MVS: The dance partner of MVP in Action

Chris Roach
7 min readJun 23, 2019

--

MVS or Minimum Viable Segment is a method of execution to gain the most reference-ability from the smallest footprint you can.

You can read more on MVS and how it relates to MVP in this excellent post https://www.startupsecrets.com/mvs/

When you’re building an early stage company, you want to be able to serve a group of customers that all share the same set of tangible needs. The goal of MVS is identifying a group of customers that share those needs,

I need X, I need X, I need Y, I need X, I need Z

Needs are often built from pain, so you should take extra care to focus on those pains rather than get lost in the wants.

If you can serve that smallest customer segment their needs and make them reference-able, you can go straight to the next customer/s who have the same needs and serve them without major changes to your MVP.

It’s easier to jump across to these customers with matching needs than it would be to jump to customers with disparate needs.

How do you identify MVS?

To begin searching for that segment, you need some customers, don’t worry if you only have a small handful, or if you have more, as long as you have somewhere to start.

Sit down with a list of those companies using your product, and begin writing down a list of attributes that each customer has. What we want to find is a common set of attributes across all of your customers.

Some tips:

  • Take as many of your current customers you have now.
  • List as many attributes as you can for each of those customers.
  • Don’t think too much, this exercise should only take you about 45 minutes.
  • Don’t complicate things, try and keep your attributes simple. Booleans are good here.
  • Attributes should also be as quantifiable as possible.
  • If you don’t find an attribute quantifiable, but still think it’s valuable, note it down and then try to tease out the attribute into many smaller things.
  • Ignore your gut for now. Only write down things you know for sure.

A working example:

Alice runs an e-commerce product, she has only 4 very successful customers who love her products, she wants to find more of those customers. she sits down and writes out their attributes to try and understand what makes her product work so well for them specifically.

Customer A

  • Need a small catalogue of products
  • Engage via social media primarily with their customers
  • They only sell physical products
  • Need freedom and flexibility

Customer B

  • They need mobile support
  • Engage via social media primarily with their customers
  • Have a small catalogue of products
  • Need the ability to add custom information to their products

Customer C

  • Have a small catalogue of products
  • Engage via social media primarily with their customers
  • They only sell physical products
  • Need freedom and flexibility

Customer D

  • They need mobile support
  • They sell digital products
  • Are very design focused
  • Engage via social media primarily with their customers

Now Alice has finished writing out the attributes for those customers. She begins to try and find a correlation between them, what matches and what does not.

There’s a few points that add up across her customers and a few that don’t. There’s also a very qualitative point that matches across customers “Need freedom and flexibility”, let’s try to tease it out into something more quantifiable before we go any further.

Teasing out qualitative attributes:

Alice tries to dig deeper into their exact use-cases to find out why they say it’s important to them.

Why does customer A need “Freedom and flexibility”?

  • They have complex rules on their ordering system and need flexibility in the commerce system they use to handle those custom rules.
  • They are very brand/design focused and need to bend the backend (code) to fit their designs. In other systems this is more difficult.

Why does customer C need “Freedom and flexibility”?

  • They have a complex points system on their products and want to track that with custom data attached to their product data.
  • They are very brand/design focused and need to bend the backend (code) to fit their designs. In other systems this is more difficult.

Great. We can see that design and brand are important to our MVS, there is also an element of custom logic, or rules that matches across the two customers.

Now, you should be able to say on a single piece of paper everything that those customers share, a similar set of needs that they are not getting somewhere else.

Representing some of those customer needs in a Venn Diagram you can begin to see where they align strongly and where perhaps those alignments are weaker.

Our finished list might look something like this:

  • Need a small catalogue of products
  • Design is very important to them
  • They are very active on social media
  • They require custom rules or logic to operate their e-commerce system

Some Advice

Sometimes it’s very hard to try and reconcile what an exercise like this might show you about your product, vs how you emotionally feel about it.

Sometimes you’ll find that everything lines up nicely with your expectations, other times, you’ll find a list of attributes that you feel don’t amount to what your product is, why your product exists, or how you want your product to be used.

Don’t let that cloud your judgement.

It’s likely you won’t get those attributes perfect first time, so you need to see MVS as an iterative approach. You will probably rinse, test, and repeat these steps in the wild many times over before you reach good traction.

Keep your attributes simple, be prepared to test and fail quickly, and understand this is a process, not a quick fix.

Next Steps

Congratulations!

The hardest part of this was really getting those attributes written down, into some form of quantitative state, and finding attributes that matched across customers. We’ll call the finished list a profile.

Alice’s Customer Profile

Customers who find my product great for them:

  • Need a small catalogue of products
  • Say design is very important to them
  • Are very active on social media
  • Require custom rules or logic to operate their e-commerce system

The next step for Alice is to test her profile.

Alice has a salesperson called John. Together they work out a scoring system for the profile attributes.

Each customer will be given a score against an attribute, higher being heavily aligned with the attribute, lower being not aligned at all.

In some circumstances you might want to weight attributes if they’re particularly important to the business, or you feel like they are worth more than some others.

For example, Alice and John might give “Are very active on social media” a lower weighting than perhaps “Need a small catalogue of products”.

In this example however, Alice and John decide that weighting is not particularly helpful for their first iteration.

Whenever John talks to a potential customer he grades them against the attributes in the profile. After a speaking to a customer, John might get something that looks like this:

New Customer X

  • Need a small catalogue of products — 10
  • Say design is very important to them — 8
  • Are very active on social media — 5
  • Require custom rules or logic to operate their e-commerce system — 9

New Customer Y

  • Need a small catalogue of products — 1
  • Say design is very important to them — 9
  • Are very active on social media — 8
  • Require custom rules or logic to operate their e-commerce system — 1

Once a day, John returns to Alice and they look through the customers and how they fit against the profile.

In the examples above. New Customer X, scores 32 out of a potential 40, New Customer Y, scores 19.

When they hit 30 or more (about 75%) Alice and John put those customers in touch with a small group of people at the company designed to help the customer get through the product easily and quickly.

Anyone who doesn’t that mark of 30 or more Alice and John make a list of the things that they missed and then contact the customer to say “Thank you, but no thank you”.

Is it working?

There are a number of ways for Alice to know if she’s found a strong MVS.

  • Alice company will say “No thank you” to many more customers than they are now.
  • The customers Alice is winning will be happier on her product.
  • John will be more effective, as he focuses more on only the customers that are a good fit.
  • There will be less feature requests for Alice product.
  • They may find that requests for issues or features starts to align more across their customer base.

Iteration

Now that Alice is finding a pattern in her successful customers she might consider several ways of iterating over the process.

If the segment is small

  • Alice can look at adjacencies she can serve without or with minor changes to her MVP.

If the segment is large

  • Alice can keep plugging away at the segment bringing more customers on board.
  • Alice can also talk to her customers and find out what they need to grow, so she can pre-empt their next moves and grow them into more valuable customers.

A Pinch of Salt

Our example of Alice e-commerce business is an ideal scenario, but there are a few realities we need to cover.

  • It’s unlikely you’ll get this right first time. So you’re going to have to be prepared to fail, just make sure you get back up and run the process again. The worst thing you can do is fail once and say this doesn’t work.
  • Any failure is good. It shows you attributes of a profile that you can immediately ignore, the more variables you can ignore, the more focused you can be.
  • You will need to continually rinse and repeat this process to get traction and grow.
  • Alice uses her sales team to ACID test her profile attributes, you may be a purely inbound organisation, so your messaging might be the first thing you change to align with your profile.
  • Talk to your customers as much as you can. You need to know everything you can about them.

--

--