It can be difficult to define an MVP through specific examples, given they can vary so dramatically in form, function, purpose, and audience. Instead, it is best to clarify the concept by its core idea. An MVP is the first form of a product that can be released to users, with the intention of adding user data to determine future development.

MVPs have cultivated a reputation as something of a ‘humble beginning’ for tech brands – likely owing to the huge number of unicorn companies having started out with one. Everyday productivity utilities such as Slack and Dropbox, and industry-defining giants like Amazon and Facebook, all launched initially as scaled-back, barebones versions of their planned product offering.

Globally, a number of factors have conspired to create a ‘startup boom’ in recent years. In the UK, studies have found record numbers of new business creations, including a 14% increase in 2020. With more startups launching than ever before, early-stage founders are faced with a number of crucial considerations before launching to a crowded consumer and business market.

MVPs, for many, afford a valuable touchstone that clarifies the value of the core idea, and amplifies potential issues in a safer environment than a full product launch.

This is preferable to full launches for startup founders, simply because MVPs are optimal within the modern startup culture. Lean, agile operations looking to ‘fail fast’ benefit from testing their products and ideas quickly, in as many scenarios as possible, while allocating the fewest financial and time resources necessary to deliver.

What should an MVP include?

At the heart of the ‘lean’ startup methodology is the principle of getting a product into the hands of customers as quickly as possible, eliminating wasted resources, and maximising efficiency. Then, it can be tested on real use cases, and the value of the product assessed. Internal feedback loops can be more quickly sidelined, and organisational resources can be concentrated on building and learning – allowing startups to authenticate the worth of their concept and present a robust case for viability to potential investors.

Accordingly, it is important that founders consider carefully what their MVP will do, and what information will be most valuable.

The most common type of MVP is a prototype product. This will usually be a functional but strictly fundamental distillation of the product into its essential features. This allows startups to avoid the expense of developing a finished product for an idea which does not service the market, or waste time building features that may fail to connect with users, such as design elements built around imagined prospective user journey behaviour.

This is the most crucial aspect of the ‘build-measure-learn’ tenet of lean methodology – more user testing at an earlier stage is likely to result in a more refined and relevant end product.

At the more minimal end of the MVP scale, others may prefer to launch a representative landing page to indicate the function and value of the product without developing its core features – or release a demo video to highlight the potential of a product and gauge interest in a controlled environment at reduced cost.

Suitability for these lower-scale options will vary from case to case. For instance, a web3-facing application involving an emerging product like AI, which is likely to be complicated and expensive to produce even in prototype, may find the use of a landing page limited only to facilitating interested users to register for a developed product – which amounts to effective marketing, but ineffective user testing. Instead, a demo video can be effective in communicating to users what they aim to produce, and what value it will add.

Conversely, startups who plan to enter crowded marketplaces with an innovative iteration on existing products – for example a new commerce platform – may need to build a full prototype to effectively communicate why their product stands out from existing offerings.

In each case, it is important that startups know what information they intend to gather from an MVP, rather than just launching a working model of a product for the sake of it. For many products, this will look at user engagement metrics, how successfully the idea has translated into use in practice, and whether it has achieved its value at the user end.

MVPs can also test whether the revenue model of a product is market-viable. If the core target market proves unwilling to pay for the core features, a startup will discover this early enough to explore alternatives.

There is also a value in discrete analytic metrics, such as user registration and retention. Naturally, in a scaled-down version of the product, often lacking sophisticated marketing and CRM functions, these can often be taken as indicators, rather than conclusive evidence of success or failure.

Learning from failure

On a personal level, founders and developers can find the failure of an MVP dispiriting – after all, who doesn’t want an instant success? It is important to remember that failure is part of the reason we build MVPs.

All feedback is valuable. In practice, this means analysing as much user data as possible and building a framework for future development around the learnings of the core product launch.

This means reviewing the pre-launch actions the product had aimed to fulfil and identifying user behaviour that led to these being fulfilled – and when they failed to do so, where this occurred – and then build a case for why it happened. All of this is valuable information for late-stage development; a robust internal understanding of how user journeys map out, and where pain points infringe on the product’s value, will ensure a more complete and purposeful end product.

The underlying question of MVP development is this; can there be such a thing as a ‘bad’ failure? If benchmarks and goals are set appropriately ahead of launch, it is certainly rare to find no positives to be taken, and no prospect of remedying a product with user feedback. Certain cases may test this; when a product is simply too early for market, it may find its sense of credible ‘minimum’ stretched, and its viability compromised. If users cannot either experience, or perceive through demonstration, the value of a product, there is little value in the data they can provide back to developers.

For this reason, I have found in my own experience working with startups on new products, that the key to a positive MVP launch is to do it with a clear purpose in mind. Testing ideas in the market is a valuable resource when assessing the worthiness of a potential product, while user data is critical in assessing the necessary features for an end product. With the right performance testing criteria in hand, and a willingness to be agile to new ideas, there is no real ‘minimum’ for what these soft-launch products can do - MVPs are instead a flexible tool that can deliver credible ongoing steer for start-ups.