Patterns are everywhere. We use patterns in architecture, design, software, chess, and cooking.
Patterns help us codify knowledge into repeatable packages of wisdom that we can reapply in other contexts.
Business models are no different.
In today’s issue, I’m going to share three archetypical business model patterns you can use to
- model any product,
- quickly identify top risks and
- formulate a starting validation strategy.
3 Business Model Archetypes
When people bring up business models, they often use a bunch of terms such as software as a service (SaaS), enterprise, retail, e-commerce, ad-based, freemium, viral, social, not-for-profit, marketplace, et cetera.
We end up with dozens of business model descriptors because we attempt to label how a business model creates, delivers, and captures value. For instance, the difference between SaaS, enterprise, and open-source business models is how they deliver and capture value. One could implement a freemium or trial-based pricing model even within a SaaS business model. Trying to create a list of business model types gets complex pretty fast.
Instead, I will take a different approach — categorizing business model types by the number of actors (or customer segments) in the model. When you view business models this way, three business model archetypes emerge:
- Direct,
- Multisided, and
- Marketplaces.
1. Direct Business Models
Direct business models are the most basic and widespread type of business model. They are one-actor models where your users become your customers.
A coffee shop is a simple example.
The coffee shop attracts visitors to its storefront by its ambiance and promise of great drinks. When a visitor, now a user of the coffee shop, purchases a drink, they become a customer, and some of this value is captured back as money.
As long as the coffee shop creates more value (even perceived) for its customers than it captures back, it creates a happy customer and has a compelling value proposition. And as long as the coffee shop can capture back more value than it costs to deliver this value, it has a sustainable business model.
In a direct business model, monetizable value is extracted directly from your users, who become your paying customers, which is simply the net revenue realized over the life of the customer.
Traction in a direct business model is the rate at which you turn nonpaying users into paying customers.
Other examples of one-actor direct business models are:
- Retail
- Software as a service (SaaS)
- Mobile apps
- Physical goods
- Hardware
- Services
2. Multisided Models
Unlike a direct business model where your users become your customers, a multisided business is a multi-actor model where your users and customers are different actors (or segments).
In a multisided model, the goal is still to create, deliver, and capture value from users, but that value is monetized through different customers. Users typically don’t pay for usage of your product with a monetary currency but with a derivative currency. When compounded across enough users, this derivative currency represents a derivative asset that your customers pay to acquire.
Let’s look at some examples that will make this more concrete:
Ad-Based Business Models
Products like Facebook, Google, Twitter, and YouTube fall under this group of business models. We’ll use Facebook as an example. Facebook creates and delivers value to its users through its social network—but doesn’t charge its users directly. It still captures some of this value back, albeit through a derivative currency (user attention, in this case).
Facebook then trades this derivative currency on a secondary market of advertisers (its customers), who pay to reach these users.
We can describe the same business model with Google’s search engine business, substituting its search engine for Facebook’s social network. In both these examples, the derivative currency is attention, which is monetized by converting attention (from users) into impressions and/or clicks for advertisers (their customers). This conversion of the key monetizable user activity into actual revenue is the derivative currency exchange rate. For ad-based businesses, this is typically described as CPM (cost per thousand impressions), CPC (cost per click), or CPA (cost per acquisition).
Monetizable value is a function of the derivative currency exchange rate, which we can use to calculate the effective monetizable value of users (or an average revenue per user—ARPU) even though they aren’t directly paying us. As of Q1 2015, Facebook’s annualized advertising ARPU was $9.36.
Traction in a multisided business model is the rate at which you capture monetizable value from your users as a derivative asset.
Big Data Business Models
Attention isn’t the only kind of derivative currency. Another example is data. You might give away a free mobile fitness app to your users and aggregate their usage data into something more valuable that an insurance company, for instance, may want to purchase.
Now for a few not-so-obvious multisided models.
Enterprise
The traditional enterprise product can also be described using the multisided model. Organizations (our customers) are made up of people who play different roles in the business model. The business model usually has at least two (and sometimes more) roles.
Users here are the employees who use the product to help the organization realize the value proposition of the product. The customers here are the decision-makers who purchase the product for the employees. Some other key roles worth modeling might be the influencers in the organization—for example, the IT department—that have a say in the buying decision.
The primary value flow, however, remains the same. Users of the product create a derivative asset, which, in this case, can be measured as a productivity gain or an improved business process that helps the organization capture more value from its customers. As long as this asset creates more value for the organization than what the decision-makers paid to acquire it, it represents a net positive ROI and a compelling value proposition.
Not-for-Profits
Not-for-profits can also be modeled as multisided models. Let’s take the Red Cross as an example. The users of the Red Cross are the people in need that the organization serves. And donors are the customers. Because these models are usually impact-driven, the number of people helped represents the derivative asset that donors fund. If the Red Cross stopped serving these people, the donations would dry up accordingly.
3. Marketplaces
Marketplace models are a more complex variant of the multisided model that warrant their category. Like multisided models, marketplaces are multi-actor models of two different segments: buyers and sellers.
eBay, AngelList, and Airbnb are all examples of marketplace business models. The transaction is the key activity that creates happy customers.
Monetizable value in these models is typically captured as a percentage of the transaction value created between buyer and seller as a commission, listing fee, et cetera.
Traction in a marketplace model is NOT the rate at which you create buyers or sellers (listings), but the rate at which you bring both sides together to conduct a transaction.
Finding more complex models in practice that layer these basic archetypes is possible. The thing to keep in mind is that even these more complex models started with a basic model at the outset.
Gall's Law states: A complex system that works is invariably found to have evolved from a simple system that worked.
Strive for simplicity, not complexity. Simple is hard
Turning these Business Model Archetypes into a Validation Strategy
When evaluating any idea, I first model it into one of these archetypes — one archetype per canvas, which informs strategy. From there, I move to recipes or tactics.
It’s important to highlight that all models are flawed because they simplify reality. But good models are useful precisely for the same reason.
So, use the following sections to guide your starting validation strategy and evidence-based learning to refine from there.
Finally, to stay within reasonable limits for a newsletter article, I will outline these strategies and recipes at a high level. I cover them in more detail in the Continuous Innovation Playbooks — the first of which is included with a paid newsletter subscription.
Validating Direct Models
Being one-actor models, these are the simplest of the three models, but simple doesn’t mean easy. One could counter-argue that the bar is a lot higher since all users are potential customers.
The biggest risk in these models tends to be pricing the product right, and the biggest mistake is deferring price/value testing.
The good news is that these models typically have many customers, making customer/problem discovery easier.
Business Modeling Patterns
- Identify a single customer segment on your Lean Canvas. Keep it simple.
- Identify the smallest subset of distinguishing characteristics for your early adopter segment.
Validation Patterns
- Use a Mafia Offer campaign (high-touch problem discovery, solution design, offer testing) to uncover your early adopter’s unmet wants and design/refine/position your product to cause a switch.
- Don’t defer price testing.
- Scale your Mafia Offer campaign to a more suitable and scalable campaign such as a landing page, webinars, or crowdfunding.
Validation Anti-patterns
- Starting with freemium
Validating Multisided Models
The common theme across multisided models is that there are user and customer sides. The user side is often the riskier of the two sides because that’s where monetizable value is created in the form of a derivative asset.
There are two challenges with derivative assets. First, this asset needs to be aggregated over a tipping point of users to make it valuable for customers. For instance, a ten-user social network is not all that interesting to advertisers. The second challenge is that the derivative currency exchange rate (how much an advertiser would pay in this example), like any derivative asset, is not a given and fluctuates over time. For these reasons, an effective validation strategy is first to tackle the user side of the model until a sufficient tipping point is achieved.
The key in multisided models is establishing the derivative currency exchange rate early. This helps demonstrate the business model story, which drives the valuation of the business. The more liquid this conversion, the higher the valuation. This is precisely why Facebook commands a higher valuation per active users than Twitter, which commands a higher valuation than Snapchat.
Business Modeling Patterns
- Identify both sides of the model on a Lean Canvas.
- Use labels like user/customer, user/decision-maker, or beneficiary/donor to capture both perspectives.
- Are the goals of both sides aligned (e.g., donors and beneficiaries) or potentially misaligned (e.g., Facebook users and advertisers)?
Validation Patterns
- Divide and conquer: As the user side is riskier, start there.
- Use a Mafia Offer campaign (high-touch problem discovery, solution design, offer testing) to uncover your early adopter’s unmet wants and design/refine/position your product to cause a switch.
- Users in this model don’t typically pay with money but with engagement. Identify a key action that aligns with the right engagement, e.g., publishing posts.
- Use the Mafia Offer to land superusers, e.g., influencers.
- Then, scale your Mafia Offer campaign using good product-led recipes.
- With this underway, start engaging the customer side earlier than you think. Again, use the Mafia Offer campaign to understand and gain tangible commitments with a timeline.
Validation Anti-patterns
- Architecting for or relying on virality
Validating Marketplace Models
Unlike the multisided model, where users are on the riskier side and can be tackled serially before customers, in a marketplace model, both the buyer and seller sides must be tackled simultaneously.
Sure, some marketplaces will naturally be buyer-led while others will be seller-led, allowing you to start building out one side before the other. But ultimately, you need to bring both sides together simultaneously to conduct a transaction.
This is the most complex business model archetype because two customer factories need to be firing together.
A key pattern for success with this model is first identifying a preexisting marketplace with lots of transactional friction. If you can remove some friction for your early adopter buyers and sellers, you represent a compelling value proposition that draws buyers and sellers from their existing alternative(s) to your marketplace.
- eBay did this for the collectibles marketplace, where the existing alternatives were garage sales and antique shops.
- AngelList did this for the startup funding marketplace, where the existing alternative was hitting the pitching circuit.
- Airbnb did this for the rooms marketplace, where the existing alternatives were hotel rooms and couch surfing.
Business Modeling Patterns
- Identify both sides of the model on a Lean Canvas
- Use labels like buyer/seller to capture both perspectives.
- Do the early-adopter triggering events of both sides bring buyers and sellers together?
- Determine the riskier side (usually demand) by calculating the expected buyer-to-seller ratio.
Validation Patterns
- Overcome the cold-start problem by adopting the non-riskier side, e.g., Airbnb becomes the first seller on the platform and rents out their room with an air mattress.
- Maximize buyer/seller collision (a good thing) by narrowing down on
- Geography, e.g., Uber starting in San Francisco
- Category, e.g., Amazon starting with books
- Triggering events, e.g., Airbnb chasing sold-out hotel cities
- Use a Mafia Offer campaign (high-touch problem discovery, solution design, offer testing) to uncover your early adopter’s unmet wants and design/refine/position your product to cause a switch. Do this for both buyers and sellers.
- Use a Concierge (+ Wizard of Oz) MVP to launch and test the marketplace, i.e., become the broker.
- Understand buyer/seller motivations and tackle both simultaneously, keeping the ratio balanced.
Validation Anti-patterns
- Starting with a marketplace platform
