Brief note on innovation, distribution, and moats

Distribution >> technology

A friend wrote a post in which he argues software complexity predicts economic growth; he suggests that more complex technology confers a deeper moat. I disagree with this thesis: I don’t believe technology confers a moat. Technology, unless patented, is easy to copy; this is especially true of software (which doesn’t even require significant capital expenditure to produce), and even more so in the age of AI, when computers will soon be able to reproduce code cheaper and faster than humans.

It is an unfortunate habit of engineers, tech-enthusiasts, VCs, and product managers to overplay the importance of product, and downplay or completely overlook distribution. The fact is, having a strong distribution footprint - i.e. having a product available in the channels in which consumers are likely to purchase or use a good or service - confers an almost unsurpassable moat. (This definition of ‘distribution’ is also overlooked - people sometimes think of distribution in very traditional terms, like brick and mortar).

Existing distribution channels are very hard to break into - even for companies that have all the resources to do so, and have a foot through the door, too. For example: when I worked at P&G, we launched a new product line to compete in the adult incontinence category, which had long been dominated by TENA. We strongly signalled commitment to the market: we invested God-knows-how-many millions in setting up factories and new production lines, as well as in marketing (including costly TV campaigns); despite this, and despite the fact that we already (obviously) has strong relationships with retailers, we struggled to convince supermarkets to give us as much shelf-space as they gave TENA. Inertia, force-of-habit, and the instinct to stick with what we know and works are strong forces.

Which is why when start-ups beat incumbents, they do so through disruption, not superior technology (and, by the way, it’s maddening that even though the Innovator’s Dilemma was published a quarter of a century ago, people still cannot distinguish between disruptive and radical innovation): disruption circumvents distribution advantages because it either invents entirely new channels (e.g. e-commerce vs brick-and-mortar, or DTC vs retail model) or because it focuses on an underserved market segment, whose members will go out of their way to purchase a new product that meets their needs better (and will thus eschew distribution channels that do not stock the product).

This is not to say that technology doesn’t matter; incumbents who don’t innovate will lose market share over time. But this is not the same as saying that these companies’ technology produces a moat; it just means that companies require innovation to stay in the game. (At most, one could argue that a company’s ability to innovate, anticipate customer needs, and bring products to market efficiently yields a competitive advantage, because these things will help the company increase prices and margins; but the ability to do these things is not necessarily a function of technological complexity.)

All this said, I agree that software (and tech) complexity probably predicts growth (both at the company and nation level) - but I don’t think there’s a causal relation there.