That is the question that I’ve been putting a lot of thought into recently. Two ideas collided this week that gave me some insight into the question. First, in the comments on my post about different business models leading to different outcomes, Greg Satell said:

A business model changes over time, while architecture tends to be much more sticky. You can’t adapt it to market conditions. You are pretty much what you are.

I think a business model must be more maleable than organizational architecture. Value shifts and so does the basis of competition.

The second idea came from this tweet by Alex Osterwalder:

Greg’s comment is interesting, because while it is true, it is also true that organizational architecture is part of your business model. Which means that if you want to change your business model, then you will probably need to change your architecture.

While challenging, this is pretty doable for startups, and there are a lot of resources designed to help you do precisely that.

The statement by Osterwalder is also interesting – because there is nothing inherently dynamic in the business model canvas. In particular, when it is applied in larger firms, it can be just as static as any other planning tool tends to be.

Paul Hobcraft captures these issues:

Is it also because it might be hard to imagine that you can capture a complex picture within a one page canvas? Within large organizations you deal in pages (I mean pages) of business plans, marketing plans, briefing papers and countless justifications. Could a business model be captured so easily? Well actually yes, if anyone is wondering.

Larger organizations are slower in adoption; it is just a matter of course. Be this in new technology, new processes, social media or a host of new things they tread more cautiously, they are slower to adapt. Those that get ‘it’ internally spend often frustrating months, sometimes years, trying to convince others to make a change, to change a process, a planning structure, even a widget within an established product that often holds it back.

This is a general problem in larger firms – innovation can be generally harder in many ways. And in business model innovation, the challenge is changing some of the things that are deeply embedded – like organizational architecture.

So what can we do?

Osterwalder’s quote provides some guidance – start experimenting.

There are several good tools around now that can help you design and evaluate business models. This one is a good example of one that is essentially static. It’s a nicely constructed tool, but it’s essentially static.

However, there are two nice tools that allow you to build some dynamics into your business model assessment. The first is the Lean Launch Lab, which is built around the Business Model Canvas:

The second is the Lean Canvas tool, based on the work of Ash Maurya. In assessing the Business Model Canvas, Maurya replaced four of the boxes from the Business Model Canvas with things that he and his community thought were more useful for startups, adding Problem, Solution, Key Metrics, and Unfair Advantage:

Both of these tools are free for only one user, and have a monthly fee if you want to have multiple people contributing to the development of a particular model. Both of them have one feature that is great: experiments.

Both tools have been heavily influenced by Steve Blank’s approach to using Business Models in customer development – and experimenting is central to this approach. It’s useful to think of this as hypothesis testing, and it’s something that you can do in both startups and larger organisations.

There are several interesting points that follow from all of this:

• Finding the right metrics are a critical part of experimenting. For example, if you are building a business model for a web-based product or service, Rob Fitzpatrick points out that using the standard metrics (traffic and revenue) will lead you astray. Instead, he says:
  

  The lean startup book’s engines of growth is a great starting point for picking good metrics:

  Growing virally? Focus on referral.
  Users never leave (e.g. subscription webapps)? Focus on retention.
  Buying new customers via ads or salespeople? Focus on cost of acquisition versus lifetime value.

  If those aren’t relevant then take your pick from amongst McClure’s pirate metrics, Ellis’ 40% disappointed test, the net promoter score, or something of your own creation.

  The hypothesis testing approach works when you establish a hypothesis about what you should be doing, figure out a way to test it, measure appropriate metrics, learn and adapt your approach.

  Everyone should be doing this – and it is built in to both of the LeanLaunchLab and Lean Canvas tools.

• Larger organisations need to pay attention to their value network. Maurya hypothesises that this is less important for startups – a proposition that should probably be tested. However, this is where the organizational architecture question that I’ve been discussing with Greg comes in. As I discussed in the post, how you structure the value network can have a huge impact on how profitable you are. But it’s hard to change.
• A great value network is an excellent example of an unfair advantage, which you need to succeed. In an outstanding post, Jason Cohen describes the types of things that make up an unfair advantage, and why you need one:
  

  To summarize: Anything that can be copied will be copied, including features, marketing copy, and pricing. Anything you read on popular blogs is also read by everyone else. You don’t have an “edge” just because you’re passionate, hard-working, or “lean.”

  The only real competitive advantage is that which cannot be copied and cannot be bought.

  What might these be? An impossible to replicate value network is one. Others that Cohen describes are: insider information – holding unique knowledge, single-minded, uncompromising obsession with one thing (think Apple + Design), authority, a dream team, strong relationships with existing customers.

  The interesting thing about all of those things – they take time and effort to build. This is an area where existing organisations have a huge advantage over startups.

Business models are an invaluable tool for organisations, and business model innovation is a particularly powerful form of innovation. To successfully innovate a business model, you must use the tool dynamically.

This is particularly important in larger organisations. Their business models can become stagnant for many reasons. As we’ve discussed, having an well-established value network is one of them. Using the business model approach to test the basic assumptions behind your approach is one good way to break out of this rut.

Building dynamic business models requires a hypothesis testing approach to support experimentation. And if you get it right, over time you can build an unfair advantage yourself.

Note: The conversation on the previous post obviously triggered some thoughts – Greg has written an excellent post on the topic today too, which you should check out.