What Makes a Business Defensible?
Your moat won’t build itself…
When a business is profitable, others businesses will copy it. Over time all of the companies that do the same thing are forced to drop the price of their products to compete with each other until, eventually, there is no profit to be made by anyone. This is the Tragedy of the Commons: when every player takes an action to improve their lot, everyone ends up losing together. The classic solution of Collaboration is illegal in this context so that’s not an option either. The real trick to business is not just to provide value, but to find the value that only you can provide. Or as Peter Thiel says:
In the real world outside economic theory, every business is successful exactly to the extent that it does something others cannot.
This is one of the reasons launching a business is so difficult: it has to be difficult. If it was easy everyone would do it, until there was no money to be made anymore. A simple mental model is this: the profits your company makes are a function of the difficulty it overcomes.
The corollary of this is that businesses with small/no barriers to entry are never profitable. Sustaining a simple digital marketing company is nigh on impossible when every kid with an Instagram account can copy what you do. Building a profitable online store is vastly more difficult since cloud computing, online commerce platforms, and payment APIs removed the need for capital and technical talent. As starting a business becomes easier, sustaining it becomes harder. For most people, the difficulty of achieving great success in business is inversely proportional to how difficult it initially appears to be.
This is also why your idea for a startup is not valuable. Anyone can have an idea. Worse, anyone can copy your idea. It is in the execution of difficult tasks that wealth and value is created.
So how can a business owner create value that cannot be replicated? One simplistic approach is to spend a lot of time working on things that are difficult, just because they are difficult. That alone won’t be enough, but it is a start. You could be doing difficult things that provide no value whatsoever, so correct planning is also important. But the kind of person who seeks out difficult tasks for the sake of conquering them is more likely to become very successful than the average Joe.
So what else needs to be considered to create value in an irreplicable manner? We actually know many of the levers that can be used to create competitive advantage:
- Economies of scale
- Network economies
- Regulatory protection
- Switching costs
- Process power
- Cornered resources
1. Economies of Scale
The bane of a startup’s existence: economies of scale are the primary force keeping incumbent businesses in control of their respective markets. They make it cheaper for organisations that already dominate to compete. Would-be challengers who compete with an incumbent that has a strong scale economy will find that offering a similar product at the same price point hurts their P&L, and raising the price or reducing service to cut costs results in loss of market share. Products that require significant upfront investment to develop but low marginal cost to distribute can avail of particularly strong scale economies. Software companies, telecoms, media, infrastructure, and anything that leans heavily on R&D and intellectual property are prime candidates.
An incumbent with the benefits of scale will use it liberally to out-compete their smaller rivals. There is no solution to this problem. Challengers must offset this disadvantage themselves using one or more of the factors listed below to compensate.
2. Network Economies
Network economies, or network effects, are very well understood in the technology sector and not so well understood elsewhere. A product with a network economy is one whose value to each consumer increases with the number of people who use it. Think telecoms networks, social networks and office software. In these cases the cost of the product can be reduced as adoption increases, increasing the value of the product to each consumer.
Most products are unchanged by how many users they have. Your breakfast cereal does not taste better because more people buy it. Some products even have anti-network effects. Precious metals and gemstones derive their value from scarcity and would be significantly less valuable if the supply was not artificially restricted.
Network economies are common with software businesses and are the basis of digital marketplaces: they tend to be particularly difficult to get started but when a marketplace reaches a critical mass of users it becomes incredibly hard to compete with, and incredibly profitable. They form natural monopolies due to the exponentially increasing value they provide as they scale. This effect can be so strong that managers of successful network-driven businesses often spend more time thinking about how to slow growth so their operations can keep pace than the inverse.
3. Regulatory Protection
One of the greatest public sector innovations in history was developing a system to incentivise private corporations and individuals to invest in public works, vastly reducing the onus on governments around the world. Traditionally, governments award some form of monopolistic privilege to private enterprises in return for full or partial funding of an infrastructure project. This privilege allowed investors to predict the return on their investment much more accurately, and also to extract more value from the investment in the first place. Nobody wants to invest in a bridge if someone else can build another bridge a mile away. You want to invest in the only bridge across the river, and to charge appropriately. Therefore, governments often offer to legally block competitors from building another bridge if you agree to fund the project.
If you want to start a new bank in just about any country in the world, one of the first requirements is to get a local banking license. This is an extraordinarily difficult, expensive and time consuming process that can extend over several years working closely with local regulators. That is a huge advantage for the firms that already hold banking licenses. The barrier creates an advantage.
Governments are very conscious of the perverse outcomes regulatory protection can generate: it is important to ensure critical systems like the financial markets maintain their integrity and trust of the public, but preventing competition often leads to the system deteriorating over time anyway. As a result some regulators, such as the Financial Conduct Authority and the Monetary Authority of Singapore, have begun to explicitly support innovation with the introduction of sandbox programmes in strategically important areas.
4. Switching Costs
Switching costs occur either when use of a product makes it more difficult to stop using it, or when the product is especially difficult to replace. It can often be identified by customer surveys showing high retention rates coupled with low satisfaction. It lends itself to subscription based products that offer a free trial and involve substantial effort to set up. Good examples include: ERP systems, operating systems, your bank, Facebook, Evernote, Bloomberg terminals, Powerpoint, Excel, and many, many more.
There are 3 kinds of switching costs to consider. Financial switching costs describe situations where it is expensive to move to a new product or service provider. Procedural switching costs describe situations where users cannot switch because they need to invest significant time in setting up an alternative (SAP) or because they are familiar with their current solution and can’t invest the time to learn to use alternatives (Excel). Relational switching costs describe situations where the customer is reluctant to move to an alternative because it would affect their self-image or personal relationships (WeWork).
Branding is defined as:
The durable attribution of higher value to an objectively identical offering that arises from historical information about the seller.
Of all of the factors on the list, Brand is probably the one you are most familiar with. Branding is the reason iPhones sell at over $1,000 apiece even though they are taking in almost 50% in margin while other manufacturers struggle to break even. Branding is also the basis for the high fashion industry. Branding is why you pay twice the price for a bottle of Coke than a bottle of Cola. One reason branding can be a source of irreplicable value is because it invokes the seller’s reputation, trust, and other implicit associations that affect how the product is perceived. Another reason is that branding reduces uncertainty around the quality of the product.
An important caveat: brand awareness does not necessarily result in good, capital-B Branding. General brand awareness can be a side effect of economies of scale. Hamilton Helmer, a Stanford instructor and CIO of Strategy Capital, describes it like this:
Coca-Cola can sponsor Super bowl ads while RC Crown cola cannot because the ad cost is only justifiable for an entity of Coca-Cola’s size. A strategist would gravely err in classifying this as Branding. RC could make all of the right Branding moves and still be at the same disadvantage due to relative scale.
6. Process Power
Process power is usually best illustrated with an example.
One of the inspirations for the founder of Toyota, Eiji Toyoda, in scaling up the company was the manufacturing revolution pioneered by the Ford Motor Company. In 1950 he visited Ford’s flagship “integrated factory” and was shocked at the wastefulness of the processes he observed. Ford had been keeping huge stocks of inventory and running below capacity so they would have the ability to smooth out variations in the factory’s output, a feat they were very proud of. In response to this Toyoda returned to Japan and laid out a new set the operating principles in his factories, principles which ultimately led to the famous Toyota Production System.
To the surprise of everyone in the business the fabled TPS was not considered a trade secret to be protected, and Toyota regularly held tours to demonstrate the efficiency of their production lines. Hundreds of thousands of executives had toured the factory and learned how the system works. They learned the principles on which it operated: focus was given to continuous improvement, inventory control, and defect detection. And yet, no matter how smart they were nor how much money they had behind them, they simply could not replicate the system.
The reason the TPS could not be replicated is because it evolved over time, adding complexity with each tweak and change made to improve performance. The complexity of the system was overwhelming and the details mattered enormously. The TPS system was understood but could not be replicated in the same way we understand how the human body works and what it is made of, but we still cannot create a living person by sitting down and arranging all of the molecules.
Process power is closely linked to advantage bestowed by proprietary technology. A deep learning system can learn to play Go at a world-class level by processing the outcomes of millions of games played. But no programmer could sit down and develop a Go playing bot that could beat the world champions. The degree of nuance and detail required could not be designed in a lifetime.
Google makes money not because they provide a search engine but because they provide the best search engine, and no amount of money invested in Bing will make it better than Google Search overnight.
Process power is one of the most valuable form of competitive advantage, available to companies of all sizes, and rarely discussed explicitly. It can usually be seen hidden in the stories of overnight successes that were a decade in the making.
7. Cornered Resources
Every business can be reduced to a function with inputs and outputs. If there is a scarce and limited supply of the inputs then there is an opportunity to create an advantage by controlling all of the available supply. This input could be anything from capital (in a capital-constrained environment, not today), talent, expertise, or raw materials. The biggest tech companies pay and treat developers well not because they have lots of money (the shareholders would like that money) but because technical talent is the limiting constraint in that industry, not capital. So retaining that talent is more important than retaining that money. The caveat here of course is that the impact of the marginal engineer (or whatever the resource happens to be) must create more value for the company than it costs to retain.
The last source of competitive advantage is the most important for startups, innovation teams, and what Google would call “Other Bets”.
Do things your competitors cannot, or will not, do.
Take John Bogle and Vanguard as an example. Vanguard pioneered the concept of index funds and became one of the largest asset managers in the world. Any established asset manager in the world could have done what they did. However, to do so would have severely cannibalised the fees on their higher margin products. They just stood back and watched Vanguard take over the market. The decision to not offer index funds wasn’t an oversight. It was a carefully deliberated strategic decision, albeit one made on a short-term time horizon. At any given time an incumbent that enjoyed high margin active management strategies would have been worse off by adopting the new product, and yet many went bankrupt in the end because of a failure to act.
This illustrates the principle: a startup has a huge advantage over incumbents when their business model cannot be replicated because of the incumbents’ existing business models.
This essay leaned heavily on the following books: