“Organizations which design systems are constrained to produce designs which are copies of the communication structures of these organizations.” – Melvin Conway
Coined in 1967, Conway’s Law neatly articulates the linkage between organizational structure and software development. While the connections between what is essentially a business abstraction and software architecture can seem non-obvious to those without exposure to software engineering, academic studies have supported the hypothesis, albeit in limited studies. Even absent that evidence, most of the industry would continue to take it for granted simply because it makes intuitive sense.
What about the world outside of the software itself, however? Are there oranizational constraints that impact the way that it’s marketed and sold? In general, the answer appears to be yes, in a particularly unsurprising way.
Much as software itself was once imagined to be an entity independent of mundane constraints such as reporting structures, so too is marketing commonly assumed to be a manifestation intrinsically tied to the product itself. But this is frequently, if not universally, not the case. Specifically, the marketing and messaging around a product is substantially more likely to be determined by the go to market for a product than the product itself.
In practice, this means, as but one example, that developer-oriented tools and infastructure are frequently not marketed as such, at least not directly. It shouldn’t come as any surprise that marketing is explicitly designed to service how a product is taken to market, but when the buyer for a product is distinct from a user the delta between the two can lead to substantial misalignments of messaging.
Imagine if Apple, for example, marketed the iPhone or iPad to the corporations that purchase them in volume rather than individual buyers. No “Shot on an iPhone” or “There’s An App For That,” but “Mac vs PC” with just PC. This is counterfactual to a degree that it’s nearly impossible to imagine, but it is the reality for thousands of enterprise suppliers whose product is built for developers but whose buyer has a different set of fundamental concerns.
Cloud-based services mitigate this to some degree, as the expectation that services will be free is less common than it is with on-premise software and therefore most users are buyers at least to some extent, but given an all cloud future is at best years away buyers are still typically a population distinct from users. Logically, go to market strategies are reflective not only of this dichotomy, but tend to focus on the population with budget.
The end result of which is marketing that speaks to organizations, rather than the individuals that inhabit them. Instead of Conway’s Law, then, we have something like the following:
“Organizations which sell systems are constrained to produce systems marketing which are copies of the go to market structures of these organizations.”
There is, to be clear, nothing inherently incorrect about this situation. Marketing needs to reflect and cater to buyers. But in areas in which the users are not the buyers, and yet have substantial influence over product selection, marketing cannot end simply with messages to buyers. Additional – and if necessary, entirely separate – messaging must be created to entice even users who have no intention or ability to pay for a product.
The last thing marketers typically want to hear is that they need to create not just one but multiple sets of product messaging, but when your buying and using audiences are fractured and have very different interests and preferences, this is what successful companies do. If you’re Apple, you can perhaps get away with a single marketing message, but if you’re here, let’s be honest: you’re not Apple.
