“A MVP is that version of a new product which allows a team to collect the maximum amount of validated learning about customers with the least effort“
-Eric Ries
Smallest amount of a proposed product that you can build, push out there and then get real feedback from users whether or not they are actually interested in it.
MVP is an early version of a product that can help spark an interest in the early adapters and validates a product or a feature idea in the very early stages of Product Development cycle.
When you run a MVP experiment you are always going to present it in some way shape or form, something that implies that your new product or feature is already real or coming soon. In this experiment ideally the users should know that they are being tested, That pure data is the one that we crave the most.
They are designed to:
– Test Hypotheses
– Test Assumptions
MVP is a process not a thing, its kind of a science experiment.
You should start with
– Assumptions
– Hypotheses
– A minimum criteria for success
Experiments rely heavily on a concept called “validated learning” they draw a distinction between learning and validated learning.
Validated Learning
Its anything that you learn from a customer thats done in a test environment in a way that you don’t bias your customer and you are studying them in an environment where they have no pressure and they act naturally.
Learning
Learning on the other hand is just pretty much anything else you do thats not designed like an experiment and the results you get from it are not that much to be trusted.
For example: If you are standing next to three paths and you don’t ask which leads so, you go down every single path and find out which one is the best.
Things to remember about MVP’s
One thing to remember is that MVP’s are not a prototype, MVP is just a basic version of what you are going to build.
“MVP is like idea validation at the end of the day and what it takes to get to that point”
Other core reason we build MVP’s is to mitigate risk as analysing the chosen path helps us save money, resources and things like opportunity cost.
Last thing about MVP’s is that we need to know about the speed at which the product needs to be built, thats where the mantra fail fast comes from.
If you fail fast and build various MVP experiments you need to run them quickly this way you can run more MVP experiments at the same amount of time and collect more data only then you are more likely to find a product or a feature that works and will be successful.
Key Points
– Fail Fast
– Run MVP experiments quickly
– Collect more data
– More likely to find successful product
What is an MVP for a PM
Risk of losing resources
– Time
– Money
– Opportunity Cost
Startups have a pretty low tolerance for risk as if your product fails you die and in a start up a PM will most likely be a founder as well. In the case of a larger company they have a higher tolerance of risk because larger the organisation, higher the cushion, higher the tolerance for risk.
For example: You work at google and your product or a feature fails the worst thats going to happen is you will lose a bit of money
Once the risk tolerance increases companies start caring about their brand or opportunity cost and even if the product fails there are come cases even failure can help your brand.
For example: Google glass failed but it gave them a lot of press saying that they are trying to innovate and come up with something new. Google glass is still being used in their warehouses.
Steps to running an MVP Experiment
– Problem/Solution Ideation
– Identify Assumptions (find the riskiest)
– Building hypothesis around them
– Establish minimum criteria for success (MCS)
– Pick MVP strategy type (how big the company is, risk factors)
– Exercise, Iterate, Evaluate (Is it worth your time and resources)
Problem/Solution Ideation
In simple words this can de explained as you figure out what the problem is and think about its solutions i.e what is the target audience/market? will the customers use it? will the users pay for it? will it get us enough ROI to continue? Is it worth our time?
This what needs to be figured out first.
Identify Assumptions
You have an idea that you are trying to build.
– What customers are you targeting?
– What solution you are trying to propose?
We assume we know things but sometimes we don’t, for example you assume your car is going to start, you have a full gas tank, your engine is fine.
Ideas are based off of some type of observation or some kind of intuition.
Like if i build something cheaper than my competition, I assume people are looking for something cheap.
If I build something convenient then I assume people want something convenient
Will the users be willing to pay for my product? This is a big enough problem if your business model relies heavily on subscriptions or services. Sometimes there are no satisfactory substitutes, sometimes there are so you need to come up with something that is at least satisfactory than the competitors for example a weather app it was once widely used but nowadays a simple google search can return the forecasts for the whole week.
Let’s talk about identifying the most riskiest assumptions and the reason we do this is to be lean and come up with a proper MVP because a PM is all about mitigating risks and saving resources. You need to focus on the most most riskiest assumptions first there is no need to focus on less risky assumptions as you may end up losing resources.
For example there was a time when people were completely satisfied with their feature phones but the likes of Steve Jobs thought people need more out of their phones this was the risk and it paid off at the end.
All this kind of depends upon marketing as well, the marketing team needs to tell people to care about something that they didn’t care about before.
For example Dominos in 30 mins, people before just wanted pizza now they want it in 30 mins (which is ruled out now because the delivery guys started ending up getting in accidents) but this is a very good marketing example which changed the game.
Don’t build solutions in search of a problem and that never goes well because the lack of problem is by far the riskiest.
Building Hypothesis
You have created a list of assumptions now is the time to put them to the test and look for potential issues that might lead to a disaster at the end because the assumptions are not really precise and they are not particularly actionable for example an assumption is people are not happy parking in a garage they want something more, I know what you are thinking “this doesn’t make any sense”
We need to take our assumptions roll them out e more specific and easier to deal with the hypotheses creation.
What is a hypotheses?
Good question, it’s a single written, testable statement of what you believe to be true with regards to the assumptions you have identified.
“MVP’s are sometimes called MVP Experiments”
Hypothesis sometimes brings clarity to not only you but to your team, its completely up to you to decide whether you want to build a hypothesis for every single one of your assumptions also sometimes one experiment can rule out all the other assumptions so sometimes not building a hypothesis for every single assumptions saves time.
Putting together a hypothesis
Hypothesis is actionable because it has specifics we define a target group and we define our expected outcome of our test, how our subjects do something and how do we plan to get there these are few of the things to keep in mind
– Target Group
– Potential Problem
– Action
– Expected Outcome
Let’s continue with our Dominos example
Target Group = Most of the people if not all
Potential Problem = Pizza gets delivered late, its cold, guests waiting, people avoid ordering at the last moment this in turn harms the daily sales
Action = Pizza Delivery in 30 mins
Expected outcome = People happy, business increased
Now we need to test the hypothesis
Right now we are just establishing that people are still interested at all, PM’s often use MVP tests and hypothesis to make changes to an existing product in order to create a desired outcome.
Minimum Criteria for success
There are three different outcomes to an MVP test
– Your hypothesis is false and not worth doing
– Your hypothesis is true without question
– Your somewhere in the middle
90% of the MVP experiments your are going to run as a PM or as an entrepreneur, you will end up in situation 3 that is somewhere in the middle.
You need a line between “Worth it, all systems functional and not worth it” because MCS gives clarity and meaning to whatever your idea is.
If you are unable to get MCS you can still go and start building but you need to proceed with caution even if you barely validate it.
Minimum Criteria for success for start ups
When you have exisiting products, exisiting user base and potentially existing revenue you are going to be more concerned with certain metrics and less sensitive to certain type of risks and costs but in the case of startups you are looking for what we call “Validation Metrics”
Metric
What is a metric? Metric in terms of product development is an indicator that shows how users interact with your product/feature and it can be anything such as signed up users, active users, no of subscription etc.
Validation Metrics
It is something that demonstrates real interest from your potential customers.
– Percentage of people that signed up
– Percentage of people that interacts with your posts
– Average purchase price
– No of people that open/respond to your emails
It is fairly hard to identify just how much money a feature will make but a product is relatively easy so we will probably look at things like
– Lifetime Value
– Gross Margins
At this point we know what we are going to build whether it is
– a product
– a feature
– a big change to the product you already have
We know what our assumptions are and which ones to test and we also have an idea what outcomes do we want to get out of these tests.
Techniques for MVP Experiments.
These are of three types mostly
– Email MVP
– Shadow Button MVP
– 404/Coming Soon MVP
Email MVP
You need
– Email Client
– Email List/User base
– Basic composing skills
Usually in an established company you already have a list of users and you contact them frequently. Every company uses email marketing in some way.
Your users are already used to
– Buying things
– Signing up for things
– Engaging with the content
– Emails that they receive
They are already ready to have a test run, you email them a pitch for a new feature or a new product this way you can exactly see how they will react.
What to do if you don’t have an email list? You can gather a list from your potential customers i.e the people that fit your. target demographics but keep in mind to maintain a personal tone to avoid. being reported as spam.
Example: A daily deals website that sends out regular emails asking to sign up and sometimes they have a buy now button in place to see how many customers are willing to buy.
Shadow Button MVP
This is generally specific to PM’s. It requires more resources compared to email MVP but its very easy to pull off.
Companies instead of building out a new feature will put a button within their already made product that kind of links the user to their new feature.. If they click on it, it registers as the user did it and the button either does nothing and says the page is broken or the page says “Sorry, this feature is coming soon”.
We do this because people act differently if they believe this is real.
Example: You want to add a social media log in option for your app, all you have to do is create a page and send it out to your users and collect the data.
404/Coming Soon MVP
Over here act like you are adding a new feature or a product but when the user navigates to that page or a section it displays 404/server error page or a simple Thank You for your interest page
Example: Amazon does this a lot to test out their new features or a side project.