Tag Archives: lean startup

Test-driven Business Model Development for Web Startups

Business plans are out and business models are in. The waterfall method is out and agile development is in. Now an entrepreneur will iterate the business model through test-driven approaches using agile software development and customer development.

How would you use test-driven development in building your business model? A good start would be to iteratively identify and test key hypothesis for each component with your business model.

Previously I suggested a mash-up of the Lean Startup/Customer Development methodology and the Business Model Canvas. On the higher level, the “back-end” business model structure aligns with agile development principles, and the “front-end” structure with customer development principles. Causally executing product development alongside customer development would optimally lead to problem-solution fit, followed by product-market fit corresponding to your value proposition.

On the lower level, typical Lean Startup and Customer Development principles are segmented according to each of the nine business model building blocks. These are the techniques and tools that you would take into consideration when testing your business model hypothesis.

As all business model components have their own characteristics they also have their own hypothesis testing schemes. The following is an overview of techniques and tools to be used for testing each and one of your business model components from a Lean Startup and Customer Development perspective.

Customer segment is where you identify with customer prospects a specific problem that they agree upon. You would put the Customer Development methodology in order by getting out the building and mapping out a Customer Problem Presentation. For what questions to ask, the guys at Survey.io provide an excellent starting point.

Value proposition. Here you test to see if your solution fits the customer’s problem. The key is Minimum Viable Product. A Minimum Viable Product has just those features (and no more) that allows the product to be deployed and tested. Especially if you are in the online business you would consider putting up landing pages. Further, A/B testing, split testing and multivariate testing are super techniques for testing hygiene factors. This allows you to test and validate customer demand and positioning. Unbounce provides you with many of these tools.

Distribution channels. Do conversion optimization whether it includes your web page, social media accounts, Google Adwords or partner’s pages. Ask your self what is most cost-effective channel. Eric Ries wrote these awesome posts on how to use search engine marketing and Google Adwords in testing demand and value propositions.

Customer relationship. This one is bit more qualitative in nature than the distribution and product offering schemes. It is more about making an arena for collecting feedback from your early customer segments. You would consider using tools such as wikis, forums and social networks as a basis for collecting data. It may include making a beta testers’ community allowing for feedback, and distributing surveys and newsletters for measuring reach. Survey and form services such as KISSmetrics and WuFoo are useful tools for gathering feedback on-site.

Key activities. “Release early, release often” is a key tenet in agile software development. Question your self – how often you ship product. Do continuous deployment to learn and adjust. Testing should be a key activity itself. Automattic CEO Toni Schneider reported that WordPress.com averages about 16 product releases a day. Beat that!

Key resources. Arrange for the tools that you need in order to test the remainding components. For an example by signing up to Google, you can run Optimizer for split testing, Google Analytics for conversion optimization and Adwords for testing clickthrough on different value propositions. The good news is that this is cheap. Perhaps the most important resource – your co-founders and team members should embrace a learning culture and test-driven environment.

Partners. Here you would consider a similar approach to that of the customer segment. Your partners may also be the ones who provide you with key resources, such as Amazon for hosting or Google for distribution.

Revenue streams/cost structure is about defining the equation of your business model. If you are going for that 1 % of China’s users cliché you are pretty soon on thin ice. Instead, do it bottom-up. For an example, numbers of users times ad revenue per user. The point is to convert your assumptions into metrics that are actionable. Sooner you would be a low-burn startup. By removing what is broken, test-driven business model development enable you to do so.

The bottom line: In systemizing hypothesis by the business model components you will simplify testing methods and reduce risk in building a lean business model.

Did you like this post? Please, feel free to subscribe or follow me on Twitter.

How Lean paradigms go beyond gorillas’ assembly lines

Lean Manufacturing, Lean Production, Kaizen and Continuous Improvement (dear child has many names) have created buzz for many years. Originated with Toyota’s production system, “Lean” was primarily created with manufacturing businesses in mind. Now as manufacturing businesses are increasingly rendered by service and network businesses, how does Lean paradigms go beyond gorillas’ assembly lines and keep up with the new economy?

In one of my favorite articles, Casting of the Chains (pdf) from 2003, authors Ø. Fjeldstad and E. Andersen make a fruitful observation.

The world has changed. From 1960 to 1999 manufacturing companies’ share of GNP in the US, as well as its workforce, fell from 30 per cent to 15 per cent, with the consequence that such businesses are now a minority of the S&P500. Banks, transportation, building, healthcare, research pharmaceuticals and other services companies have taken over. Strategic models of the world, however, have not changed. When managers develop strategies for their companies, they still use the tools and language of the manufacturing organisation.

In brief, the authors found limitations in applying Porter’s Value Chain to other than traditional assembly line-based manufacturing businesses. Consequently, the authors extended the Value Chain to two more models; the Value Shop and the Value Network. The Value Shop creates value by scheduling activities and applying resources in a fashion that is appropriate to the needs of the client’s problem (typically management consulting, lawyers and doctors). Value in a Value Network is created by linking clients or customers who wish to be interdependent (typically banking, social networks and dating venues).

Business Model Patterns on Value Chain, Shops and Networks

When I joined my current employer to work with the web and startups, I did at the same time choose from working on Lean methodology alignment with one of Scandinavia’s leading media companies (see also Bharat N. Anand‘s Harvard Business Review case). Regardless of my interests in innovation methodologies and owing my conviction to entrepreneurship, I wanted to work with growth ventures rather than cutting down “corporate bacon” (is that innovation?). Later I discovered the Lean Manufacturing Startup, which basically adopts Lean Thinking and Customer Development to early-stage ventures and startups.

Although Lean Startup principles are argued to be generally applicable, it has mainly been applied to Enterprise- and Consumer Software cases. However, as far as Microsoft Windows creates value by linking consumers with third party software developers, and Google links consumers with advertisers, the software business generally acts as a value network. Hence, I believe that we start to see cases with the lean paradigm being adopted to the new economy.

In this manner, I assume that new schools of Lean methodologies not only help traditional management thinking avoid cramming business models with manufacturing approaches, but also preserve new-product introduction and disruptive innovation alongside continuous improvement.

Getting Lean: A Startup’s Key Learnings from Web 2.0 Expo

Being fortunate to win a scholarship for this year’s web 2.0 Expo in San Francisco (appreciation to @ericries, @SarahM and @TechWeb for having this opportunity), it is also in courtesy that I share a startup’s takeaways and lessons learned from the conference.

In my approach writing for the scholarship program for lean startups, I pitched how EasyPeasy could learn from attending the conference and the Lean Startup Intensive. I described how EasyPeasy currently is seeking to validate Product-Market fit and, presumably in the customer validation phase, searching for its first transactions.

The Grand Pivot

Arriving at the conference after a 20 hours travel from Oslo, Norway and suffering from a mild jet-lag, I finally got to see Eric Ries, Steve Blank, Sean Ellis and Dave McClure and the reminder of the Lean Startup movement at work. In nearly every event that I was attending that week, either it was the in the talks, keynotes or unconferences, “Lean Startup” was buzzing. One major highlight was attending Matt Brezina‘s talk: 5 stages of Xobni’s growth and 5 pivots along the way.

Having been a fan of Xobni for some time, I was excited to learn from their practical implications in pivoting from offering an e-mail analytics suite to that of “just” an outlook sidebar plug-in. Nevertheless, by talking to @brezina and @hnshah I had my thoughts about customer validation and pivoting matured. One week later, EasyPeasy is smoke testing for a new product offering.

Minimum Viable You

During the Lean Startup Intensive one common denominator came into view. Steve Blank, advocate of validating hypothesis about business models, talked about how a startup must strike balance between a product’s minimum feature set and maximum sales. Accordingly, Dave McClure talked about how a startup must balance between user hypothesis and revenue. Lessons learned, and one key tenet with the Lean Startup methodology is that a focus on validated learning will enable startups to mitigate risk in new-product introductions, often by including continuous deployment, arguably a start-charging-now and learn fast/fail fast philosophy.

This is where the minimum viable product comes into the picture. At the second day of the conference a case of minimum viable tactics was elaborated by @drewhouston and @asmith in their excellent talk: From Zero to a Million Users – Dropbox and Xobni lessons learned. I think that this presentation gave a great many startups, including EasyPeasy, a lean toolbox at hand.

People Analytics

The Lean Startup advocates use of metrics for validated learning. Yet, there still is one vast amount of metrics to track, dependent on what line of business you are in. At the second day, Neil Patel gave an exciting talk on Web Analytics – Tracking People and Not Just Numbers, and an amazing Q&A session about what metrics you ought to track. In plenary, Neil asked the audience for their websites’ URL, he quickly analyzed their pro and cons, and gave the audience applicable tips and tricks, all in real-time. I highly recommend his slides to be found on Slideshare, 5 Metrics You Ought to Track.

iPad, Flash and HTML5

Soon after arriving San Francisco I found that nearly every overseas visitor who I got to meet at the conference, in their very first morning in town, had been rushing over to the Apple store. Ourselves, arriving one day after our Norwegian fellows (we organized by extending the #w2e Twitter hashtag with #w2eNOR), we found that the iPad 3G edition was already sold out. We put our names on a notification list, hoping that another delivery would arrive at the store before we were on our way back home.

Nevertheless, web 2.0 Expo hosted a couple of very popular talks on the tablet innovation, here and here. iPad were on everybody’s tongue and the big discussion tended to be about Flash vs. HTML5 and Adobe vs. Apple. Among a many good keynotes, Brady Forrest‘s chat with Adobe CTO Kevin Lynch did comb the debate. As when it comes to EasyPeasy, currently offering an open source operating system for netbooks, you might have a clue about how this might affect a pivot.

Learn Fast, Fail Fast

Truth is that there was many interesting talks, and unfortunately one did not have the time to attend all.  The main theme of the conference being Web as a Platform, there was a lot of interesting stuff on social media marketing, cloud computing, mobility, usability, virtual and social gaming. Many more events and people should be mentioned, yet I think that the above is sufficient to draw some main characteristics.

Pivoting, minimum viable products and analytics were just some of the themes that were buzzing throughout web 2.o Expo. What these subjects have in common and what was an overall takeaway, whether we are talking about ever changing technologies, markets or startups, is that a strong learning culture can be source of success, or at least to fail fast. I certainly will attend the next year’s web 2.o Expo.

Did you like this post? Please, feel free to subscribe or follow me on Twitter

What’s in a Startup Methodology?

Back when I did the transition from studying software engineering to entrepreneurship, one question kept coming to me – while methodologies are inevitable for successfully building software products, why isn’t there an integrated methodology for launching and taking that product to market?

Take for an example the Perpetual Beta and Spotify, a successful online peer-to-peer music streaming service. Spotify did not only use beta versions to test engineering requirements with early users – using beta-invites enabled Spotify to create demand for their service at the same time. Software developers have been embracing practices such as Continuous Deployment for years, but merely from a technical viewpoint. I believe that the Perpetual Beta represents a new line of ambidextrous practices that not only enables a startup to plan, test and build – but also serve, distribute and market their product at a lower cost.

Astonishingly, methodologies that help entrepreneurs facilitate software product development alongside commercialization are still scarce. At the one hand there exist a variety of methodologies to manage risk in agile software product development, including Scrum, Extreme Programming and Adaptive Software Development. However, none or few of these methods, to my best knowledge, encapsulate risk in commercialization. To simplify, think of it as Scrum + marketing (Scrumm). At the other hand traditional management practices have been argued not to fit the extreme uncertainty in startups, and often comes to short in terms of aligning with disruption driven by Internet technologies.

At that time I was also more than inspired by Crossing the Chasm that addresses the specifics of marketing disruptive high-tech products. However, I still find the Pre-chasm phase left unintended. In my last post I wrote about diagramming the product-market fit, in which I have aligned with the Technology Adoption Life Cycle below.

Pre-Chasm Startup Methodology

In many ways I think that new-product introduction is about ramping the Pre-chasm, where building and taking new products to market are not two separate activities. Obviously a startup needs ambidextrous qualities, and working engineering and marketing in parallel will enable a startup to discover the holy grail of product-market fit – a key tenet with Customer Development and Lean Startup thinking. Together with elaboration on 37signals’ Getting Real and Rework, I look forward to seeing what this emerging school of startup methodologies brings.

Did you like this post? Please, feel free to subscribe or follow me on Twitter

Diagramming Product-Market Fit at Lean Startup

Continual iteration is a fundamental principle in agile thinking as well as in startup methodologies such as Customer Development and Lean Startup. Basically, its premise is that a startup will mitigate risk and uncertainty by shortening product and customer learning loops, and adjust its product-market fit accordingly.

A startup achieves product-market fit when it masters the balance of building a solution, or product that acts on a customer’s problem, or vice versa. Both product development and customer development each has its own iterative loop structure. I believe that the two should be reciprocally acting and proceed in parallel towards the goal of product-market fit and that this makes a multiple-loop system. This invokes an ambidextrous challenge to early-stage ventures.

Building on my former post on Disciplined Creativity with Mihály Csíkszentmihályi’s Flow diagram, I would add to the Lean Startup model.

The diagram above shows product-market flow as a result of efforts in parallel iteration between agile product development (at the y-axis) and customer development (at the x-axis). In order to achieve a product-market flow state, that is product-market fit, a balance must be struck between customer development and product development. If a startup is drifting too far along one of the axis without iterating, flow cannot occur.

I believe that iteration beyond the product-market flow zone could be considered pivoting – that is when you change a fundamental part of your business model in regards to products and customers. To successfully iterate between product and customers and achieve product-market fit, you would develop a minimum viable product offering that enables you to learn about your customers needs and wants.

At startup you must pay close attention not only to the iterative tasks within customer development and agile product development separately, but also to the feedback loops in between the two. However, time is limited, and you should be aware of trade-offs in achieving flow in a Lean Startup. This is where continual iteration and validated learning allows for greater risk reduction under extreme uncertainty.

Did you like this post? Please, feel free to subscribe or follow me on Twitter.

6 Essential Startup Decks

Among the great many startup 101s’ that are out there, it can be hard to separate the wheat from the chaff. Here is an overview of the most essential startup decks that I have taken notes of. The list is merely a snapshot of hands-on startup methods and tools but a complete list of entrepreneurship literature.

Mint.com Pre-Launch Pitch Deck makes a great example of a startup pitch. For more on pitching  it is definitively worth looking into Pitching Hacks: How to pitch investors from Nivi and Naval at VentureHacks (PDF and free sample) to be used in companion with Dave McClure’s Startup Viagra: How to Pitch a VC.

Mint Founder Institute Accounting, this time by Mint founder Aaron Patzer, is a must read on how to draw your business case together and getting the numbers right.

Startup Metrics for Pirates by Dave McClure is the nuts and bolts of how to formalize traction and accountable metrics for your startup, including methods from the Lean Startup.

No One Cares About Your Stupid Little Startup by Xobni founder Matt Brezina takes us through its phases and tactics for building critical mass.

The Lean Startup methodology by Eric Ries combines the best from the agile development paradigm with Steven Blank’s Customer Development methodology (which might as well have been included) into a low-burn startup methodology.

Drop the business plan?! The Business Model Canvas by Alex Osterwalder, is an one-slide template and systematic approach to analyze, brainstorm and sketch out your business model. If you are not familiar with the framework you might want to start with the Business Model Generation book, the lean startup business model pattern and Steve Blank’s post on how to process it.

There sure are several interesting startup decks, and no such list goes without mentioning Guy Kawasaki’s The Art of the Start and Garage‘s Perfecting your Pitch. I will read SEO Moz’s Venture Capital Process again. Not to forget Getting Real and Rework – the business, design, programming, and marketing philosophies of 37signals’ (see aslo Eric Santos’ summary). Neverthless, keep up to date with Lean Startup news.

Do you know of other essential startup decks, we would love to know about them.

Did you like this post? Please, feel free to subscribe or follow me on Twitter.

Minimum Viable Product revisited – the MVP Curve?

My professor once introduced me to a concept he called FRUST – an acronym for frustration. Its premise is that products should be built solving a problem or pain for the customer. So is the minimum valuable product (MVP) a systematic, disruptive approach to product marketing. Google, Twitter and Spotify apparently get it. Say no more.

At its shortest, Eric Ries coined the minimum viable product:

.. the minimum viable product is that product which has just those features (and no more) that allows you to ship a product that resonates with early adopters; some of whom will pay you money or give you feedback.

In this Venture Hacks interview with Eric Ries, it is put even more simple:

The minimum viable product (MVP) is often an ad on Google. Or a PowerPoint slide. Or a dialog box. Or a landing page. You can often build it in a day or a week.

More recently, voices of the Customer Development/Lean Startup community have made an excellent effort in elaborating the idea, e.g. Andrew Chen; Minimum Desirable Product and Ash Maurya; How I built my Minimum Viable Product.

If a picture is worth a thousand words, then the minimum viable product would deserve its own least common multiple. Adapting the Featuritis Curve, here is a conceptualization as a basis of discussion.


The MVP Curve questions whether resonance with Early Adopters is relative to the number of features or amount of complexity offered. The minimum viable product does not necessarily mean that the product should be dead simple. Rather, the resonance with customers should peak when the product offering is designed to solve their core problems or jobs-to-be-done, as suggested by Clay Christensen. In accordance with the Customer Development model, this implies not only listening to the customer, but getting out of the building and carefully studying the customer.

I leave to you the question whether the minimum viable product can be conceptualized. In different use cases, what would be the pitch of the curve? Further, the curve might be tested by applying metrics to it: how to measure resonance with early adopters over features, and are there alternative variables?

Did you like this post? Please, feel free to subscribe or follow me on Twitter.