16

Minimum viable product (MVP)

The minimum viable product (MVP) is a concept (first coined and defined by Frank Robinson within software development) that helps to identify and scope the minimum viable activity that initially can prove, and later deliver, the benefits required to achieve a viable solution.

A key component of an MVP is that while it is minimal, it is also viable; the solution must be viable in that it adequately represents the value. This may be the overall goal, a specific feature or component where more insight is needed. An MVP is not a cheaper or diluted version of the final product; it is a representation of it, either in whole or in part, that delivers value or represents the tangible activity or solution in some way.

Rather than trying to do everything, an agile approach looks at what the MVP might be, so that a relatively small amount of work can be undertaken that then can be reviewed and improved upon. The ideal early MVP is able to showcase and test the solution without doing a lot of work, so that the concept can be tested and proven before further resources and costs are committed. The MVP should validate if it provides a better solution than the current alternative and will be an attractive and viable option.

This concept can be used to test ideas and, as an ongoing approach to developing solutions, the benefit of using an MVP approach throughout development is that time and resources are retained, allowing for staged development, which itself allows the scope to change. By delivering value early, it also means that a return may be gained much earlier in the development cycle, which can be reinvested into additional time and resources for further development.

An MVP helps you to learn more about the scenario and identify whether a particular approach or solution is worthwhile, and whether it should be developed further or not. It is a great way to test your ideas at the start and validate any assumptions you might have, as well as verifying the information provided on which you are making decisions.

Early on, when a goal or a solution is identified of which we have little or no experience, early learning is a key action to finding the best approach as fast and as efficiently as possible. An MVP approach helps to reduce the associated risk by validating and testing assumptions, value and expectations. At this stage, the MVP is purely a representation of the final solution that can be tested to validate the concept and the value it could deliver. By creating a minimum viable product, tests can be carried out to identify if the chosen option is the best choice and if amendments need to be made.

Agile provides a framework to act and learn fast, whether initial concepts and ideas are the right and best thing to progress. Early MVPs provide an opportunity to find out more about the environment and inform our future actions and decisions. The MVP will help you to understand which elements of the solution are of the highest value and which are less important.

By creating an early representation of the solution and releasing it as an MVP, feedback can be gathered that informs decision making in the next stage of development and improvement. An MVP may take a number of forms; it could be a part of the solution or a representation of the solution.

Often, what is thought to be the best option may have unexpected outcomes or consequences when tried. Testing and trialling early can help to identify these and make changes to remove or mitigate the risks identified.

The concept of an MVP builds on the ideals of continuous delivery and feedback, where the definition of the MVP can be repeated to establish the next batch of activity.

By using the initial MVP to record early performance metrics, we can validate and revise our scope and estimations for how long a solution may take and how many resources are needed, and establish early improvements to processes and methods.

The concept of an MVP can be used in various contexts when planning and developing new ideas for growth and improvement.

Three contexts in which the approach of an MVP can be used are:

• Ideas – a test to validate needs and value.
• Change – quick improvements to mitigate urgent problems.
• Solution – a viable solution for delivering value early.

Testing ideas

Using a test-driven approach to solutions development, the MVP is a representation of the solution to be delivered that will test the assumed or proposed value, and provide an insight into whether the solution will fulfil needs and is an attractive proposition.

An MVP is different to a prototype as, typically, this is created before the solution has been developed. The MVP tests should help to define what the initial solution will include and what value it will deliver.

The objective of an MVP is to gain feedback and deliver some value early that can be measured to see if it is attractive and meets needs. The MVP at this stage may not be scalable: for example, the MVP may be quite clunky or time-consuming to undertake, but it is shared with a limited volume.

Quick change

A quick and simple way to solve a problem is to look for small options that quickly can mitigate the impact of a problem, or improve an existing process by identifying simple changes. This is useful especially for improving workflows, processes and systems in the short term that have become inefficient or ineffective. In this context, the MVP works to remove some of the friction or issues that are related to the problem to be solved by implementing small valuable changes.

Quick changes work to reduce the impact of the problem quickly and improve the situation immediately. These changes may be an interim solution or may prove to be a permanent solution.

Viable solution

Often, when a solution is scoped, there are more features and options identified than there are resources and time in which to complete it.

A key aim of agile is to deliver a solution that takes the optimum approach to delivering a solution that is fit for purpose, and satisfies needs enough to be a viable solution for adoption.

The optimal minimum viable solution ideally is the 20 per cent of the solution scope that delivers 80 per cent of the value, as per the Pareto Principle. However, this may not always be realistic and so this theory should be adopted with some flexibility. The MVP solution would be the first release of the solution that would be usable and deliver value, with the use of continuous improvement and delivery to develop, maintain and improve the solution further.

Analysing the work that has been scoped, estimated, ranked and prioritised and the results of any early testing, MVPs will help to plan a way forward and define the initial scope of work to deliver a new MVP working solution.

The minimum viable solution is not just all the ‘must have’ actions: just because something is an essential must have does not mean it has to be actioned first. For example, a new library requires system training on lending books but time available for this is limited. On the first week the team needs to know how to lend out books and receive them back. Since books are loaned for one week, the team does not need to know how to deal with overdue books and related fines until the eighth day, when the first books lent are due to expire.

In this library example, initially the MVP would be to train the team on lending and receiving books, while training for overdue books can be provided during the following week when there is more time available for further training.

Because of the nature of an MVP, the type of people on whom it is best to test are those who would be natural early adopters. The reason for this is that these early testers are less demanding of tried and tested solutions. They are more interested in improvement and added value than efficiency. The adoption lifecycle is explained in more detail in Section 24, Part 4.

MVPs work in all contexts to find the best way forward by working closely with the users and beneficiaries of the work we do, understanding that actions are inter related and dependent on one another. Change in one area impacts on others and this needs to be considered when we decide what activities are needed and what flexibility is needed for change: MVPs can help to identify and take these into account.

Mapping an MVP

Mapping MVPs is a useful exercise to do when establishing a way forward, both individually or with clients or colleagues. It can be useful especially where there is more work than resources.

Draw three interlocking circles. Label each circle ‘Jobs’, ‘Value’ and ‘Resources’ and fill each with sticky notes with items relevant to that circle (see Figure 16.1).

• Value: the benefits that the solution will deliver in order to meet its objectives.
• Jobs: the activities/components of the potential solution.
• Resources: the resources available, such as cash, time, people, materials, systems.

Move items into where the circles overlap to form potential MVPs that can be discussed and considered. based on the MVP goal:

• Test: select the value to be tested and then identify the associated jobs and resources needed to create an MVP to test this value.
• Change: select the job to be improved, identify the value that this will deliver and the resources needed to create the MVP.
• Solution: the ideal MVP in this context is one that delivers 80 per cent of the value defined by the objectives by doing only 20 per cent of the jobs.

Figure 16.1 Minimum viable product

This exercise may create a number of combinations and options that can be reviewed and tested.

Testing with MVPs

Two entrepreneurs wanted to set up a retail online business that generated sales through pay-per-click search engine advertising. The pair had done their research and found two products that received a high volume of keyword queries on search engines, but the competition for these keywords was low. The research suggested a gap in the market for retailing both these products online.

Taking an agile approach, they built their website incrementally in order to test and validate if people would buy Product A and Product B from them, if they were to offer them for sale.

• Test 1 – Google AdWords campaign – the team purchased a small volume of pay-per-click search engine advertising for the key phrases and the adverts linked through to a simple website holding page. The team measured if the adverts were clicked on and at what rate.
• Test 2 – A webpage with a number of potential product options was set up and, running the campaign again, the team tested if a visitor selected an item for more information and measured the click-through rate.
• Test 3 – A shopping cart was added to the site, the AdWords campaign was run again and the basket was monitored to see if items were added to basket and check out was attempted.

Product A passed all three tests and a payment gateway was implemented. The most popular products were selected for sale on the website and the business generated good volumes of sales.

Product B results were poor in both tests 1 and 2. While visitors viewed and clicked on the product options, the rate of attempted check out in test 3 was very low. This low rate of conversion meant that the cost of sales (in terms of advertising costs) was too high and, therefore, the model was not viable, no product was ordered and the website was shelved.