Small beats big – some thoughts on digital infrastructure

Some friends of mine are starting a small business soon. It’s a really cool idea, and I plan to help out with it as I can. As we talked about it on a hike this morning, it occurred to me that a lot of key factors that make the venture viable would not really have been possible until very recently. By taking advantage of free or very low-cost technology services (primarily in regards to distribution, marketing and commerce), they will be able to bring this awesome new product to market. Who knows where it could go?

roadsI’ve actually been thinking a lot about infrastructure lately. Public infrastructure, as one example, has always been a hobby horse interest of mine. We have tragically under-invested in public infrastructure here in the U.S. over the last few decades, which is directly harming our national competitiveness and economic growth even as borrowing costs are at historic lows… but others have written much more eloquently than I can about the importance and challenges of public investment in roads, bridges, ports, telecommunications, etc., so I won’t go into that here.

Infrastructure socializes costs and allows users to privatize most of its benefits (minus obvious costs in the form of taxes). In this way, much infrastructure is inherently more valuable to smaller constituents than big ones. If you could afford to build your own road to bring goods to market, then a public road is still a useful option as a cost-savings measure, but maybe not a must-have; but if you’re a small producer, then a public road is likely the difference between bothering to produce (or invest) in the first place or not. Good infrastructure makes critical steps in value creation easier and cheaper.

Interestingly, that is exactly what’s happening in technology today. Web application and development is becoming streamlined and standardized; instead of setting up your own datacenter and standing up servers, there’s AWS and Heroku; Google Apps, Salesforce, MailChimp, HubSpot, Skype, Slack and any number of other cloud-based services are making it easier and substantially cheaper (sometimes free!) to do the important stuff that used to slow down small businesses and divert their attention away from their product. Not only do these services make the same old stuff better, faster and much cheaper, but in many cases, they allow users to do whole new things too. It’s digital infrastructure.

This is why a lot of the old thinking about how incumbency and size provides such a huge advantage is increasingly outdated: tech is steadily unbundling core components of starting, running and growing a business, and in the process, lowering entry barriers. You saw this first with communication (email, IM, Skype, VOIP, chat), then in IT (storage, hosting, productivity), it enveloped marketing and sales (analytics, CRM, automation, personalization, etc.), and is now literally everywhere you can name: HR, finance, operations, distribution, merchandising, design, more. Using mostly cloud-based software, businesses are able to automate more of these operations on demand, reducing their cost and organizational drag.

Like its physical counterpart, digital infrastructure directly and primarily advantages the small guys – the ones who once might have had a great product idea, even a prototype, but who couldn’t overcome big fixed start up costs. When it becomes cheaper and easier to try to launch a new idea, more entrepreneurs will do so. Naturally, this pushes up the percentage of startups that fail, but it can just as easily reduce the cost of that failure too.

This puts many big companies in a difficult position. A lot of big companies have made huge investments in their own internal business and IT “infrastructure” which is now outdated, unnecessary or simply irrelevant. Particularly for older companies, much of this infrastructure brings with it legacy processes, operations and internal advocates that have a vested interest in continuing to exist. (I’ve seen many arguments against “the cloud” that can be attributed almost exclusively to this.) Over time, this is how big companies accumulate technical debt. Remember – lots of companies still use Lotus Notes.

We like to think that “the best product” always wins in the marketplace – but everyone knows that this isn’t really the case. There are an awful lot of steps in between building a great product and making it successful in the market: sales, marketing, distribution, and foundational services like finance, operations, manufacturing, etc. To the extent these intermediate steps get cheaper and streamlined, it allows a firm to focus on what actually matters: its product and customer experience. Big companies can certainly still do this, but probably not without overcoming a good deal more internal inertia than used to be the case.

buggyWe’re still in the early days of the internet, and likewise, our digital infrastructure is still in rudimentary form. By analogy, we’re really still in pre-interstate highway days. In ten years, people will doubtlessly look back in wonder that we were able to get anything done with the very basic digital tools at our disposal – and yet, compared to where we were just a decade ago, the capabilities at our disposal now seem incredibly powerful. It’s a great time to work on building cloud-based enterprise software.

Of course, even while our digital infrastructure is improving by leaps and bounds, the state of our physical infrastructure is in dire straits. The power of the former will be inevitably hampered by the latter – you just can’t skip building decent roads – and I do worry about the ability of our political system to meet the bare minimum requirement of rationally allocating its resources. But, again, that’s probably a topic for a separate post. (Or maybe a different blog!)


So, what do you think ?