Wednesday, March 20, 2013

Pushing it to the Nth Degree ...

I've been meaning to write something more on the whole Pioneer, Settler and Town Planner structure but things have got in the way.  Ok, first a big dose of warning.

Whilst writing out maps and using them as a visualise aid in determining strategy is something I've done for quite some time and is based upon a fairly solid ground, this next bit is pretty shaky i.e. it's based upon very few data samples.

However, it has worked effectively in overcoming many common business problems such as alignment etc. The problem of course is few data samples and the danger of backward causality i.e. company A does X which is successful therefore X must be applicable to company B.

So you need to take the next bit with a huge pinch of salt!

Let us assume you've gone through the exercise of mapping out value chains, identifying areas for efficiency, common components (services) and created some wicked strategy mixing all the goodness of ecosystems (models like ILC) and the various attacking, defensive and dark arts.

You now need to organise yourselves. 

Step 1 - Eat Pizza
Rather than the one size fits all approach or breaking down into silos based upon type of activities (i.e. finance, marketing, IT), I'd argue for use of a more cell based structure such as the Amazon Two Pizza approach of small teams connected by services.

If you've mapped out your environment, built a strategy around "why" and created a cell based structure you're in pretty good shape. You can always go and top up with more of the Next Generation practice.  Going further is a step into the unknown. So, I'll guide you with what I know but I can't guarantee the results.

Step 2 - AGAIN - Plunging into the Depths
Ok, first thing go and grab your maps and count up the number of activities, practices and data sets that you've mapped out at different stages of evolution. This is actually a good time to find and take out duplication and bias between those maps. Anyway, once you've counted up then ...

Step 3 - Plot the frequency
Plot a graph of the frequency of components i.e. number of components at each step of evolution / total number of components.  What you're aiming to build is a profile for your company. As a side note, if you've been evil and built maps for your competitors, you can build a profile for your industry as well.

Now, as things evolve (through competition) then the profile moves. However some industries are more weighted to commodity relative to genesis than others.

Step 4 - Prepare to run screaming
Ok, this bit is a huge leap of faith and the evidence for this is scant and certainly not in the hundreds of examples I'd prefer in order to show any statistical significance. However, nevertheless I've used it in the past and it has worked well but that statement means almost nothing (single example, backward causation etc).

Despite everyone telling you that you need to be an innovator or a company needs one culture, beyond the cultural bits about treating your staff decently I'm going to argue the general perception is wrong. In fact you need at least three different types of people and three different sub cultures to create a sustainable organisation. The groups, I'll refer to as Pioneers, Settlers and Town Planners. 

This is the structure I started to implement in Fotango about 2005. In its fullest implementation you need these three groups and a Chief Pioneer, Chief Settler and Chief Town Planner (though at that time I didn't call them that). Everything else is gone ... no finance department, no IT department, no silo structure based upon type of activities. 

Everything here is based upon evolution.

The balance between these three groups is determined by your profile  i.e. if your industry and the actions you are taking (i.e. activities in your maps) are weighted towards the more commodity activities, then you'll need a larger Town Planning group. Obviously you can decide to manipulate this by focusing on one area but the point is to get a relative balance.

So who are the Pioneers, Settlers and Town Planners and how does this work? 
First, the model of operation is based upon theft i.e. Pioneers create the novel and new (as in genesis), Settlers steal from the Pioneers and grow the market, Town Planners steal from the Settlers and commoditise it. This process forces the Pioneers to use components from the Town Planners in creating new higher order systems and hence a virtuous circle is created. You can do this with Ecosystems (i.e. the ILC model) and the whole point of chowing down on the ecosystem is to force it to move forward (genesis). This post is however an example of how to run this internally.

The characteristics of each group are different, so is the sub culture they will create and the people you need. I've added some details related to a more technology example below. Your profile will give you an idea of how many of each you will require.

Step 5 - Populate your cells
Ok, now you've got an idea of the type of people you're looking for and how you can create a virtuous circle of theft, now populate your cell based structure with the right type of people.


Here's the shocker. If past experience is anything to go by (which it might not, a normal caveat of too little data) :-
  • All three groups are valuable and you need equally brilliant Pioneers, Settlers and Town Planners.
  • You can accelerate your rate of genesis (creation of the novel and new) and efficiency at the same time.
  • People did seem to happier in this structure i.e. it seems to fit to some deep way of operating. There does appear to be people who like to focus on the well structured, well ordered and use their brilliance to improve. There are others who seem to enjoy the completely chaotic, failure prone and occasional areas of brilliance.
  • Business alignment is simply an artefact of our current organisational structures.
  • You can permanently eliminate inertia to change through this cycle of theft.
Now before you rush off and do this, remember the evidence for this is scant and you're talking handfuls of examples. It's a gamble. The results might well be a statistical fluke.