Differentiation, Or "2 Commoditisation 2 Furious"

A colleague of mine (hi Jake!) who subscribes to my newsletter mentioned to me recently that I use the word commoditisation a lot in many of these posts. He is absolutely correct. I think elaborating on why this is could be quite helpful to the ongoing series about digital products and their factories. This topic also has connections to the field of truth and primitives.

At a very high-level when people consume mass-produced goods, they most often do so through the lens of commoditisation. They start with a particular need, and they engage with a field of available products in a market that look like a solution. While there may be many acceptable solutions, if you were to compare two products you would find that there is probably one or more dimensions in which they are different. One may be cheaper. The more expensive of the pair may look and/or feel nicer. The points of comparison will vary depending on what the product is.

More complicated problems will have more complicated solutions. We may mix and match a basket of commoditised goods, for instance in a household we may select a different can of diced tomatoes or replace batteries in a television remote with ones produced by a different company, in either case these fit into established production processes, in the first case this would be a recipe, in the last this would be an entertainment system.

We can apply this concept of commoditisation at this larger level of granularity as well. "A recipe". "An entertainment system". These are not descriptions of a unique entity, they are descriptions of a field of things. "A recipe for pizza" is more specific than "a recipe", but it is still describing a smaller field. "A recipe for Domino's pizza" is more specific than "a recipe for pizza", but it is still merely reducing the size of the field. A recipe for Domino's Hawaiian pizza. Now we're getting pretty specific.

Domino's is a global chain, and they may offer Hawaiian pizza all over the world, but there is little to guarantee that the allegedly identical product has the exact same supply chain, machinery and production processes involved (at the very least we can safely say that the humans involved are not identical at every stage) everywhere that it is offered for sale - and therefore even for the "same" product, there will be some level of differentiation. This is before we account for the reality that any of these elements may evolve over time, for instance due to cost optimisation, forced changes to the supply chain, attempts to improve the product in some other manner.

The people who ultimately consume Domino's Hawaiian pizza might start their journey specifically seeking Domino's Hawaiian pizza, or Domino's pizza, or pizza, or much broader still, "lunch" - where the field of solutions is much broader. They engage with the market with different models of "the problem" as well as different models of "the solution", or in other words, they engage with the market to varying degrees of commoditisation.

In the physical world we can view products as commodities nearly everywhere we look, and in many domains of life these are relatively stable in time with social relevance that may span decades, centuries, millennia - it depends on what class of product we are referring to, and to what degree differentiation is important. I was writing about pizza earlier but I could have also been talking about soda. Processed timber. Glasses. Backpacks. Sunscreen. Boats.

It is a relatively rare occurrence to add new primitives in the physical world, but it does happen when there are technological advances with significant social utility. Closer to the world of computing, over the last few decades we have seen it with relatively high frequency. The desktop computer, the laptop, the internet, the search engine, the video game console, the smartphone, the cloud, the streaming service, and recently "the agent". There are many more examples. In computing, there will probably be more in the future.

In software, differentiation is a powerful force in both the concerns of the digital product and the concerns of the digital factory.

Software developers make technology choices based on requirements that are often provided to them in a piecemeal fashion as per breaking the work up and specifying the product. They navigate the field of available primitives in the global digital supply chain and make decisions based on the known requirements. As mentioned in the article introducing the digital product, often requirements are introduced that if known earlier, would have resulted in different technological choices. This is a consequence of building strongly differentiated products, of participating in a market where integrated value-chains are largely governed by rules of ownership.

After having released software, developers often face challenges communicating differentiation. It is sometimes extremely difficult to clearly communicate which problems a unit of software is attempting to solve, and which problems it is not attempting to solve. Users naturally place software into categories, and modern storefronts (where software is obtained by users) reinforce this mode of thinking by doing the same.

Many modern applications tend to attempt to solve a range of problems, and while at some level of commoditisation (e.g. "Travel app") two apps may be compared, they may have significant differentiation in terms of functionality. This is really consequential and the source of a lot of the pain between users and the digital experiences they have, and a clear break between norms in the physical world and the digital world.

There are many driving forces behind this phenomenon. For economic reasons (related to revenue), as a developer, you want to attract as large an audience as possible. For economic reasons (related to time), as a user, you want to solve your problem, you generally don't really want to invest a lot of time reading about solutions and their limitations. Often there can be a great deal of overlap in the problems two "similar" units of software attempt to solve, which only makes it a greater source of pain when we land on the wrong side of differentiation.

Until next time.