A recent study from Pendo shows that 4 out of 5 features of the average software product have lower usage than expected. For a startup, such a failure rate can be devastating, and for an established company it may turn into a substantial profit loss. A minimum viable feature (MVF) can help reducing the failure rate of a new feature.
One of the reasons of such a surprising statistic is that new feature ideas often go the market without a proper analysis of their customer value and ultimately return on investment. Even when you execute proper analysis, a product feature may still turn out to be a fiasco. This happens when the feature is rather new compared to the product. For example, the product feature targets new customer segments or offers a value proposition that the product has not yet provided before.
The natural question at this point is how to avoid failure when a feature has uncertain customer value? Clearly, before committing any resources to develop the feature, getting customer feedback is crucial. It allows to get their interest to adopt, use and even share the feature.
In order to get feedback, there are a different ways to go. An effective way is to launch a minimum viable feature (MVF), which serves the same purpose of the minimum viable product (MVP) used according to the lean startup.
This post describes the importance of an MVF and shows an example from one of the most innovative and successful platform businesses out there, Airbnb. But first, how can an MVF make a difference when it comes to launch innovative product features?
Why a minimum viable feature
Let us step back to the moment when a product feature is create and conceptualized. If we have already conducted thorough business analysis, why do we still need a minimum viable feature? There are 4 main reasons why you may need an MVF, especially for features that offer new value propositions compared to the product or offer an existing value proposition to new customer segments. The reasons are need validation, time to market, customer touchpoints, behavior testing. Let’s look at them one by one.
Need validation
The first reason for using an MVF is need validation.
You have studied the market and found an unsatisfied need that your feature can cover. The gap looks promising and you are ready to go on with developing the feature. However, you may wonder, what if that expected need is not there? What if it is just a blunder? Indeed, one thing is to hypothesize a need, another thing is to actually see it. Validation is key for features that address new needs with respect to the main product. An MVF allows to validate an expected customer need by testing actual preferences of customers through their actions. Of course, there has to be a logic explaining why a certain action validates a customer need. However, betting on the logic is way less risky than doing it on the actual need.
Time to market
The second reason to opt for an MVF is time to market. This is because, in many industries market and competition are dynamic. Take the e-commerce case, for example. Regulation, technology and customer preferences are evolving very quickly in many countries, during and after the COVID-19 pandemic. This is why any company who wants to enter or thrive in a dynamic industry, needs to have a fast product innovation cycle, from the moment a product feature concept is created, to the moment the feature is launched on the market.
Waiting too much can alter the market outcome for the worse. So, here is where the MVF comes in. It allows to quickly test if it makes sense, from a market perspective, to push for building the product feature or if it is better to pivot to some other feature. In a nutshell, the sooner you find a product feature to be a failure, the more the resources you can save for the next feature.
Customer touchpoints
Another important reason to adopt the MVF approach is customer touchpoints. There are always aspects about your customers’ needs and your value proposition that are unknown. The more you interact with customers, the more you get to know them and their needs, especially when your product is young and your knowledge about customers. In fact, trying different product features by minimum viable features, each can help to look for product-market fit. An MVF is a relatively low-effort and quick way to learn from real customers and grow from there.
Behavior testing
Last but not least, an MVF tests customer behavior rather than intent. In fact, while you can get early customer feedback through interviews, surveys, focus groups and more, be aware of the social-desirability bias. This reflects the tendency of people to answer questions in a way that is desirable by society, although people would act in a different, less socially desirable, way.
For example, if your product helps to save CO2 emissions, asking someone a question like “From 0 to 10 how likely is that you would make the effort to use this product if it was available?” may return not honest answers. The average answer will most likely be well above “5”. This is because people would just report “good behavior” in order to appear socially likeable. On the other hand, if you checked how many of those people would really use the product, you would record much less interest than what you recorded from their answers. This is because when you record actual behavior rather than just intent outcomes are closer to reality. This is why testing actual behavior is generally better than capturing intent, especially when there is risk of incurring into the social-desirability bias.
Testing behavior is not always easy or even possible, especially when on a tight budget or schedule. A product manager would lead the discussion about what type of customer testing works best for the specific case.
Anatomy of a minimum viable feature: example from Airbnb
So, what is an MVF really and how it differs from a full feature? An MVF is the minimum functionality that can allows to test the core customer value that the feature is expected to deliver. In other words, it is the minimal version of the feature that can be used to test if a feature really creates value for customers. Value creation can be tested in different ways. For example, it can be tested by checking customer interest, adoption or willingness to pay with respect to a potential new feature. In a nutshell, an MVF is a feature experiment to validate a hypothesis and reduce the risk of the feature being a fiasco.
Before looking at what constitutes an MVF, let us start with a concrete example to make things easier. The example is mainly about a web page and a web link introduced by Airbnb. Let us first look at the context. Then we jump to analyzing the Airbnb page and the specific link. Finally, we go back to spot where the MVF is in this case.
1. The context
On the 6th of May Airbnb announced the introduction of a new travel insurance page for guests. It came a bit as a surprise. Indeed, after having previously introduced a comprehensive liability insurance for hosts, now travel insurance was just, well, a web page. After all, most of Airbnb competitors offer travel insurance as part of their service. So why is Airbnb apparently so cautious about introducing travel insurance? Maybe they are concerned about whether guests are really interested in travel insurance. Who knows. The point here is not to talk about Airbnb’s product or marketing strategy. We are more interested to look at an example on how to deal with potential about market interest.
2. The Airbnb page on travel insurance
Let’s look at the travel insurance page piece by piece to find out. Below there is a screenshot of the page title. I took this screenshot at the end of May 2022.
Below the title, there is a paragraph, with small characters, that reads: “Airbnb is providing this information for general educational purposes only. Airbnb does not currently offer travel insurance or endorse the products or services of any travel insurance companies, agents, or brokers.”
Thus, the disclaimer and the title clearly say that the page is just informative. There is no attempt to actually sell or offer any insurance feature.
If you scroll down the page, there are some details about what travel insurance is, what the benefits are and when it makes sense to buy it. Below there is an excerpt about the benefits, just in case you are curious.
It seems that Airbnb is putting efforts in letting guests understand what travel insurance is all about in an impartial and informative way. In other words, Airbnb does not apparently try to convince guests to buy travel insurance.
3. The link to get an insurance quote
Now comes the interesting part. if you keep scrolling down the page, there is a paragraph where the customer can get a quote and even purchase insurance.
Although the paragraph shows a link with the name of the insurance provider, there is no particular advise to choose this provider. In fact, Airbnb clearly states there are many other providers that offer the same service. So, the main question you may ask is why the link is there in the first place.
There is certainly a reason for the link. Here is what a product manager would probably say about it. The link measures if guests are interested to buy insurance or not. Clicking on that link means there is one potential customer that wants insurance. So, for Airbnb product managers each click on the link would be a statement like: “yes, I am interested”. A little disclaimer here. This logic is not something that I have heard from anyone at Airbnb, either publicly or privately. It just comes from my own interpretation.
4. Spotting the minimum viable feature
Why are we talking about a product feature if this is just a web page then? The reason is that this web page looks pretty much the first version of a feature. Indeed, most customers will learn something new from this page. Some of them will interact with it by scrolling it up and down. Some other will click on the link. It probably did not take much of Airbnb resources to build up this page. Beyond the web page, all it took was, apparently, just a marketing campaign and some kind of partnership with the insurance provider. But what is interesting about this web page is that it is more than just a common feature. In fact, this web page is an MVF. Because it helps Airbnb understanding if guests are interested in buying travel insurance or not.
Characteristics of a minimum viable feature
After seeing an example of a minimum viable feature, let us consider what sets an MVF apart.
An early version of a product feature
Although an MVF may look incomplete and somehow out of the schemes, it has all the characteristics of a feature. If we look at the Airbnb case, there is value creation, because customers get the value of learning about travel insurance directly from Airbnb. There is action, because Airbnb guests can experience travel insurance through the insurance partner. There is return on investment, because Airbnb can learn more about customers’ interest in travel insurance. So, it is actually a feature.
Despite of the minimal set of functionalities, an MVF should not compromise on quality. This is because customers will see and use it as a normal feature. So, you do not want to disappoint them. Once you define the scope of an MVF, it still requires testing and business acceptance before launch. This means that, like any other version of a product, it needs to have a certain level of quality.
Hypothesis-driven
People often confuse an MVF with an MMF (minimum marketable feature). Indeed the two types of features are quite similar. However, they have a clear difference in terms of objectives. By definition, an MMF has the objective of providing value to customers by delivering the minimum set of functionality possible, but without any objective to test a hypothesis.
An MVF, on one hand, aims at validating a hypothesis. Therefore, it provides with the minimum required functionality to test the hypothesis. All that matters for an MVF is to validate or disprove a fundamental hypothesis on the value created to customers by the feature. In our Airbnb example, a possible hypothesis behind the travel insurance web page could be “80% of our existing visitors ask for an insurance quotation if informed about insurance benefits.” By setting a clear and quantifiable threshold, such as 80%, we can validate or disprove the hypothesis simply by measuring customer actions. In general, hypotheses can be about customer interest, customer adoption, customer willingness to pay, customer retention, and others. For the Airbnb example, Airbnb would validate the need when a sufficiently high number of page visitors click on the link to get a quote.
On the other hand, you would use MMFs when you have already proven the value hypothesis. However, you still want to committ limited resources, get customer feedback early in the product lifecycle and incrementally improve the product.
Effectiveness more than efficiency
Another benefit of an MVF is reaching the market early and with limited resources. In this way, if the value hypothesis is validated, the business can allocate more resources to the feature. In this sense, a MVF is effective to get the answer that product development needs in order to proceed with the feature.
However, an MVF may require higher running costs. This is because MVFs does not target aspects that increase efficiency, such as back-office operations and partners integration. For example, if you want to tests whether a new pricing model will appeal to customers, the MVF can be just a section describing the price with a button that allows the customer to pay and get onboard. Once the customer pays, there might be no automation for billing customers efficiency. So, back-office and administration personnel will have to do some manual work.
Time-limited
Higher running costs urge you to use an MVF with parsimony. Testing the value hypothesis should last not more than strictly needed to right amount of customer feedback. So, this is why it is important to establish in advance how long the test will take and what criteria will define the possible outcomes. Indeed, there can be three major outcomes.
First, if the feature turns out to have market potential, you can plan a full release or incremental releases in your roadmap. Second, if the MVF shows that the feature is not successful, you should stop investing on it and concentrate on some other feature. But there can also be a third outcome. This is when the hypothesis is neither validated nor disproved. In other words, there is need to repeat the MVF by changing its setup. This can happen when there are not enough customers seeing the value proposition or if the MVF appears to be not representative of the feature, for example when customer do not understand it.
What to get from a minimum viable feature
A minimum viable feature helps to validate a new product feature that offers a different value to customers than the rest of the product. As soon the new product feature involves a new customer segments, a different value proposition or a different user experience, an MVF is recommended. After all, learning more about customers is rarely a wasted investment.
Once you have the result of an MVF, you have more reliable information to decide whether the product feature is worth further investment or not. The uncertainty is still there, but it is lower than before. All of this without committing huge amounts of time and resources on something with uncertain outcomes.
If you found this post interesting or you want to share this experience, please leave a comment.