stratechery
stratechery

Business Models for 2014

Perhaps the greatest continuing disconnect between technologists and normals is the value ascribed to software. As Marc Andreesen famously declared, “Software is eating the world,” and while that is absolutely true, software, especially in the consumer space, is not eating value, at least not directly. In fact, if 2013 taught us anything, it’s that the value consumers place on software is lower than ever.

This phenomenon was seen most clearly in the app store economy, where paid downloads plummeted in popularity. Multiple developers have written about how increasingly difficult it is to make money on the store, and despair about the relentless march to free (with in-app purchase). I myself have joined in, excoriating Apple for not doing more to help developers create sustainable business models in the app store. Ultimately, though, the removal of friction from the software buying process has – just as it did for music and other media – reduced or removed the willingness of many consumers to pay.

It’s telling that software and music share a common heritage: round, shiny pieces of plastic in a shrink-wrapped case. While the very idea of driving to a store and buying a physical object already seems absurd, it seems it was the physicality of that CD – and disks and tapes before that – that drove customer willingness-to-pay. Consumers may not understand at a technical level what is happening when they download an app or a song, but they seem to implicitly know that they are simply copying bits, a process which has a marginal cost of zero.

As I wrote in Open Source Apps, over time the price of a product moves to its marginal cost, and if the marginal cost is zero, that means free is inevitable.

As competitors enter the market, the price moves towards the marginal cost. In the case of software, that is zero.

As competitors enter the market, the price moves towards the marginal cost. In the case of software, that is $0.

What makes the software market so fascinating from an economic perspective is that the marginal cost of software is $0.2 After all, software is simply bits on a drive, replicated at the blink of an eye. Again, it doesn’t matter how much effort was needed to create said software; that’s a sunk cost. All that matters is how much it costs to make one more copy – $0.

The implication for apps is clear: any undifferentiated software product, such as your garden variety app, will inevitably be free. This is why the market for paid apps has largely evaporated. Over time substitutes have entered the market at ever lower prices, ultimately landing at their marginal cost of production – $0.

On the flipside, though, tangible products, which by definition have marginal costs of greater than $0 – continue to be valued by customers. No one expects a free microwave, or car, or even a candy bar. Consumers understand that making, packaging, and shipping such products costs money, and there is no compunction to spending money for that proverbial latte that is more than an app.

This is critical to understand while thinking about consumer business models: consumers pay money for tangible goods; they don’t for virtual goods (in-app purchases for games is a glaring exception here). It then follows that the two primary ways to make money in the consumer market is through 1) the sale of physical devices and 2) advertising (usually for physical goods). Ultimately, though, success in either area is dependent on software. It really is eating everything.

Physical goods.

Perhaps the classic children’s gift is the remote-controlled car. Who doesn’t have memories of eagerly eyeing the Christmas lineup at your local Radio Shack, eagerly opening your presents, scrounging for batteries, and finally driving your car like the maniac you are. A classic physical device, sold for cash.

It’s this context that makes sense of Apple’s odd opening to last year’s WWDC. The only app demo was of Anki, today’s version of the remote controlled car. Of course these cars are nothing like those pieces of junk at Radio Shack: Anki cars have artificial intelligence, the parameters of which are set with an app. The app, obviously, is free, but the value of the software within the app is realized through the $199 price of Anki drive.

This is one of the primary ways that software will be monetized going forward: hardware sold at a significant margin that is justified by the differentiation provided by software. In fact, I just described the business model of Apple itself. The physical components of a MacBook Pro or iPad have a marginal cost that is significantly lower than their retail prices; the margin is provided by software.

Another great example – and particularly relevant for this blog and this topic on this blog – is Paper. Last year I decried the fact that Paper couldn’t build a sustainable business on the app store:

With a small amount of seed funding, the original five employees set out to build Paper, the best place to capture your ideas. After launching a year later, they began to reap the rewards through in-app purchases. They kept improving the app, and came out with a significant update in October – the Mixer – for $1.99.

And then they realized that they were five people living in New York City without an obvious route to sustainable revenue.

Instead they are building physical devices: while they may have had limited success charging $4.99 for an in-app purchase, they are presumably having much more success charging $59.99 – including a substantial profit margin I’m sure – for a physical good, the Pencil.

Advertising

The size of the consumer market in technology is greater than it has ever been before. Not only do more people have access to computing devices, but, thanks to mobile, a much greater percentage of a person’s time is spent with computing devices. And, crucially, there is one industry that values consumer time and attention far more than consumers themselves do: advertising. This reality is at the root of Facebook’s resurgence over the last year, and for the initial bullishness on Twitter’s IPO.

The continued strength of the advertising model – and the data collection and analysis that goes along with this – is the flipside of consumers’ unwillingness to pay for software: consumers demand great products such as Google search, Facebook, or Twitter, and have demonstrated clearly that they will tolerate advertising in exchange for getting such experiences for free.

In fact, I think one of the biggest misconceptions in technology is that companies like Google or Facebook will inevitably irretrievably damage their products because their true customer is the advertiser. To be sure, there is a cost from advertising – this recent WSJ profile of Facebook said mobile advertising reduced activity by 2% – but that is necessarily countered by a focus on improving a product more rapidly than advertising degrades it.

Consider the incentives:

  • Company A competes in a marketplace dominated by paid products; the fact you have to pay to switch builds a bit of a moat around their products, reducing the incentive to improve them rapidly
  • Company B competes in a marketplace dominated by ad-supported free products; switching costs are minimal, creating huge incentives to ensure the impact of advertising is much less than the increase in value of the product

In short, I am not at all certain that advertising breeds poor products; it is possible it does just the opposite. What I am certain is that advertising as a way to monetize consumer-based software is here to stay, and that any service that can capture an outsized amount of consumer attention – e.g. Snapchat – will be very valuable.

Software as a Service is for Business

It’s almost cliche at this point to talk about the consumerization of IT, but, as I wrote last year, this is a very product-centric view:

Phones are the clearest manifestation of this trend; the iPhone, Android, and even Windows Phone are first and foremost consumer devices; it’s corporate-focused BlackBerry that is all but finished.

But while this trend may be true when it comes to the technology itself, the differences in business models and go-to-market strategies seems more distinct than ever.

Start with advertising: no business is going to rely on an advertising-based service for critical line-of-business needs. Beyond the lack of professionalism, data security concerns makes any consumers service a non-starter.

Devices, though, are becoming just as distasteful, especially for publicly traded companies. Devices such as servers are a capital expense, and have a significant impact on the denominator of the most important ratios used to evaluate companies: the return on invested capital.

Instead, it is businesses who are increasingly willing to do the exact opposite of consumers and pay for non-tangible goods in the form of software-as-a-service:

  • Software-as-a-service, as noted, has a positive impact on a company’s financial structure, and means more stable and predictable operating costs
  • Businesses still typically have centralized purchasing which responds well to productivity arguments. making it easier for a SaaS company to articulate the value of software. This is a marked contrast to consumers who can only be reached through blunt mass-market advertising. It also means that successful SaaS companies will have well-developed sales teams
  • Businesses often require customization or have specific feature requests, and are willing to pay for it. Successful companies balance responsiveness with a clear vision about where their customers ought to go; the fact SaaS is cloud-based and thus easily and quickly iterated upon makes this balancing act much more possible than on-premise software with multi-year upgrade cycles

That said, some of the dynamics that apply to consumer-focused advertising-based businesses apply to SaaS as well: specifically, switching costs are lower, and employee knowledge of alternatives is likely to be greater. This will continue to drive a much higher standard for business software, magnifying the challenge faced by traditional vendors who have focused on maximizing lockin and multi-year engagements.


To be sure, other business models will continue to exist, including paid downloads, consumer-focused software-as-a-service, and hardware for businesses. Moreover, there are exception to this trend: the entire in-app purchase gaming segment, for example, or Evernote, which is finding great success as a predominantly consumer-oriented SaaS offering. However, going forward, I expect companies built with such models to be increasingly specialized and smaller-sized, and those which are already large – such as Dropbox – to continue pivoting to the enterprise.