I’ve read Adrian Slywotzky’s The Art of Profitability three times now, and each time I learn something new. According to its Amazon page, the book “offers 23 business lessons via the tale of a manager’s quest to learn the “art of profitability” from David Zhao, a wise master.” Each lesson starts with basic principles and uses the Socratic method to build up to a profitable business model. I think current founders, aspiring founders, investors, and employees would find this book valuable and practical. Here are some notes on the profit models that are most relevant to software companies – although the notes are no substitute for reading the book!
The main theme: “The path to profitability lies in understanding your customer.”
Assorted Profit Models
Key insight: Different segments of customers want different levels of quality/service and have different abilities to pay. Product lines that map well to different customer segments can capture more profit.
Explanation: A pyramid is a system of three profit tiers:
a defensively priced bare-bones product to keep competition out.
a standard product for most people.
deluxe version to get maximum profit from those with a high willingness to pay.
Many SaaS businesses follow this profit model: there’s a free/cheap tier, a medium priced tier that is appropriate to most customers, and an enterprise tier for customers with deep pockets who want the most features and services. Another example of this profit model would be different car models for a single brand (e.g. Nissan Versa vs Altima/Maxima vs GT-R).
Key insight: Customers can have different price sensitivities for the same item in different contexts. Someone who needs a resume proofread might pay $25, but someone who needs it proofread in the next 24-hours might pay $100.
Explanation: You can often sell the same thing in different contexts and for different prices. For example, you can sell a bar of soap for $1 at 7-11, a 6-pack $3 at a grocery store, and a 30-pack for $10 at Costco. Similarly, you can rent out a hotel room to individuals, wedding parties, or business conference attendees – the prices and profitability levels will vary greatly across these use cases, but the room is always the same.
Key insight: In contexts where assembling a package of related goods and services takes a lot of effort, customers will pay a premium for pre-assembled packages.
Explanation: There are situations where a customer prefers to buy a package of several interlocking pieces instead of hunting down each piece separately. The book uses the example of movie studios, which need a director, a script, and a star to make a movie. Studios are willing to pay a large premium to an agent who can deliver a bundle of all three things. Heroku is another example of this model: they offer bundles of software and hardware services, and people pay a premium for not having to deal with installation, integration, and deployment. The more services Heroku supports out of the box, the more it can potentially charge because the cost of configuring, deploying, and monitoring every service individually is too high.
Key insight: In some markets, the path to profit is to produce blockbusters. The movie industry is one such market; the pharmaceutical industry is another one. Because profit in these markets is very hits-driven, companies need to be very comfortable with risk, and they need to understand the components of a blockbuster as much as possible.
Explanation: In a blockbuster-driven industry (e.g. consumer software, mobile games, etc.), R&D can be a huge money loser if you are doing research in the wrong areas or in an area not worth researching. Focus on understanding blockbusters as much as possible, so that you can focus your R&D spend on the most promising, most profitable areas. It’s a shame when someone invests a lot of time and money into developing a product that people don’t want.
Key insight: You can improve the effectiveness of R&D by increasing the amount of profit that successful projects produce. If you can turn one successful product line into five then your ROI will be much higher. This is especially true in software businesses where different product lines can share a lot of code and infrastructure.
Explanation: Many companies can take a single skill or asset and repackage it in many ways. For example, Honda is good at making motors, and they use that skill to sell cars, boats, motorcycles, lawnmowers, etc. Similarly Disney’s IP is used for movies, TV, merchandising, music, etc. Unlike the Multi-Component model, which involves selling the same product in different contexts, the Profit-Multiplier model is about creating new products that are repackaged versions of a core asset.
Key insight: If you understand a problem better than anyone else, you’ll be able to create better products, and customers will pay a premium to work with you.
Explanation: Become a domain expert in a new discipline, then use your expertise to generate profits. Being an expert gives you the knowledge to lowers costs and the reputation to justify higher prices. This model focuses on learning everything about a specific problem. Palantir is a good example of the Specialist approach: by setting themselves up as the foremost experts on data management and analysis, they command higher prices and, as one of my friends said, “they [de facto] have first dibs on any interesting big data problem.”
Installed Base Profit
Key insight: Customers who already use your products are a great market for upgrades, add-ons, related products, and so on.
Explanation: One particularly effective business model is to sell products at a low profit margin, then sell add-ons, consumables, upgrades, support plans, and so on at a higher profit margins. Examples of this model include printers (low margin, but ink is high margin), Amazon Kindle (no margin, but ebooks have good margins), and cars (lower margin, but dealer maintenance services are high margin).
Specialty Product Profit
Key insight: Specialty products usually earn much higher margins that commodity products (although not for long).
Explanation: Unique products that serve a small niche can make a ton of money, especially in the absence of competition. However, as the specialty products become more common and more commoditized, their profit margins drop dramatically – an event that companies need to plan for. A lot of recent SaaS services, which are becoming more and more specialized, are good examples of this model. There are now CRM programs designed for dentists, cloud-based auditing tools for CPAs, and the like.
Local Leadership Profit
Key insight: Having a near-monopoly is one geographic area can be more profitable than owning a small piece of the market across many locales.
Explanation: This is the playbook for a lot of on-demand startups: Homejoy, Uber, Move Loot, Spoon Rocket, etc. If you’re competing with a lot of companies for the same location, profits plummet (or go negative). On the other hand, if you have a monopoly on a location, you can enjoy high profit margins while your competitors struggle to get even a small foothold.
Relative Market Share Profit
Key insight: The higher your market share, the more advantages you have in terms of cost structure, distribution, marketing cost per unit, R&D cost per unit, and so on.
Explanation: The biggest player in the market can spread their fixed costs across many more units, which provides the flexibility to decrease prices or increase advertising spend or take other actions that make it even harder for others to compete. Companies that understand the importance of relative market share do whatever it takes to become #1. Think Uber vs. Lyft as a great recent example.
Another advantage of higher market share is that you can accumulate more data, which increases competitive barriers. For companies like Yelp (business review data), LinkedIn (professional network data), and Google (search query data), the ability to learn from more data makes it next to impossible for someone to compete, even if they have a better fundamental product. For example, someone who makes a better version of LinkedIn will still struggle to get anywhere because LinkedIn has so much more data about what kind of job recommendations people respond to, how to find colleagues that someone might know, how people in different companies are connected, and so on.
[Data network effects are a key reason for my fund, Susa Ventures, having a strong focus on companies accumulating valuable datasets.]
I think that the best way to use The Art of Profitability is to try applying different models to your own business and see if any of them could fit. Is there a core asset you can repackage into different products? Do customers have different price sensitivities for your product in different markets? Can you sell your product as part of a pre-assembled bundle to save your customers from integration headaches? And so on.