Horizon scanning 2: internet wars

A couple of years ago I wrote a post about horizon scanning: taking a big picture perspective on how the economy as whole is changing and what this implies for future growth opportunities. The main theme was how the internet is transforming the economy – while some sectors of the economy are being disrupted, others are growing rapidly. This transformation is the most significant change happening to the economy by far and there’s still a long way to go. From an investing perspective, I think it’s critical to understand as much as possible about who the winners and losers are going to be. So I’m following up on my previous post with a more detailed look at the dynamics of competition between internet-related businesses…

A central idea from neoclassical economics is that market forces will naturally promote competition. Markets where businesses are able to earn high profits will attract rivals who want a share of the pie. In highly cyclical industries (eg airlines or any sort of commodity), this dynamic is very clear. It is present to a lesser extent in many others. It should cause businesses that happen to be experiencing particularly high returns to suffer from ‘reversion to the mean,’ as more focused or innovative competitors are attracted by the opportunity to eat their lunch. The idea that investors will often fail to anticipate this and over-extrapolate future returns based on historic success is one of the possible foundations for value investing.

However, it’s long been known that there are exceptions. While many markets experience ‘reversion to the mean’, some experience the opposite dynamic, where scale becomes an ever-increasing advantage. For example, a successful brand may become better known over time, develop its distribution network, operationally leverage its cost base and improve its expertise and capability for innovation as it scales up. Businesses that have these sort of scale advantages are often then able to allocate capital more profitably than smaller rivals and successfully deliver compound growth for much longer than expected. They have tended to make very good long term investments, as short-sighted investors have under appreciated the extent and longevity of the benefits that accrue over time.

The internet has amplified this second dynamic in a fairly dramatic way, by enabling new classes of platform business model that benefit from more extreme scale advantages. In the traditional world, economies of scale tend to be more subtle. There are often many dis-economies as well, as bigger corporations become less focused and harder to run. Smaller rivals could be successful in most markets as long as they had a good idea and executed well. However, for online platforms, scale advantages can be so great that it is difficult to compete effectively against large incumbents, no matter how good your idea is or how well it is executed. In many cases, scale advantages are so great that the market tips towards a single winner.

Platform business models

The internet supports platform business models based on connecting consumers and businesses to information or each other. Most of these fall into one or more of the following categories:

  • B2C Content delivery: platforms that provide digital content, such as films or computer games direct to consumers. They can have a variety of pricing models, including by-transaction, subscription or funded by advertising.
  • Ad-funded social networks: platforms that connect users to one another and content produced by other users. These tend to be free to users (so they can be adopted widely and rapidly to generate network effects) and then funded by advertising.
  • Transaction platforms: a variety of platforms focused on enabling various types of one-to-one transaction between two groups of users, typically businesses and consumers. These include payment networks, online marketplaces, comparison websites and search engines.
  • B2B enterprise software (SaaS) platforms: these provide various types of software to support businesses with back office tools, productivity and customer relationship management etc.

Sources of scale advantage

Various aspects of the competitive dynamics affecting all of these business models result in scale advantages.

Cost structure

Typically, large fixed investments are required to develop and maintain an online platform, while the costs of serving additional customers tends towards zero as the platform scales up. This allows platforms to benefit financially from operational gearing, making them extremely profitable once they have sufficient scale. More importantly, making substantial fixed cost investments in improving platform quality creates a first-mover advantage and a strategic barrier to entry. Rivals are deterred from making similar investments as they would need to compete with the incumbent, reducing the profits they would make even if their investments were successful.

This sort of cost structure isn’t unique to online platforms but it does tend to be more pronounced. Easier access to capital in recent years has allowed technology businesses to make huge strategic investments and scale up rapidly, rather than sacrificing their scale advantage by trying to become profitable too quickly.

Network effects

Many of the scale advantages online platforms benefit from could be described as one sort of network effect or another. Network effects occur when the value of a service increases with the number of its users. When you have very strong and direct network effects, as in communications or social networks, markets can tip to a single provider very rapidly as everyone needs to be on the same network as the people they want to connect to.

Two-sided transaction platforms are also characterised by powerful ‘cross-platform’ network effects. On an online marketplace, buyers want to go where they can access lots of sellers and vice versa. This means a transaction platform faces a ‘chicken and egg’ problem in needing to build up both side of the platform before it can become viable. This can make it very hard for rivals to attract business away from an incumbent that has already scaled up on both sides of the platform.

Transaction platforms typically need to compete more vigorously for consumers, who are more readily able to switch to alternatives, than for businesses, who need to go wherever there are consumers to buy their products. In practice, this tends to mean that transaction platforms offer their services to consumers for free (or even provide them with incentives) while cross-subsidising this with revenues from business users. Attracting businesses with lower fees is of limited benefit if there are no consumers for these businesses to transact with. Most of the time winning over consumers is key and everything else follows.

Use of data

Online platforms increasingly rely on algorithms to present users with the information, content and functionality they want, often tailoring their services to individual users. There are indirect network effects inherent in this process, as greater usage of the service generates more data that can be used to further improve the algorithms. This means that having more users indirectly contributes to a higher quality service, making it easier for the platform to gain even more users. The importance of this ‘flywheel’ can vary – for instance it is absolutely critical for a search engine, where the core functionality is an algorithm that predicts what website a user is looking for based on a search query. For other platforms it can sometimes be a more minor aspect of the overall service.

Convenience, ecosystems and bundling

The core reason the internet has been so transformative is that it offers convenience. It lowers the costs of searching for information and services and can reduce transaction costs to near zero. Above all else, convenience is probably the main reason that we consumers make the choices we do. After all, we all have a lot on our plate and want to prioritise our attention for the things we care about. The convenience brought about by the internet has affected competition in all markets, not just between online platforms, by making it easier for the biggest and best brands to prevail and making it a bit harder for businesses to exploit narrow niches eg from local distribution networks and customer relationships.

Convenience is also one of the main drivers of competition between online platforms. The convenience of being able to connect with multiple other users in a single place is part of the reason for network effects. This also extends to the idea of being able to access multiple related services in one place. Once an online platform has the attention of consumers using one core service, it then has an advantage in providing other complementary services. Conversely it may face competition from other platforms doing things the other way round.  This means there are strong incentives for platforms to aggregate multiple related services into a single ‘ecosystem’. One of the reasons the tech giants can keep on growing is that they often have plentiful opportunities to exploit their advantages in expanding into related services to build out their ecosystems.

To reinforce the convenience that comes from accessing multiple products or services in one place, platforms can use subscription pricing models that effectively bundle these services together. This reduces the incentives for users to ‘multi-home’ across multiple platforms and preserves direct and indirect network effects.    

Sometimes the relationships between different services have a specific direction, in that one service is an access point or gateway for another. The ‘customer journey’ to buy a product online might involve ‘traveling’ via a smartphone, a mobile operating system, a browser, a search engine and a comparison platform before finally reaching a supplier, each step having some influence over access to the next. These access points can have huge strategic significance for the platforms that operate them. This is illustrated by the fact that Google, despite having a near-monopoly search engine, pays a surprisingly large share of its search advertising profits to Apple in ‘traffic acquisition costs’ simply to be installed as the default search engine on Apple devices.


As well as offering convenient access points to each other, many online services are more valuable when they can interoperate with one another. This allows different technologies to be used together rather than independently. For example, when I publish this blog it will automatically be posted on social media without me needing to do use both services independently. Often there are important and complex interdependencies between different services, for example where the data output of one application is an input for another. Interoperability between services is especially relevant for businesses, where the activities and information generated is complex, highly interrelated and often needs to be monitored and analyzed. For example, many businesses may want various HR functions to be integrated into a single system, or to integrate their customer databases with the platform they use to buy advertising. 

Realising all the potential benefits from making services interoperate well is much easier when they are provided by a single entity, as this reduces the need for agreeing various standards and protocols and allows further innovation to be delivered more flexibly.  This creates a scale advantage, or strictly speaking a ‘scope advantage’, for a provider that has aggregated multiple complementary online services into a single ‘ecosystem’. 

Of course, interoperability between services provided by different entities is also possible. In some cases it’s in everyone’s interest for services to interoperate as well as possible, as the value of all services to users is improved.  However, often larger providers with other scale advantages may not want to allow too much interoperability with the core functionality of their services as this can undermine their value. If different networks are able to seamlessly interoperate with one another then it matters less which network users are on and network effects are diminished. Scale is also often an advantage when negotiating the terms of interoperability. Providers of larger, more important services have a much stronger position in dictating the terms of interoperability. This can be used strategically. For instance, a small platform may initially want to allow for as much interoperability as possible to incentivise complementary technologies to be developed for its platform, increasing the overall value to users. At a later point, once it is more powerful and is in a position to start providing these complementary technologies itself, it may then decide to reduce interoperability. 

The dynamics of interoperability are complex and hard to generalise about but nonetheless very important. Without getting into specific cases, for investors I think the high level points to bear in mind are that these dynamics tend to reinforce existing scale and scope advantages and that they create additional opportunities for larger businesses to exploit.

Implications for investors

As I explained in this post, the distribution of long term returns from investing in different stocks has always been extremely distributed, with a handful of long term winners being responsible for the lion’s share. It doesn’t take a wild imagination to see where things are going – the transition to a more online economy and the huge scale advantages inherent in these markets should exaggerate this further. The largest and most successful global businesses are already a small number of online platforms that clearly benefit from many of the advantages described above. They have grown incredibly rapidly, continue to get bigger and I predict will continue inexorably for a long time. Of course, it’s not just the acronym of tech giants that will be successful. There is room for more larger players to emerge and there are plenty of smaller players that look set to dominate specific niches.

This will be reinforced by a trend towards convergence, as convenience and interoperability lead to ever-increasing advantages in providing various complementary online services together. While the whole ICT sector is benefiting at the moment from strong secular growth, a tide that is currently lifting all boats, as time goes on these businesses are going to be forced to butt heads more and more. The outcomes of this are hard to predict and will take a long time to play out. Having larger scale and scope is likely to be a significant advantage in a world where being able to aggregate multiple services is important, but it’s not the be-all and end-all. There are many examples of the tech giants unsuccessfully trying to muscle in on a new market in competition with a more specialised incumbent. Most of the time these attempts fail but every so often they succeed. Another advantage of having a large and important platform is that there are many opportunities to experiment.

This means that identifying the high quality winners matters more than ever. I don’t think investors collectively are very good at keeping up with this sort of structural change, preferring to steer by looking in the rear view mirror at what has tended to work in the past. Many seem inclined to see the recent outperformance of quality and technology businesses as more of a temporary sentiment-driven phenomenon rather than a reflection of structural changes to the economy. This scepticism creates favourable conditions for these businesses to continue to outperform.

The factors that predict quality, ie the extent and consistency of long term future growth in profits, are also likely to be changing. I think qualitative factors to do with competition and strategic opportunity are likely to become increasingly more useful predictors than the traditional backwards-looking profitability metrics many investors are more accustomed to. These traditional metrics rely on the insight that consistently profitable and growing businesses are likely to have some sort of secret sauce that ensures success in the future. This insight is still useful. However, as markets have become more dynamic, I think it’s becoming a less central part of the picture. Companies can expand more quickly in online markets and these days have plentiful access to cheap capital. Given the advantages that come from scale and the favourable economics in operating a capital-light IT business, achieving scale is often more of a priority than proving that high profitability can be achieved. This should mean that it is increasingly possible (and necessary) to identify winners some time before there is any track record of profitability. Businesses like Amazon and Netflix epitomise this (as do the other tech giants to some extent). They may seem like exceptions but I believe the reality is that these dynamics are becoming more prevalent everywhere. 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s