its result concerns feasibility.
A prototype represents aspects of a proposed product. It can be a physical model, an interactive interface, or a partially functioning system. Prototypes can support research about appearance, interaction, fit, or performance without being offered as complete commercial products.
An MVP places a limited offering into a setting that supports learning about the business proposition. A pilot is a bounded deployment or engagement, often with a specific customer and agreed scope. A pilot may use an MVP, but it can also evaluate an already mature product in a new organization.
The distinction affects interpretation. A prototype test can show that participants understand an interface. A paid pilot can show that a customer accepts a defined commercial engagement. The evidence follows the activity actually performed and the question that activity was designed to examine.
A Minimum Viable Service
Service businesses can begin with a narrow, repeatable outcome. A reporting service might deliver one defined report for one type of customer. A maintenance business might offer a limited inspection covering specified equipment. A professional adviser might provide a fixed-scope diagnostic engagement.
The scope identifies what is included, what the customer supplies, how delivery occurs, and what result is returned. Those boundaries allow the business to observe both customer value and the work required to provide it. The first offering becomes a practical unit of delivery.
Manual work can reveal the decisions a later system would need to support. An analyst preparing reports personally may discover which inputs require clarification, which outputs customers use, and which exceptions consume time. That knowledge can inform automation and service design.
The labor involved remains part of the experiment’s record. A service that customers value may require more work than the eventual price supports. Recording delivery time and direct costs provides information about the operating model alongside customer response.
Software MVPs and Complete Tasks
A software MVP can focus on one complete customer task rather than several incomplete functions. A scheduling tool might support selecting an available time, confirming the booking, and receiving a usable record. Completing that sequence creates a coherent experience that can be observed.
The implementation can remain limited in other respects. Administrators may perform some setup manually, and the system may support only a defined customer segment. These choices keep development aligned with the experiment’s scope without changing the outcome promised to participants.
Usability affects the evidence. If customers cannot understand how to perform the task, abandonment can reflect the interaction rather than the value of the outcome. Observing where participants stop helps distinguish those explanations. The measurement becomes more informative when it records the sequence of behavior.
The same principle applies to accessibility. A test that excludes the normal interaction methods used by its intended customers examines an incomplete experience. Input methods, readable content, and compatibility with assistive technology can be part of the limited task itself.
Physical Products Require Application-Specific Tests
A physical-product MVP can use a constrained application, a small production run, or a supervised deployment. The scope depends on the product’s function and the conditions of use. A desk accessory and an industrial control component present different testing and delivery situations.
Technical evidence and customer evidence can be organized separately. A device may perform its measured function, while customers evaluate installation, handling, or the usefulness of the result. Both matter to the offering, but they answer different questions.
NASA’s technology readiness framework provides a technical vocabulary for demonstrations under progressively more representative conditions. An MVP adds a commercial learning objective around the customer experience. The two approaches can be used together without treating them as the same assessment.
For a hardware startup, the limited offering can also reveal service requirements. Customers may need instructions, replacement parts, calibration, or help interpreting an output. Those activities are part of the delivered experience even when the experiment centers on one physical function.
Defining the Customer and the Test Setting
The test population affects what the results describe. Positive technical users may accept setup work that a mainstream customer would not. Existing personal contacts may provide extensive feedback because of their relationship with the founder. Recording participant characteristics gives the results a clear scope.
The setting matters as well. A guided demonstration differs from ordinary use without the development team present. A short trial may examine initial usability, while repeated use can examine whether the offering fits a recurring task. The experiment can select the setting that matches its immediate question.
Customer discovery helps define that choice. The NSF I-Corps program uses direct market investigation to connect inventions with potential applications. In MVP development, similar conversations establish which task and customer circumstances the first offering is intended to represent.
New Space Economy’s discussion of demand-led product development provides related context. The initial offering follows a specific customer situation, which gives the experiment a practical purpose and a defined audience.
Measuring Behavior and Delivery
Useful measures correspond to the intended customer outcome. They may include completed tasks, repeat use, time to completion, requests for another engagement, or payment for a defined service. The measure is selected because it describes the business question, not because it is easy to collect.
The denominator matters. Ten completed tasks have a different meaning among 12 attempts than among 1,000 attempts. A record of invitations, starts, completions, and repeat activity explains where the experience succeeds and where participants leave it.
Qualitative observation helps interpret the numbers. A customer may stop because the product failed, because the task was no longer necessary, or because another person had to approve the next step. The same recorded event can have different commercial meanings.
Delivery measures describe the supplier’s side of the experience. Staff time, rework, support contacts, and direct expenses show what it took to provide the outcome. Combining customer and delivery evidence creates a more complete account of the limited offering.
Paid and Unpaid Experiments
A paid experiment introduces a commercial decision. The customer compares the offer with its price and terms, then commits funds. The resulting evidence includes both the use of the service and acceptance of that particular transaction.
An unpaid test can be appropriate when the immediate question concerns interaction, technical fit, or a task the participant cannot yet perform reliably. Its value lies in the question it answers. A free evaluation and a paid engagement produce different information because the customer makes different commitments.
In either case, the offer’s status can be stated clearly. Participants can understand the scope, duration, delivery conditions, and support available. This makes the experience interpretable and provides a shared basis for the relationship.
The test can also specify what happens when its period ends. A customer may receive a final result, return equipment, or consider a separate commercial agreement. The endpoint prevents a limited experiment from becoming an undefined service obligation.
## A Worked Service Experiment
A hypothetical startup offering inventory analysis could begin with one report for independent retailers. The customer supplies a defined export of sales and stock records, and the startup returns an agreed analysis for a fixed fee. The first version can be produced manually with existing tools.
The experiment can examine whether the report supports a specific purchasing decision. The team records whether the customer supplies usable data, understands the output, and uses it during its normal ordering process. It also records preparation time and questions raised during delivery. These observations concern both the customer’s experience and the supplier’s work.
The initial offering does not require a dashboard, automated integration, or support for every retail category. Those features become relevant when they address an observed part of the recurring task. If customers repeatedly request the same clarification, the next version can change the report or its accompanying explanation.
A later version might automate data preparation while retaining expert review. Another might add a recurring delivery schedule. Each change corresponds to learning from the earlier engagement. The MVP develops through evidence about a defined task rather than through the accumulation of unrelated features.
The commercial record can also distinguish the fee for the first report from the price of an ongoing service. This preserves the conditions of each transaction and makes repeated purchasing visible as a separate result.
Learning Changes the Next Version
Results can lead to changes in the target customer, the outcome, the interaction, the price, or the delivery process. These are different adjustments. Recording which assumption changed preserves the reason for the next version and makes later comparisons possible.
A team can retain the parts of the offering that produced useful results and revise the parts connected to observed difficulties. The next test then examines a more specific proposition. Development accumulates through a sequence of decisions supported by customer experience.
An MVP has completed its immediate purpose when it has produced the evidence needed for the defined decision. The next offering can expand its scope, improve repeatability, or serve another test. The practical result is a product or service whose development history is connected to observed use and commercial behavior.
As an Amazon Associate we earn from qualifying purchases.
Useful Books Available on Amazon