A previous article described how a scrum team could measure the value delivered by completed product backlog items. The approach is based on creating a product hypothesis for each item, describing how the value will be measured. By implementing an item (user story) and then measuring the results, the team is conducting an experiment to validate their beliefs about the value of the item.
A minimal viable product (MVP) allows us to implement less than the full feature, and still gather data about how valuable users find it. While an MVP is often used to test an entire product idea, the approach can also be used to test new feature areas of an existing product. An MVP is a simple implementation that allows us to gather real data from our users.
In the previous article, the team implemented a business center and then measured how much customers used it. You may recall that the user story:
As a businessperson,
I want a self-serve business center with printing, copying, and faxing capabilities,
so that I can get a little work done without leaving the resort.
This will be confirmed when at least ten guests do printing, copying, or faxing each week during the first month that the self-serve business center is open.
Instead of implementing the full business center with computers, furniture, and its own room, we could implement an MVP. In this case, an MVP might be accomplished by purchasing a print/copy/fax machine and putting it at the resort’s front desk. We could put flyers in guests’ rooms notifying them that this free service is available. Now we can test our product hypothesis as follows:
If a month goes by and the machine is getting no use, that might lead us to reassess our assumption that we need a business center at the resort. Maybe we should build a swim-up bar instead.As a businessperson,
I want a business center with printing, copying, and faxing available,
so that I can get a little work done without leaving the resort.
This will be confirmed when at least ten guests do printing, copying, or faxing each week during the first month that the free service is available at the front desk.
With an MVP approach, we can test minimal builds of our features and iterate based on what we learn. This can go a long way toward avoiding overinvesting in feature areas that won’t be valued by our users.
Types Of MVPs
Wizard Of Oz MVP
The Wizard Of Oz approach involves implementing the back-end of the feature manually at first. The users think they are interacting with a fully implemented product, but in reality parts of the functionality are being done via low-tech means behind the scenes.
More About Wizard Of Oz MVP
- Michael Flowers on MVPs
- Product Experiment: Wizard Of Oz
- Video: The Wizard Of Oz Minimum Viable Product
Concierge MVP
A Concierge MVP is similar to a Wizard Of Oz except that there is no pretense of having a fully automated system. The concierge approach involves giving people the value that our product will deliver in a completely manual, high-touch way. Rent The Runway tested their online dress rental concept with pop-up shops where people could view and rent designer dresses.
Learn More About Concierge MVP
Piecemeal / Mashup MVP
A piecemeal or mashup MVP involves building the first version of a product or feature by cobbling together existing services. The initial system likely won’t scale or be able to support the full feature set that is ultimately envisioned, but it can be built quickly. This allows us to test our product hypotheses without investing in expensive custom development.
Learn More About Piecemeal / Mashup MVP
Single-feature MVP
Before adding a new feature set or wide-ranging capability to your product, you might want to pick one small part of the feature set to implement as a way to test your belief that the larger feature set is valuable.
Learn More About Single-Feature MVP
- Single Feature MVP By Michael R. Flowers
- What Can You Learn From A Single Feature MVP?
- Product Experiment: Single Feature
Landing Page MVP
Landing pages are simple websites that can be used to judge the interest in a product idea. Advertisement and organic search results direct interested people to the landing page. Based on the number of people who land on the page, and perhaps interact with it, we can judge the level of interest in the new product idea.
Learn More About Landing Page MVP
- Product Experiment: Spoof Landing Page
- How to successfully validate your idea with a Landing Page MVP
- Landing Page MVP: How to Get Insightful User Feedback on the Product You Haven’t Built Yet
Feature Stub / Fake Door MVP
With a feature stub, or fake door MVP we provide a way for users to access a feature that doesn’t yet exist. For example, if we think people will want an AI grammar check feature, we could test that belief by adding a button that says “Click here if you’d like to run an AI grammar check.” When users click on the button, they get a message that explains that the feature isn’t currently available. By tracking how many users click on the button, we can judge the interest in the feature before we build it.
Learn More About Feature Stub / Fake Door MVP