Thread by @swardley

Excellent talk by @tehviking on hacking the tech treadmill with pioneers, settlers and town planners (PST) - ... but word of caution, the hype cycle does not show evolution but hype and this can be dangerous to mix with model such as PST ...

... to explain why, let us take a map of ERP. We can even use this map to anticipate what next for ERP. There is consistency of movement ... take the bold ERP, 2002 ... you know what happens next, more industrialised forms ...

... because of this consistency of movement, we can determine patterns and structures to cope with constant change. Hence the Pioneer, Settler, Town Planner model ...

... so, if it's 2002 and I'm looking at ERP, I know where it is going and what sort of attitude I'll need to manage it - more settlers as ERP consolidates and finally town planners as it turns more commodity / utility ...

Now, let's change the axis. Let us not use evolution as the axis but instead the hype cycle. For interest, I went through this pain long ago - blog.gardeviance.org/2015/03/evolut… ... but let us return to using our ERP example ...

... I've changed x-axis to hype and positioned nodes based upon where they appear on various dated hype cycles. If I'm in 2002 then Enterprise ERP (2005) is to right but Cloud ERP (2011) is to the left, there is no consistency of movement. Try placing utility ERP on this "map"?

The problem stems from the same issue with trying to map with diffusion curves. Evolution consists of multiple diffusion curves and also multiple hype cycles (assuming these are in any way real) - blog.gardeviance.org/2011/02/decons… ...

... the "hype cycle" map is not in fact a map as there is no consistency in movement. It's a bit like a map where if you walk 2 miles North then you end up 15 miles South West of where you started ...

... where this creates big problem is if you add a PST model to it. Let's add PST to the hype cycle "map" - now we have no idea where utility ERP is going to be, so let's deal with what we know. We end up with pioneers building a cloud service where we want industrialisation.

This flaw can be more clearly seen by comparing the map and the hype cycle "map" next to each other. By not using a map based on evolution we lose both anticipation and ability to spot & apply evolutionary patterns such as PST.

The x-axis on the map isn't random, it took over 6 months of hard work to create it, over a decade ago and it has stood the test of time so far. Be careful when changing it. Think about movement and consistency of movement. You'll need that unless you want a right old pickle.

Don't confuse evolution with either diffusion curves or the hype cycle. They are not the same. #WardleyMaps

Oh, and "Wardley’s concern is that two modes aren’t enough. He proposes three" - issurvivor.com/2015/07/27/tri… ... for reference, Pioneer-Settler-Town Planner was implemented 2005 - 2007 after a previous failed two mode effort and long before "bimodal" was a twinkle in Gartner's eye.

I still however cannot emphasise enough that embarking on any such organisational change to create a complex adaptive structure is pointless until you have the basic principles of doctrine in place (all the way to Phase III)

... far too often I see execs fiddling with organisational structure as though will magically make all things well whilst lacking any of the basic principles such as challenge, learning, user needs etc. It's like moving the deck chairs around in order to stop the Titanic sinking.

Also please don't try and mix PST with the Horizon model. Let's take a map from 2005 (when I was CEO of an online photo service) and to this map add Horizons ...

Now, let us add PST to it. As you can see my Horizon 2 is full of town planners (industrialisation of compute) and Horizon 3 is full of pioneers (the things built on cloud). But this is 2005, pre launch of AWS ...

Now, let us roll the clock forward to 2010. Let us take the role of the cloud provider. As you can see our horizons are different. Horizon 2 is full of pioneers, Horizon 3 is full of settlers (those new higher systems built on top of cloud starting to become products themselves).

There is no link between Horizons and PST because Horizons vary with where you are on the map and PST is based upon evolution i.e. components at this stage of evolution are suitable for this form attitude / culture.

This also doesn't mean that PST is right. It's just the method I found useful to organise back in 2005-2007. It does require a lot e.g. maps, the principles, the system of theft. The only thing I'm confident on is "bimodal" is flawed - blog.gardeviance.org/2014/11/bimoda…

Give it another decade, with examples of success and failure, with work such as @KentBeck 3X (a similar model) and examples like "Power of the Elastic Product Team — Airbnb’s First PM" - firstround.com/review/the-pow… ... and we'll have a better idea.

Of course, my two favourite articles on PST are ...
@drewfirment cartoon - cloudrumblings.io/a-pioneer-a-se…

and the exceptional article by @andy_callow on exploring pioneers, settlers, town planners which has the truly marvellous video by Mia -

see - medium.com/@andy.callow/e…

There are also lots of interesting people out there exploring the whole interaction of evolution and organisational design e.g. @markusandrezak - gotocon.com/dl/goto-berlin… or @chadcthomas or ... well, there are many.

Oh, and if it isn't obvious ... you need all three - pioneers, settlers and town planners - but not necessarily in the same company. To start with, you can't pioneer and explore the adjacent unexplored without town planners having industrialised what was once novel i.e. ...

... when you look at a map and see higher order systems built on industrialised components, you might just think "Town Planners" here and "Pioneers" here but ...

... the map is richer than this, because we can show the past and the future. It's all connected. You needed more industrialised pulley systems (the Plymouth system) to build the ships we needed to navigate, explore, colonise and create trade with different lands.

Yes, you can explore using underlying components that are not industrialised however it is more costly, less effective, less stable and with a few more layers breaks down due to unreliability - a tower of Babel. Our progress depends upon industrialisation of lower order systems.

In short, you need all three - pioneers, settlers and town planners - to create a sustainable system unless you wish the entire system to eventually collapse ... but ... and there's always a but ...

... you don't have to do all three yourself. The purpose of the ILC model that I wrote about over a decade ago was to get others to explore (to Innovate for you), to Leverage that ecosystem to identify patterns and to Commoditise. Hence ILC - blog.gardeviance.org/2014/03/unders…

... the effect of ILC is it enables you to increase innovation, efficiency and customer satisfaction all at the same, contrary to old ideas of you had to focus on one. As you do so, you're constantly building up the industrialised value chain e.g. look at cloud or serverless.

... pioneers and settlers still exist under ILC but you push them outside of your company. You get others to innovate and listen to them grumble "they've eaten my business models" as you harvest the ecosystem to the delight of everyone else who builds on your newly components.

There are many ways to play this game, to exploit the environment ... but this all assumes that you map - medium.com/wardleymaps

If you don't map, then ... well, this probably seems like indecipherable magic voodoo to you. C'est la vie.

The model of Pioneer, Settler, Town Planner that I used within a business 2005 - 2007, of course replicates how we work as a society - it doesn't matter whether we talk about the history of McMurdo station in Antartica or the running of a software company or policy in Government.