Perfection By Subtraction – The Minimum Feature Set

By knowing things that exist, you can know that which does not exist.”
Book of Five Rings

I was having coffee with a former student who was complained that my idea of building a first product release with a minimum feature set was a bad idea. (One of the principles of Customer Development is to get out of the building and understand the smallest feature-set customers will pay for in the first release.)

“Steve, you’re wrong. I can’t get more than one of ten potential customers to think that this is something they’d buy.”  I asked, “So what does the one who likes it say?”  “Well he didn’t like it either.” he replied, “but when I started talking about our entire vision, he couldn’t wait to help get our product into his company.”

The Minimum Feature Set is Not The Goal
This minimum feature set (sometimes called the “minimum viable product”) causes lots of confusion. Founders act like the “minimum” part is the goal.  Or worse, that every potential customer should want it.  In the real world not every customer is going to get overly excited about your minimum feature set.  Only a special subset of customers will and what gets them breathing heavy is the long-term vision for your product.

The reality is that the minimum feature set is 1) a tactic to reduce wasted engineering hours (code left on the floor) and 2) to get the product in the hands of early visionary customers as soon as possible.

You’re selling the vision and delivering the minimum feature set to visionaries not everyone.

Why A Minimum Feature Set?
The minimum feature set is the inverse of what most sales and marketing groups ask of their development teams. Usually the cry is for more features, typically based on “Here’s what I heard from the last customer I visited.”

In the Customer Development model, the premise is that a very small group of early visionary customers will guide your product features until you find a profitable business model. Rather than asking customers explicitly about feature X, Y or Z, one approach to defining the minimum features set is to ask, “What is the smallest or least complicated problem that the customer will pay us to solve?”

This rigor of “no new features until you’ve exhausted the search for a business model” counters a natural tendency of people who talk to customers – you tend to collect a list of features that if added, will get one additional customer to buy. Soon you have a ten page feature list just to sell ten customers. Your true goal is to have a feature list that’s just a single paragraph long that you can sell to thousands of customers. Your mantra becomes “Less is more.”

You’re Selling The Vision
Most startups following Customer Development and a Lean Startup methodology understand the idea of a minimal viable product.  But they get it wrong in thinking that’s the point.  It’s not.

Most customers will not want a product with a minimal feature set. In fact, the majority of customers will hate it.  So why do it?  Because you are selling the first version of your product to Earlyvangelists.

Earlyvangelists = Early Adopter + Internal Evangelist
Earlyvangelists are a special breed of customers willing to take a risk on your startup’s product or service. They can actually envision its potential to solve a critical and immediate problem—and they have the budget to purchase it. Unfortunately, most customers don’t fit this profile.

Earlyvangelists can be identified by these characteristics:

  • They have a problem.
  • They understand they have a problem.
  • They are actively searching for a solution and has a timetable for finding it.
  • The problem is painful enough that they have cobbled together an interim solution.
  • They have, or can quickly acquire, dollars to purchase the product to solve their problem.

These Earlyvangelists are first buying the vision and then the product. They need to fall in love with the idea of your product.  It’s the vision that will keep them committed the many times you screw up.  You’ll have bugs, your product will eat their data, you’ll get the features wrong, performance will be bad, you’ll argue about pricing, etc.

But Earlyvangelists will stick with you through good and bad because they share your vision.  In reality Earlyvangelists are now part of your team. If you’re selling to a business, your Earlyvangelists will end up using your slides and metrics to help sell your product inside their own company!

This means Earlyvangelists, particularly in corporations, will be buying into your entire vision, not just your first product release. They will need to hear what your company plans to deliver over the next 18 to 36 months.

That means your Product and Customer Development groups must agree that:

  • The minimum feature set is spec’d with Earlyvangelist interaction,
  • You will provide a one-page product vision or roadmap (typically 18 months to 3 years out) that’s shared with Earlyvangelists
  • Everyone (including the Earlyvangelists) understand the vision is subject to change.

In Customer Development your goal is not to avoid spending money but to preserve your cash as you search for a repeatable and scalable business model. Seeing a repeatable pattern of sales to Earlyvangelists is a sign you may have found your first scalable business model.

Lessons Learned

  • Minimum feature set (“minimum viable product”) is a Customer Development tactic to reduce engineering waste and to get product in the hands of Earlyvangelists soonest.
  • Earlyvangelists require a 18 – 36 month product vision past the minimum feature set.
  • You’re selling the vision and delivering the minimum feature set.

Add to FacebookAdd to DiggAdd to Del.icio.usAdd to StumbleuponAdd to RedditAdd to BlinklistAdd to TwitterAdd to TechnoratiAdd to Yahoo BuzzAdd to Newsvine

46 Responses

  1. Steve, this is excellent advice for entrepreneurs. We are currently developing a business model software for the iPad. We have a vision of how a complete software could look like in 2 years, however, we are starting with a minimum version that caters to Earlyvangelists – this will allow us to seduce them, get their feedback, focus our development on core issues, and to build a foothold market.

  2. I get and like the whole MVP concept. 2 questions:

    1. You say “… the smallest feature-set customers will pay for…”. What is the analogue for purchase when it is a free consumer web service that is monetized in other ways, and how to best measure that in meaningful way?

    2. Again, in a consumer web service, how do you communicate your vision to those earlyvangelists (like that term) without broadcasting it on the web for the world and all of your competitors to see?

    Thanks in advance and love your blog.

    • If I may offer a couple suggested answers to your questios:

      1. The analogy to buying for a free web service might be: (1) time spent on the site/service (2) number or frequency of return visits (3) number of referrals made by a user. Each of these would indicate value to the service or parts of the service.

      2. If looking for “earlyvangelists” you might consider a closed, invitation only beta. Then you could somewhat control the distribution. However as you go wider you should expect competitors to see and find out about your service – and be ready to move faster and smarter. If you are depending on secrecy as your only competitive advantage you’re doomed, IMO.

  3. [...] Perfection By Subtraction – The Minimum Feature Set « Steve Blank [...]

  4. The last two posts are wonderfully succinct, compelling and, I can tell you, all too true in my experience.

    I think the competitive analysis issue often becomes endemic within companies or product groups whose marketing departments, through hubris, spend more time worrying about and listening to thier competitors than to thier customers. Sales, too, through the odd apocrypal tale of a major losses due to single issing features, can contribute to this. The end result can often be the locking of products into specific architectures for the sake of maintaining the checklist, quashing innovation. Even entire competitive segments can end up in a race to create the most bloated products.

    Thanks for your insights, they’re terrific.

  5. A chicken and egg question: How would you sell a 36-month plan to Earlyvangelists before all of them tell you what features they’d like to have?

    • You start with your vision. If they don’t react positively they’re not earlyvangelists or you have a bad idea.

      • So when you say plan, you don’t mean a too-detailed schedule but more a vision and
        “Everyone (including the Earlyvangelists) understand the vision is subject to change.”

        Gotcha.

  6. This post partly answers my earlier question (which I asked in no-accounting-for-startups): how do you “unsell” what you have sold during the customer validation step when you realize that you need to go back to customer discovery once again. The whole vision is subject to change, but what about the minimum feature set? Doesn’t this mean that the minimum feature set must also be subject to change or that it will grow during each iteration?

  7. Thank you. I completely agree that vision and product are two distinct entities. Early customers need to understand and buy into that idea. They are excited about the vision and they know that they need the minimal product right now.

    I think it’s important to note that sometimes (perhaps often) ‘Earlyvangelists’ can’t finely articulate their understanding of this, but it should be fairly obvious that they are onboard.

    Also, if you are having a hard time finding these Earlyvangelists it may be because your vision/product idea just isn’t compelling – or you are trying to sell the idea to the wrong level or type of user.

    Again, thanks for distilling this concept into a very consumable post.

  8. Great post to really clarify and distill some of these principles. Even when we think we’re doing customer development, it’s easy to forget the nuances of the minimum viable product. Of course almost no one is happy with it! People don’t generally want minimum products. But it’s an amazing tool to get a toehold with customers and figure out whether your longer-term vision rings true for them.

  9. Really liked the post. In addition to converting earlyvangelists, I like to mention that the roadmap is also a self-help tool for the ambitious entrepreneur: it helps them mentally be okay with the minimum product because they can put all the features they would like to build (another source of feature creep) on the roadmap. It is kind of like “permission” to build the minimum product because they can tell themselves that the full version is coming. It also helps entrepreneurs see priorities of which features may matter more. Of course, the whole point is to avoid wasted effort on something no one wants to buy and the roadmap has lots of benefits in achieving this goal.

  10. Great post, Steve.

    I think startups often go crazy with features because they are constantly trying to please any customer they can identify. Also when you are a startup, customers are so precious, it just seems easier once you get a customer who will engage with you to keep building features for that customer than to go find another customer for your current product (god forbid losing one of the few customers you already have because you wouldn’t add the feature they want).

    In my first startup in 1999, we felt that it was easier to build features for new potential customers than to find more customers for the features we already had. This was a huge mistake and led to an overly complex product that missed release deadline after release deadline.

    After we released the product, as you have pointed out, those customers who wanted additional features were never truly committed to our product and often weren’t even around anymore by the time the feature was ready! We had much better success at that startup (and at later startups I was at) when we focused on finding customers for features we already had as part of our vision. Sure we would alter the product, pivot, etc. but not until we really, really had to and we were sure that customers were committed to that product/feature. Almost never would we accept a customer who said “well, I will buy but only if/when you build X”.

    You also have to resist impulses from your sales team (if you have one) who will constantly lament “if only we had X in the product” I could make quota!

  11. [...] all those product managers obsessed with feature-rich first products? Steve Blank would like to have a word with them. In his most recent post, he explains how the goal should be the exact opposite: “the minimum [...]

  12. “It seems that perfection is reached not when there is nothing left to add, but when there is nothing left to take away”
    –Antoine de St-Exupery

  13. [...] Perfection By Subtraction – The Minimum Feature Set « Steve Blank I'm a big fan of getting something simple into production fast, then adding on more features as the users work with the system and provide some real feedback. Steve Blank presents a nice view of why you want to implement the minimum feature set first, but how you have to sell the 18-month vision at the same time. (tags: development design agile business) [...]

    • Define that subtraction by doing it recursively. You end up with layers of abstraction.

      Further, you can sell individual minimal marketable features to customers in other markets. Look at MMC snap-ins. Or, n-tiered functionality. As long as it has a market fit, you can always add a new MMF, or a new thread of MMFs.

      We sold a thing called a node. We added optional elements to the node. We sublimated external functionality. We had other discontinuous technologies productized in other markets and cross-sold only to those in our current market that would find it useful, not everyone. Worked towards a gradual merging of those populations. We also crossed platforms. Plenty of stuff to do well beyond your 18 mo. horizon. Expected cash flow events every 3 months.

    • I’ve given this more thought. The thing we are subtracting from is a vector of differentiation. Each subtraction is on this vector. Each of these vectors is has an associated S-curve showing the instantaneous price-performance and a local slope. The vector becomes more expense and its performance improvements move from being highly leveraged to being very difficult to leverage. That S-curve is also intersected by a line indicating the point at which the customer will no longer pay for improvement to that vector of differentiation, the commoditization line. This is where you must change your vector of differentiation.

      An offer usually consists of several vectors of differentiation, all of which are being invested in and improved.

      Moore’s technology adoption lifecycle also shows how the entire company and complete offer change over time. Every offer component, again, being a vector of differentiation, not just the application or technology.

      The technology layer and product/service/experience layer(s) all have vectors of differentiation. If you consider the technology layer to be carrier, the product layer would be the carried/content layer. These layers are independent, and potentially independently discontinuous. The entire stack is a collection of vectors of differentiation.

      There is plenty to iterate at all times.

      If you want to change your vector of differentiation, it will require a discontinuous innovation long before it is ready to replace your current technology. This requires a Christensen separation, as in a division, at the technology layer. The S-curve for this technology should be at a less mature part of its curve when it is positioned to intersect your current vector of differentiation at the point in time when you want to change direction.

      All of this can take twenty plus years. There is no rush as long as you make your quarterly revenues.

  14. Hi Steve

    Here is a link to a famous skit from Saturday Night Live where the producer keeps asking for more cowbell.

    http://noolmusic.com/my_video/snl_will_ferrell_more_cowbell_-_ebaums_world_video.php

    Would you agree that “More Cowbell” is a metaphor all the things we add to products that the end customer doesn’t care about and is not willing to pay for.

  15. You create Minimal marketable features to deliver value to the customer incrementally, sooner, and to distribute the user’s cognitive load. You do not do it to serve development.

    Minimal viable product is something else. The minimal marketable feature is the deliverable. It need not be barely working and in need of further work.

  16. Steve – if I present to someone who could be – but isn’t – an Earlyvangelist, won’t I lose that client for good, since he will judge my product as incomplete?

    Since we’re aiming for the Enterprise segment, 9 future clients out of 10 looks like a pretty high price to pay.

    • No, they will not judge your product as incomplete. Not unless you purposely piss them off, act like an idiot, or go in telling them you have a complete solution ready to be bought by everyone.

      If instead you go in saying, “this may be too early for you but…” and they agree you then say, “how about I give you a call in a year or so when I have other early customers?” Most will appreciate it.

      Of course, there’s a risk some will think you wasted their time. However the alternative is sitting in your building finishing your product so it’s “done” and only then finding out that most of the features you built aren’t wanted by anyone.

      steve

  17. [...] Steve Blank – Perfection by Subtraction – Having a clear, tight vision helps to keep development scope down, but it isn’t the key to the “minimum viable product” often mentioned in discussions about product development.  The key is to get a product in front of customers who can understand the vision and who can become evangelists for it because – They have a problem your vision will solve. They understand they have the problem. They have been actively looking for a solution. They have put together some parts of a solution themselves. They have or can get a budget for something that solves the problem.  These customers can validate the vision and will actively pull it into the shape that fits their context. With them behind you – you can develop a beta product that is much closer to what the market needs.  This is also part of Bessemer’s Law #5 – Build Employee Software – which talks about the “consumerization of software” that SaaS has enabled. [...]

  18. [...] Perfection by Subtraction: The Minimum Feature Set Steve Blank is one of the great contemporary entrepreneurial thinkers. An advocate of the Minimum Viable Product, Blank argues for making the product and the market work together. The approach he lays out asks entrepreneurs to build a dialog with the customer and use that to flesh out the product. It’s a smart way to build an HR department. Ask yourself, ‘what’s the minimum level of HR service?’ [...]

  19. [...] of ideas such as Lean Startup, and Customer Development, provides a very needed explanation of the minimum viable product here.  Something any entrepreneur and/or product manager should read and [...]

  20. [...] "Mit dem Wissen, dass Dinge existieren, kann man wissen, was nicht existiert. "Buch der fünf Ringe I Idee war mit Kaffee mit einem ehemaligen Schüler, die schlecht wurde beklagt, dass meine Idee zum Bau einer ersten Produkt-Release mit einem Minimum ein Merkmal gesetzt wurde. (Einer der Grundsätze des Customer Development ist raus [...] URL des Original-Artikel http://steveblank.com/2010/03/04/perfection-by-subtraction-the-minimum-feature-set/ [...]

  21. [...] z možných prístupov je využiť princíp: Minimum feature set. V preklade: minimálna sada [...]

  22. [...] and mobilize the masses. These platforms are no longer reliant on tech geeks and earlyvangelists (Earlyvangelists = Early Adopter + Internal Evangelist) to provide them with traffic. Yet there is a distinct divide between how college students and the [...]

  23. [...] about to jump into the waters unless others have done it before me. Typically version 1 of anything has just enough features to get it into the market, or for it to be viable in the market, with a few rough edges here and there. The Minimally Viable [...]

  24. [...] Perfection By Subtraction – The Minimum Feature Set (steveblank.com) [...]

  25. [...] Rhaya Shilts join up on the team and begin building the business.  We did the right thing, build a minimally viable product, test it, pivot and build some more.  It’s been two years since we launched and we have [...]

  26. [...] you find a customer that embraces your discovery process, you’ve found an earlyvangelist. [...]

  27. [...] In the startup world there has been a lot of noise made about the Lean Startup Movement, made popular by Eric Ries, first through his blog and, most recently, through his best-selling book.  One of the tenets of a Lean Startup is the Minimum Viable Product (MVP), which he defines as the “fastest way to get through the Build-Measure-Learn feedback loop with the minimum amount of effort.”  I won’t go into the MVP too much, mostly since I can point to other blogs who have done it far better than I could (like here and here). [...]

  28. [...] this doesn’t mean you should be doing lots of press yet. At this point your customers are still early adopters and your product may well change a few more ti… It is very hard to redefine your public image once you’ve created it. Only once you’ve [...]

  29. [...] asked him how things were going and got the update on how the earlyvangelistswere using the product. As I had heard, they were ecstatic. But Bob said he was worried he hadn’t [...]

  30. [...] would like to remind of a blogpost that i came across reading through a while ago on MMF by Steve Blank which admittedly i have noted down in one of my notes months ago.According to Steve, the [...]

  31. [...] 这一阶段是与用户面对面沟通的好时机,可以进一步了解他们使用该产品的原因和方式。习惯测试是为了弄清这些“早期支持者们(earlyvangelists)”有什么独特之处,并探寻其中的洞见,以推广到其他用户。 [...]

  32. [...] nothing wrong with RORE as a guiding principle. Focusing first on an MVP which is not feature rich is a great business strategy. But it’s a business strategy, not a [...]

  33. [...] sense of their system. One of the Onyx product managers got engaged and became an enthusiastic earlyvanglist. The integration effort probably used up a calendar month of our engineering time and an few hours [...]

  34. [...] the overall direction of a product based on consumer feedback and data. People like Eric Ries and Steve Blank spend a lot of time advocating this ‘lean’ approach to product [...]

  35. [...] behind your value proposition for the most promising capability, you set up experiments: minimum viable features that enable you to test your hypothesis [...]

  36. [...] me amplify a bit on “customer development” by taking another quote from the same blog post by Steve Blank. One of the principles of Customer Development is to get out of the building and understand the [...]

  37. [...] to the Minimum Viable Product methodology prescribed by many entrepreneurial gurus, most famously Steve Blank and Eric Ries. Their advice is difficult to disagree with – don’t waste time and resources [...]

  38. [...] Take a look at Steve Blank’s definition of an “earlyvangelist“: [...]

  39. [...] 这一阶段是与用户面对面沟通的好时机,可以进一步了解他们使用该产品的原因和方式。习惯测试是为了弄清这些“早期支持者们(earlyvangelists)”有什么独特之处,并探寻其中的洞见,以推广到其他用户。 [...]

  40. [...] through the challenges and the hard times. It’s all about daily steps forward. Deliver that MVP, onboard those early adopters, raise that first round, keep listening to your customers, constantly [...]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 151,631 other followers