Do tech companies have moats?

Barriers to entry framework

• Economies of scale
• Where cost per unit goes down with scale - typically applies for businesses with large fixed cost bases
• Network effects
• Where the value of the product increases with scale and users e.g. online communities like Discord
• Switching costs
• Built-in structures designed to incentivize to stay with the product e.g. frequent flier miles programs
• Capital requirements
• Requirement for large investments in order to launch a competing product e.g. AWS
• Government policy
• E.g. patents, government sanctioned monopolies
• Brand

High-level implications for tech startups

1. Compared to companies selling physical products, tech startups typically do not have strong barriers to entry.
• Most software companies don’t benefit from economies of scale, government policy or have high capital requirements
• The main fixed cost for software companies is their R&D (i.e. technology development spend), which can be overcome more easily than for certain physical products
2. Despite this, tech companies can still have a moat. The moat is based on producing products that are 10x better, faster, cheaper than alternatives. And the moat exists because software is an incredible form factor through which products can be created, iterated and modified - much more easily than with physical products. The only real limitation is the inventiveness of the creator. However this moat can lack durability, because it can be overcome by someone else dreaming up something better.
• The ability to create 10x products is dictated by the quality of the team. This is why many VCs believe that Team is the most important asset that software companies have - and that execution is the only true moat.
3. Given this lack of moat durability, software companies should supplement their product creativity with more classic ‘offline’ moats
• It’s not enough to produce high-quality software that uniquely solves a customer need. In fact, the opposite is true: startups often lack durable advantages purely on the basis of software and should incorporate analog world strategies to fortify their business, such as:
• developing proprietary distribution channels
• maintaining a direct customer relationship
• developing skills in sales and marketing (e.g. Brex case study here)
• partnerships
• bundling (e.g. Microsoft selling Teams to existing customers at a discount)

Moats in B2B SaaS

• Most B2B software can be fairly easily replicated
• Although it may be unfashionable to say, in my experience most B2B SaaS products are replicable by small teams of talented engineers (with low $ millions budget and <12 months)
• As a recent example, despite originally focusing on employee and payroll software, Rippling was able to quickly launch a spend management product line that competes with Brex and Ramp
• Said another way, technical risk is not the key risk for most B2B software
• Nevertheless, B2B software can be designed to incorporate switching costs into the product (i.e. create stickiness)
• As an example, workflow tools with deep integrations can become hard to displace, even by more modern alternatives e.g. Workday, Atlassian
• Becoming a system of record and/or a platform can further entrench a product e.g. Salesforce
• Software enhanced by embedded fintech product lines can increase its value and differentiation in the eyes of customers e.g. Square’s cash flow advance product for merchants
• In some cases, products can be specifically designed to become more valuable to users over time e.g. as a user of Notion, you experience accruing benefits and mounting losses over time

Marketplace moats

• Marketplaces benefit from network effects, which provide a genuine barrier to entry. But these moats are only protective once the company reaches scale - and are rarer than many investors often believe
• In my experience meeting and investing in startups over the past few years, businesses with powerful network effects are rare to encounter. One reason for this may be that a lot of the more obvious ideas have already been picked over. This isn’t to say they cannot be built, just that they are fewer and further between
• Additionally, these network effects tend to only exist once the company gets to a certain scale. In the interim, the company is vulnerable to competitors who out-execute them and are better funded
• The strength of network effects varies substantially based on the type of marketplace
• 2-sided marketplaces tend to have the strongest network effects (e.g. Ebay), but are rare to uncover in 2023
• Platform network effects are powerful at scale, but there are few companies that end up becoming platforms
• Data network effects are often weaker than investors want to believe
• Hyperlocal network effects (e.g. Lyft) need to be built up in each new geography that is launched
• Labor marketplaces (e.g. Uber Freight) often have a diminishing network effect, whereby a competitor can somewhat easily launch a ‘good enough’ alternative

Concluding thoughts