On pace 🔥
Often, we need things to go at the same pace. In a product team, for example, we want to be working out what to build (Discovery) at roughly the same speed as we are building it (Delivery). If we work out what to build too slowly, we will end up building from guesswork. And if we build too slowly, we will find ourselves with a large stock of blueprints that never return the cost of creating them.
The product team example shows a system with opposing interactions on a stock – in this case the stock of things worth building. Another example is retail stores, who have to replenish their shelves at the same rate as they are depleted by sales.
Both product teams and retail stores will self optimise to try and keep these flows in balance. When customer demand changes, a store might alter its order from the warehouse. If a product team finds itself without enough great ideas to build, it will turn more of its effort over to discovering new ones.
There are limits to this self optimisation. For instance, a product team can’t usually assign itself more staff. It also can’t change the size of the shared service teams around it, like site reliability engineering or customer research operations. It’s down to management to use these levers to help teams avoid bottlenecks.
The state of balance that we aim to achieve could be called a productive equilibrium. It’s superficially similar to a market equilibrium, in that supply and demand are matched, but it’s achieved through planning rather than pricing. Many other useful terms, including stocks and flows, are laid out in Donella H. Meadows’ classic book “Thinking in Systems.”
Two important stocks in product development are the stock of things we have already built – our product – and the things we plan to build – our backlog. Both have an important characteristic: they reflect the state of our company’s goals at the moment they were created. If our goals change, there’s no guarantee stocks will remain valid. This happens all the time, and for good reason; changing the product is often the aim of changing the goals.
This explains the inherent danger in long periods of discovery focus. They may generate a valuable stock – blueprints for what to build – but the value of that stock is contingent on it being delivered one day. The longer it sits on the shelf waiting to be delivered, the more likely it is that a change in our goals will force us to write-down our investment in creating it.