Pathfinding: The Direct Route To Value

Pathfinding: a process and a mindset to get to value

A word we use quite a bit when talking to prospective clients is pathfinding. We say that where others might talk about their process, we prefer to talk about pathfinding the way to value. Pathfinding is a subtle concept because it's not really a process as much as it's a mindset. Despite the subtleties though it does seem to resonate with most leaders. We don't get too many looks that say 

‘Who are you? Bear Grylls with a laptop?.’

Pathfinding and pattern based delivery

Here's a simple explanation of what pathfinding is and why it's so important as the structural backbone for pattern-based delivery. A common metaphor used to describe software projects is climbing a mountain. It's a good one because it brings together the sense of challenge, the effort, aiming for a summit and elation at the top.

But I don't think that's how it feels in practice. To me, it feels more like traversing a valley.

And that's because, when you start out, some people (the customer) can see all the way across to where they want to get to and you (the deliverer) have to descend into the valley and make decisions about each step to take, towards the goal, but limited by what the terrain allows. The first few steps will determine how hard the later ones are. And all the while you're in a dialogue with others whose view of the destination is unencumbered by these details.

If that's not hard enough, the target itself will likely change as more detail and data emerges.

Expectation vs Calculation

These dual worlds are the result of all the conversations that happen before work starts (expectation) running headlong into the resistance of the real world of legacy systems, complexity, politics, competing priorities, market forces, budget changes, etc (calculation).

Standard delivery processes don't have a predefined step called "So all your assumptions turned out to be wrong?"

They can't because the world of expectation is a conversation about what you need whereas, when you hit calculation, it's a conversation about what the work needs. And every project will be different. Similar maybe, but different. There's no handbook for crossing the valley.

Conventional delivery that operates with agility

 As described by Dave Thomas is:

1. Find out where you are

2. Take a small step towards your goal

3. Adjust your understanding based on what you learned

4. Repeat

And if you are unsure about which small step to take then take the one that's easiest to reverse if it turns out to be wrong.

I usually add "and never procrastinate about which step to take for longer than a step would take" because the important thing here is motion towards the goal. You cannot get it 100% right, but you can deal with a lot of uncertainty by just moving and remembering that steps 1 and 3 are really the most important ones.

The Double Loop

Pathfinding takes this idea a little further though. Because sometimes there's a flaw with step 2. Your goal is some form of value perceived by the customer. And delivery (calculation) has a habit of providing insights that both challenge that value and throwing up Unplanned Additional Value.

For example, say you are building an online pet store (and yes, if you are old enough to remember it I am thinking of the Java exemplar). You might discover some unplanned additional sales order value can be found by cross-selling pet foods with pet cages and hiding irrelevant pet accessories. You might also determine that the value goal has shifted because the market for hamsters is declining but kangaroos are all the rage. Both of these represent new work. Maybe a new pet/food taxonomy for the former and a new marsupial category for the latter. You are in the valley and the goal, and therefore the route, has changed.

Pathfinding adds a critical extra step:

1. Find out where you are

2. Question and restate the goal, the mental model we have of the route plan, and also the process we're using to get there (apply new delivery patterns here)

3. Take a small step towards your goal

4. Adjust your understanding based on what you learned

5. Repeat

This is called double-loop learning and it allows us to remind ourselves every week that we're not just adjusting and optimising our route to the goal, we're optimising our understanding of where the goal is and whether our assumptions, models, tooling, process, cadence, etc are still appropriate.

Obviously most weeks there should be no change. The subtlety is really no more than making something that's usually implicit (what the goal is) explicit. And of course, the moment you feel the ground shifting that's a prompt to redraw the map with the customer.

Pathfinding is important because it's the sync between knowing where you are going and knowing you are going there. It's not always a straight line but it's the most direct route to value even when the value is moving.