The Problem: Business Model Inertia

There is a strong tendency for companies to bring a new venture to market using their existing business model or some variant of it. This can lead to a business model that is not well-suited to the new venture, and the mismatch can actually destroy value. Even when the default business model can be made to work, it very frequently leaves money on the table.

This impulse to default to the current business model has some good underlying reasons, which is why it is so common. First, the current business model has likely been optimized over many years and has a track record of delivering results—reliably, predictably, and at low cost. Why start from scratch? Second, constructing a new business model from scratch is risky. It may require developing new channels, new pricing models, and even new operating models. These take time, and mistakes are inevitable. Finally, elements of the new business model may disrupt the smooth operations of the core business, creating costs no firm wants to absorb.

These are legitimate concerns. But when a company locks onto its current business model and attempts to apply the same model to breakthrough innovations, it locks out business models that are likely to be much better.

Consider, as an example, the way that product companies have most often chosen to take their smart, connected product capabilities to market. Many manufacturers have invested heavily in the Internet of Things (IoT), adding sensors, creating cloud-based platforms for sharing data, developing intelligent algorithms for diagnosing customer problems, and creating web-based dashboards for presenting product data to users. These components frequently create value that a company should be able to capture.

Product-centric companies with product-centric business models often seek to capture value by asking a premium price for the smart product or by creating an add-on subscription service. Alternately, IoT components are often sold as adjuncts to the product, and their success is measured by their impact on pull-through sales of the core product or on maintenance contract sales. Unfortunately, customers are often unwilling to pay extra for the “bells and whistles” because the technology doesn’t, in and of itself, create value. As a result, the investment in IoT fails to capture its costs.

A shift in business model—from a product-centric to a services-led model—may enable the company both to create more value for customers and to capture more value for itself. In a services-led model, the company sells not the product but the use of the product or the business outcome the customer seeks. Digital technology, including IoT components, is a key enabler for improving the performance of the product in use from the customer’s point of view. A services-led business model broadens the value proposition to include outcomes and enables the supplier to create—and capture—substantially greater customer value. Services-led business models have been successfully employed by a number of venerable manufacturers, among them Rolls-Royce, Caterpillar, The Goodyear Tire & Rubber Company, and Castrol.

But a services-led business model is very different from a product-centric business model:

• The product itself is conceptualized differently. Rather than an object—a physical product with digital add-ons—the company sells an outcome—improved uptime or increased throughput, for example
• Revenue is based not on product sales but on customer outcomes
• Measures of success shift from product sales to customer-focused metrics; the company’s success is tied directly to the success of its customers according to their own criteria
• Design priorities change; the product may now be designed for increased life and ease of service, neither of which may be justifiable in a business model based on new product sales
• Customer relationships are reshaped. Contact with the customer is more frequent and is often focused not on the product itself but on customer outcomes and how they might be improved
• Spare parts supply chains receive closer attention. They may have to be reengineered to increase responsiveness and reduce downtime, since the supplier will now absorb more of the costs of downtime
• Contracting will evolve. Contracts may include clauses that impose liability on the manufacturer for outages or failures of the product.

In other words, implementing a services-led model—or any other model that is radically different from the company’s dominant model—may require differences in six or more elements of the business model.

What Is a Business Model?

A business model is the set of strategic choices a firm uses to create and sustain margins and to grow in a competitive environment.

The elements of a business model can be represented using the Business Model Canvas, which is one of the most popular tools used for business model innovation today [Osterwalder and Euchner, 2019]. Although the canvas is effective in creating awareness and encouraging broad thinking about the possibilities for business model innovation, it does not capture the critical elements that make a business model successful in practice.

There are three attributes of a good business model that go beyond the Business Model Canvas. These attributes are systemic: They address the interrelatedness of the business model elements and their dynamics. The three elements are:

1. Coherence: The parts of the business model are designed to reinforce one another. Brainstorming of the elements of the business model is unlikely to create a coherent whole
2. Competitive Advantage: The business model creates points of important differentiation in the marketplace; for established companies, this is often the leveraging of an existing asset in a new way
3. Economic Leverage: The economics of the business model get better as the business scales.

None of the attributes above is captured on the Business Model Canvas (though there are currently attempts to augment the canvas to address some of the gaps).

How Do You Do Business Model Innovation?

Business model innovation is difficult, but it need not be as fraught with risk as is often thought. It should be engaged in systematically and with an experimental mindset. New business models inside corporations fail for predictable reasons:

• The business tries to capture value before it has a clear view of the customer value created
• The new venture defaults to the current business model or to the one that most readily comes to mind
• The venture team fails to consider the full range of risks to the business early enough in the process
• The new venture team fails to consider fully the risks that the new business may create for the core, which leads to resistance when the business is ready to launch
• The venture team fails to do what it can to understand and manage business risks before going to market
• The business is launched without an incubation period, during which issues of profitability and scalability are addressed.

The Business Model Pyramid (Figure 7.1) deals with each of these points of failure in turn. It helps innovation teams move systematically from high uncertainty and high risk to lower uncertainty and lower risk. The Business Model Pyramid was developed at Goodyear Tire & Rubber Company, an established company with a very strong culture and an established business model, where it was used successfully to launch four new businesses on three continents [Euchner and Ganguly, 2014]. The elements of the model are:

1. Validation of customer value created
2. Active consideration of alternative business model archetypes
3. Identification and assessment of the risks of each model
4. Modeling of the specific risks to the parent organization
5. Reducing risks through business experiments
6. Incubation of the business on a small scale.

Each of these is discussed below.

1. Start with a Clear Understanding of Customer Value

It is critical to understand the value your business concept will create for prospective customers. This value should be articulated in clear and quantitative terms as early in the process as possible. Without a clear understanding of customer value creation, the business model development is likely to be wishful rather than fact-based.

It is important to stress that you cannot create value with a new business model; a new business model can only help you to capture value. One can only create value by meeting a real customer need in a compelling way for a critical mass of people. This seems obvious, but innovation teams are easily diverted by concerns about capturing value before they understand how much value they might create. The strength of the value proposition creates the value; the strength of the business model lets you capture more or less of the value you create.

A compelling value proposition must create a lot more value than it costs to deliver to make it worth the effort of developing and testing a new business model. As noted above, business model innovation is fraught with risk, uncertainty, and internal resistance. If there is a big enough prize, you can make it work; if the prize is not so big, you are very likely to fail in creating the business.

It is important to attempt to quantify customer value. The first step is usually to identify categories of value, which may require interviews and observations of customers. The next step is to make estimates of the ranges of value likely created. These estimates force you to be concrete and provide a basis for further study. Finally, you need to validate this understanding with real customers, perhaps as part of a trial or through a business experiment.

A startup I worked with uses drones to design solar installations for residential and commercial buildings. The company is able to provide superior designs, account for insolation differences in their designs, develop parts lists that minimize waste, translate the data into CAD/CAM systems—and do so notably faster than manual methods, which increases customer acceptance rates. The company had not calculated the value they created for customers. When we did the math, we discovered that their business model—based on a subscription revenue model—was capturing 3% of the value created. A different business model, which made the value creation visible to the customer, permitted an increase in value capture of three to five times, depending on the segment.

2. Identify Business Model Options

A great business model hangs together. The parts reinforce one another and bring the concept to life. There is synergy in the true sense of the word. A good business model is often replicated across industries or sectors; over time, it becomes a proven archetype.

There is a limited number of effective business model archetypes. Adrian Slywotzky has identified 23 core models (see The Art of Profitability) and explored several others (see How Digital Is Your Business?). It is useful to become familiar with all of these core archetypes, even in a digital world. They provide a basis for understanding the piece parts and power of a truly good business model: What makes it work, what levers are critical to its success, and the sources of its profitability. The underlying dynamics are often subtle, so the learning curve takes time.

The digital age has spawned a Cambrian explosion of new business models, from hyper-targeted advertising, Direct-to-Consumer (D2C) businesses, and the shared economy to multi-sided platforms and predictive selling. Many of these models have emerged in several different industry contexts, and they are evolving and being optimized over time. It is valuable to seek to understand these models, their underlying dynamics, and where they are most applicable.

Archetypes are an excellent starting place for business model innovation. Developing a new business model in practice is usually a process of structured selection: A business model archetype is selected that might be an effective means of bringing a value proposition to market, and that model is adapted to the details of the new business.

A key element in the search for business model archetypes is to look for assets that you might be able to leverage in creating a new business based on your value proposition. If you are able to build a new business on a proven business model archetype and to do so in a way that leverages core assets, you can greatly increase the chances of success and reduce time to market.

For any radically new customer value proposition, it is useful to consider three or four business models. The first may be a variant of the current business model. The second may be one that is somewhat different from your current model but comes to mind very quickly. The third and fourth alternatives will require more thought and may open up possibilities that make a real difference.

To decide among the alternatives, you need to apply them to your situation. Understanding how a model might play out in your context requires effort and attention. I suggest approaching that effort sequentially, taking the same steps for each candidate model:

1. Map out the elements of the archetype and their application in the business you are building
2. Create a very high-level P&L for the proposed business; the assumptions may all be wrong, but the exercise will give you a clearer sense of the business model and the critical economics
3. If the model looks promising, try to identify companies that are using the business model archetype in a non-competing industry; if you can, benchmark a few to understand how current practitioners think about capturing value with the model and to identify the underlying processes that make the model work
4. Get feedback on the business model from potential customers in your industry; test it as part of your value proposition.

Many companies gloss over the analysis of potential business models. Frequently, the business model decision is made by default. Even when the business team considers alternatives, it often doesn’t really study them. This superficial approach makes it more likely that the team will lock onto a model that looks attractive and lock out alternatives—perhaps missing the real opportunity to create and capture value. Adrian Slywotzky discusses business designs and how to go about creating powerful, profitable businesses in an interview excerpted below [Slywotzky and Euchner, 2015].

3. Study Business Model Risks Broadly

It is easy to fall in love with a beautiful business model, but the model must be rigorously assessed. There is a natural tendency of innovators to focus on identifying and addressing only the execution risks associated with the business. But many ventures fail because they don’t consider a wide enough range of what could go wrong. It is important to identify all critical risks early because some of them may be showstoppers.

Ron Adner has developed an approach for uncovering these risks. He calls his approach the Wide Lens [Adner, 2013]. Adner makes the point that many innovators focus their efforts primarily on reducing the risks that they understand the best—those related to execution—not on those that they understand the least. Execution risks are often managed with great care while other risks are ignored.

Adner has identified two other categories of risk: Co-innovation risks and adoption chain risks. Co-innovation risks relate to innovation that must be successfully executed by others. Such risks are increasingly prevalent in a world that relies on ecosystems for success. Ecosystem partners may or may not deliver in a timely way the critical elements necessary for the success of your new value proposition.

As development partnerships have become more common, successful companies have become increasingly adept at the management of co-innovation risks. This is true even though there are frequently issues on both sides of joint projects that impede progress. The most common of these are IP concerns, negotiation of eventual commercial terms, and arms-length development practices that assume that all of the technical challenges have been defined by requirements documents.

Adoption chain risks, Adner’s second category of risks, arise when others in the value chain to the customer decide not to cooperate in making your innovation successful. An adoption chain partner may be part of your business model’s distribution, sales, service, or supply chain. It might be a regulator or a platform provider. Anything new that is required of an adoption chain partner must be worthwhile to that partner, both economically and strategically, if you are to gain their support for your venture. If your business model design does not care for these parties, your innovation may be delayed or derailed. If, despite your efforts, you cannot build the support of a value chain partner, you may need to redesign the business model so that the particular partner is not critical to its success.

Adoption chain risks are difficult to manage. They require truly understanding the needs and concerns of partners that seem far from the core innovation. Understanding these risks requires spending time in the ecosystem and envisioning what will be required. It means understanding the incentives—monetary and otherwise—required for the stakeholders to cooperate. It also means sharing the vision—and selling it—to people who have other things on their mind. The innovation must be big enough to create a sufficient surplus for this to work.

Apple has been masterful at managing co-innovation and value chain partner risks. Steve Jobs famously pressured Corning to produce Gorilla Glass to support its entry into the iPhone market, for example. At the time of the initial discussions, Gorilla Glass existed, but there were no plans to manufacture it. Steve Jobs insisted that Corning get into the business and he set the time frame to match the schedule for the release of the iTouch, which preceded the iPhone. Jobs was actively managing a co-innovation risk.

Jobs also worked relentlessly with music producers to garner their support for the iTunes store. At the time, music industry executives were very concerned about the cannibalization of their core CD business. Jobs had to create a revenue model and a vision of the future of music sales that enticed a critical mass of music rights holders to sell through iTunes. This process, which required vision, a clear understanding of the concerns of the value chain partners, and the ability to make the economic case for participation was critical to the success of first the iPod and then the iPhone.

The first step in identifying the full set of business model risks is simply to map them. Co-innovation risks are generally easy to identify, though assessing them may be quite difficult. To start, you simply need to identify all the elements of your offering and who is needed to provide them. A contractual arrangement alone, of course, does not guarantee success. It is important to take into consideration what it will take for your partner to succeed in helping you.

• Is the task they are undertaking routine for them, or does it require innovation?
• Is the customer domain one that is well-established for your partner, or does it involve new elements of risk?
• Can the partner scale up to deliver the necessary volumes at the price point desired?

The key question is this: What needs to be true for the partner to succeed?

Managing co-innovation risks means co-innovating, not simply delegating. You need to work closely with the innovation partner to provide information that may have been tacit, and both parties need to respond quickly as new issues arise. It is important that the whole team is comfortable raising issues, which can be difficult in a fixed-price vendor relationship. Establishing a working relationship at multiple levels between the partnering companies can be critical.

Managing adoption chain risks is more difficult. You need to start by understanding how the world looks today and how it will have to change once the new venture is launched. There may be many value chains that must be addressed for the product to succeed. Understanding who must do what in the new world means spending time in the ecosystem—walking through the steps in the process, from cradle to grave; having discussions with people who would need to do something differently; and seeking to elicit potential concerns when they seem to be both distant in time and contingent on success with other parts of the project.

It is helpful in assessing adoption chain risks to map relevant value chains and to capture specific areas where a change in operations might be needed. New requirements may be as simple as the availability of spare parts or as complex as process changes and modifications to IT systems that are outside of your control. In addition, the partners with whom you need to work may be fragmented, each with its own systems and ways of doing things.

It is important to ask, at each stage of the delivery process, who will need to know and do what. It can be difficult to surface some value chain issues without creating a set of specific scenarios—and trying to anticipate what might break them. Once a basic map is in place, it needs to be validated with those closest to the work. The process is inherently iterative.

Sometimes, management of the stakeholders for an entire ecosystem is overwhelming. In that case, it may be useful to envision what Adner calls a Minimum Viable Ecosystem (MVE). In an MVE, the offering might be confined to one customer segment, for example, or to a subset of the full suite of value chain partners, in order to simplify delivery.

To manage value chain risks, it is helpful to enumerate them and to assess them using a risk matrix, similar to those used in project execution. The matrix captures on one axis the willingness and ability of the partner to execute in a way that addresses the risk; the other axis captures the severity of the consequences to the program if it cannot.

Identifying and modeling risks may seem like a lot of effort, especially early in an innovation program. It pays three kinds of dividends, however. First, it increases the likelihood that you will bring a successful venture to market. With careful consideration of risks, it is far less likely that you will be surprised later, when much more is at stake. Second, it opens the door to finding ways of finessing some risks through re-design of your go-to-market model. Finally, it is more likely that you will include and account for all relevant costs.

In the second interview excerpt that accompanies this chapter, Ron Adner discusses the Wide Lens and the concept of the innovation ecosystem [Adner and Euchner, 2014]. His work highlights the fact that innovation today is increasingly complex and interdependent. It is unlikely that a company will control all of the resources necessary for success with a significant innovation. Adner introduces practical tools that innovators can use to identify members of an innovation’s ecosystem, understand the motivations of those parties, and manage their alignment in order to increase the chances of success for your innovation.

Innovation Ecosystems

An Interview with Ron Adner

When most organizations try to innovate, the primary focus of attention is on the innovation itself, including the commercial wrapper that makes it an actual business proposition. From there, attention moves to execution … but implied within the proposition of many innovations is an additional set of factors and an additional set of actors that need to come together in order for the innovation’s value to be realized.

That larger set of actors is the innovation ecosystem, and I believe that it is important to make that set of dependencies as explicit as possible as early as possible. When you think about the strategy for an innovation, you should be incorporating into that strategy all the dependencies and partnerships that are necessary for success and how you are going to align them to make the innovation work in the real world.

[There are two types of ecosystem risks to consider] Co-innovation risk concerns other innovations that need to be successful in order for your innovation to matter. You need to ask, does anyone else need to innovate to make my innovation successful? Those co-innovations can be product technologies, but they can be other kinds of changes required for success. Innovation is change, and it requires change among multiple parties.

[Adoption chain risk concerns] who else needs to buy in to enable adoption of my innovation? It is often the case that many parties need to agree to play. The critical issue when thinking about adoption chains is that the different parties will be approaching the choice of whether or not to participate based on their own concerns. They’re going to think about their specific and idiosyncratic costs for participating and how those compare with their specific and idiosyncratic benefits of participating.

The mistake that too often is made is to look at innovation and see only how much better the world would be if it were implemented. Even if we know there’s somebody in that adoption chain who’s not going to be crazy about the innovation, the implicit assumption, the silent assumption, is that they’ll come along because the overall concept makes so much sense. Time after time, we find that there are parties that won’t come along, and it turns out that a little thing can hold back the whole innovation …

The question is, how do I create value both within and for the ecosystem? … The core tool that I developed to support this is the value blueprint. You start by depicting where you are and where your end customer is, and then you fill in all the parties necessary for success [with the new innovation.] … Go through the chain step by step and ask explicitly, “What else needs to happen in order for that party to move the innovation forward?” That process will help identify the co-innovators and the adoption chain partners …

The first thing that happens when you do a risk analysis of this broader system is that you see risk, and that’s usually considered bad news. But the reason we do a risk analysis is precisely to see the risks so that we can act to mitigate them. It takes a lot of maturity to be calm in the face of these new moving pieces that we need to manage. The natural thing is to see the adoption chain risks and the co-innovation risks in addition to all your usual execution risks and ask, why am I even bothering to get out of bed? If you can do everything perfectly to create an innovation and the other guys can screw it up, why bother? …

[T]he philosophical difference between the Wide Lens approach and a more traditional approach to innovation [is that the] traditional path starts with a prototype to make sure we have a good idea, then we pilot it to make sure that we can really deliver everything required, then we move through different phases of expansion. The focus is about how to get to scale.

The alternative path, the Wide Lens path, starts with a prototype to make sure we have a good idea. But once we’ve established that, we seek to establish a minimum viable ecosystem, an MVE. The idea is to establish a commercial footprint that lets you manage a staged expansion, where you’re adding partners over time, and with every additional partner you’re enhancing the value proposition. At the end, you’re at scale with a full value proposition, but your priority has been partners and value creation rather than racing to expand the single concept you had established at the pilot level.

It’s very unlikely that you can line up all the partners for any meaningful innovation at the outset. If that’s the goal, it’s going to lead to paralysis. The principle of the minimum viable ecosystem tells us to take a step back and think about the bigger puzzle we want to assemble and then ask, what’s the smallest subset of pieces that we can put together that allows us to create some kind of initial value? …

4. Model the Risks to the Core Business

There is a variation of adoption chain risk that Adner touches on in his work. This is the risk associated with building support within your own company for innovation. To create a successful business model, you may need to leverage the assets of your company. To do so, you will require the support of the necessary functions to be successful. You may also need the support of senior business leaders who may be skeptical of the venture. It is well worth the time to create a stakeholder map to assess who in the company must support the venture and where they stand concerning it.

A useful practice for assessing internal risks is to explicitly model the effect of the new business on the core. Again, the model will be based on assumptions, but these can be reviewed, discussed, and tested. Such a model can shift the tone of internal reviews from a search for what might be hiding in the numbers to an assessment of how to reduce risks in order to move forward. Specific items modeled for Proactive Services were pull-through of tires from non-customers; increase in the share of a customer’s wallet as a result of increased customer contact; reduced roadside assistance due to lower incidence of failure; increased capture of service performed as a result of the integration of the NewCo with the service network; and increased service revenues for installation of the telematics system. Just creating the list makes it clear that there are both opportunities and threats.

Finally, it can be useful to track (as best as you can) the metrics above. The Profit and Loss statement for the NewCo may not include these items, but an extended P&L can, and it makes clear the benefits of the new business to the larger company.

5. Reduce the Risks Using Business Experiments

As noted above, a business experiment (or what Eric Ries calls a Lean Learning Loop) starts with a hypothesis—which may be about almost anything important to your venture’s success. The hypothesis needs to be clearly stated, in terms that are testable, in order to design an experiment to learn more about it. The experiment itself needs to be carefully designed so that it is effective, is executable, and can be undertaken quickly. The goal is to design a good-enough experiment that will reduce uncertainty quickly and cheaply.

Business experiments are in the real world, with real people in real contexts. Unlike science experiments, you can’t really do a business experiment in the lab. Once the experiment is designed, you run it, which can sometimes be tricky. If you designed the experiment well and you conducted it honestly, you have either validated or invalidated one hypothesis. In the Lean Startup method, building a business is setting up a series of such hypotheses and knocking them down.

Designing a business experiment is both simple and difficult. It is simple in concept, but designing the experiment so that it can be conducted cheaply and provide a “good enough” answer quickly can be challenging. The first challenge is to define the hypothesis clearly. Nothing is more detrimental to good experiments than vaguely defined hypotheses. The goal is to isolate a particular important unknown and to state clearly what you think to be true.

The hypothesis may be related to sales: “We can sell the product through an email campaign.”

It may be related to the product: “Our target segment will prefer a solution that is eco-friendly.”

Or it may relate to costs: “We can set up a customer with our new product in less than 2 hours, on average.”

To test a hypothesis, then, it needs to be stated in very specific terms. It needs to be designed to answer the question: “Is the hypothesis true?” The hypothesis to be tested might be made more specific with a concrete claim: “We can sell the product through an email campaign using our current customer list, and we will get a response rate of 5%.” This is specific enough to test.

Executing a business experiment can also be complex. There is a tendency to want to do it perfectly—to design it as if it were part of a development activity rather than a lean experiment. It might be nice to integrate the email campaign with the customer relationship management system, for example, or to develop a great landing page for the product. The key, however, is to do something pragmatic to get a “good enough” answer quickly. This often means that everything is manual, the graphics are quick and dirty, and the prototype (if one exists) doesn’t really function at all. All that is required for a good experiment is that it is close enough to reality to give a good indicator of truth.

The experiment may yield surprising results. Hypotheses in the realm of new venture creation are often wrong. You have to be willing to listen to the results of these experiments, whether they are positive or negative.

A well-designed business experiment has five attributes [Ganguly and Euchner, 2018]:

• It is focused on a key variable (not on many unknowns)
• It is specific and measurable
• It is appropriate to what needs to be learned (i.e., at the right level of fidelity)
• It is out in the world—with the market
• It is as fast and cheap as possible.

Business Design

An Interview with Adrian Slywotzky

This discipline of business design is [getting to] the answer to … six seemingly very simple questions, which turn out to be quite difficult to answer well.

The first question is, who is my customer and who is not my customer? There are a lot of problems if the answer is “anybody who pays.” In business design, deciding who the customer is that we’ll serve with our offer, and who we will not serve, is the point of departure. The first tough test is to assure that what we built really matches what this customer needs, wants, and will pay for.

The second question is, what is our unique value proposition? Why should the customer buy from us rather than from somebody else?

The third question addresses the profit model: how do we make money from this transaction or this relationship? The reason that this is a big question today and wasn’t such a burning question 20 years ago is that 20 years ago the answer was simple: the player with the highest market share made the most money, so that’s what companies aimed for … [Market share is still] very important, but it’s no longer the imperative. The key question now is, where and how will we be allowed to create profitability?…

The fourth in my list of questions and arguably the toughest is, what is our source of strategic control? That is to say, how do we protect the profitability that we create? It is important to protect profit not only from competitor imitation but also from growing customer power in the B2B world, or growing customer choice in the B2C world.

The fifth question is one of scope: what do we do ourselves and what do we procure from others or partner with others to do? … The winning models from the early part of the 20th century into the third quarter of the century were integrated … That has changed a lot, and everybody knows that.

But the question of what do we do ourselves and what we partner with others for is actually quite difficult for a couple of reasons. One is that we can’t be brilliant at everything. We need to ask what are we planning to be world-best in. Second, not everything is really important. How do we make sure that we either undertake those important things ourselves or have a supplier or partner that does not get us into trouble? Finally, especially when you look at patterns of industry evolution, the important place in the value chain often changes with time …

The last question to ask about business design is, how do we organize ourselves to make this happen? We need to think not just in terms of organizational structure but also in terms of the talent we develop or hire and the culture we create.

Answering these questions well is not easy. The overriding issue is that the answers match, that the model is coherent, so I don't wind up with a Jaguar engine and a Mercedes body and a BMW transmission and a Cadillac seat. The elements need to be mutually reinforcing; the model needs to work for both the customer and the economics …

[When considering business designs, it is important] to develop alternatives. It is common for engineers to test two or three alternative versions of a product; companies should make it a practice to develop not just one good business design but three or four alternative designs for taking the same product or service to market. This exercise can be extremely powerful; there have been differences in the value created by alternative business designs in the same sector that are not 20 to 30%, but 5 to 10 times the value capture … That simple question—What are three alternative business designs to take our new value to the customer, and which one will create the most value for us?—knocks people out of assuming the default business model is the best …

The most efficient way to develop a model for your innovation is simply to leaf through [the two dozen models in The Profit Zone] and ask, are there three or four or five alternatives that are in the feasible set that we ought to explore?…

[Companies] used to be able to live in their industry’s value chain, to butt heads in their value chain, to compete in their value chain. But this is breaking down. Now tech companies are competing against telco, against media, against consumer devices. They are expanding towards financial services, towards retail, towards education, towards health care. The thought process for successful companies is to focus on the hassle map of the customer, not on what value chain they are supposed to be part of. Their fundamental question is, how do I do a radically better job for the customer in fixing their hassle map? [M]ore and more competition will be across two, three, or four different value chains, not within a firm’s original value chain …

The complexity can kill you if you’re trying to defend your original fort. The flip side is to seek simplification, which starts with the customer. If you can see the world through the eyes and emotions of the customer, you’ll invariably see a mess. There’s no industry that has been anywhere close to optimized. The winners are people who figure out how to not get anchored in the complexity but who look at the mess through the lens of the customer. Forty percent of that mess you might be able to solve yourself, but you may have to get parts from other places. Maybe that’s a big difference: In the past, maybe 80 to 90% of the mess could be solved within a firm’s value chain. In the new world, a company will probably be able to solve 20 to 30 to 40% by itself. If we’re really looking at the problem from the customer’s point of view, there will probably be a lot of pieces that we will have to license or hire in or acquire or work with the customer to solve.

Designing good business experiments is not easy. Three mistakes are common. First, the team can do an experiment that is just too vaguely defined. Often it is the hypothesis itself that is vague and, therefore, not testable. This can result in inadvertently testing several variables at once (and learning nothing). At other times, the measures of success are not defined ahead of time, so the results are subjective. These are not business experiments—even if they are out in the world—they are explorations.

On the other end of the spectrum, a team can make the experiment too complex or too high fidelity. The prototype used may be far closer to a product than is needed for the test, for example, or the team may seek to automate data collection instead of using people to do the grunt work. An over-designed prototype costs time and money, and it usually adds very little value. But many teams feel embarrassed to go to experiment with something simpler.

The best antidote to this tendency is for the team to envision three experiments to test the hypothesis and to challenge itself for the fastest, cheapest way forward. Sometimes it is helpful for outsiders to critique the experimental design and ask simplifying questions:

• Can we use a nonfunctioning prototype?
• Can people behind the scenes make the system seem real?
• Can we collect data on paper?
• What do you expect the output to look like?
• Where can we use proxy data to simplify the experiment?

The key is to challenge yourself to be simpler and faster—and then to challenge yourself again.

A third pitfall is to allow yourself to be constrained by perceptions of what you might be “allowed” to do.

• We can’t talk to customers without going through sales
• We can’t share an idea until we have filed the intellectual property
• We can’t sell something that we don’t have
• We can’t show a customer a prototype that doesn’t meet our brand standards.

On the one hand, these are real issues that must be navigated. Often, however, they are excuses for staying in the lab and bypassing good (and valuable) business experiments.

Business experiments can be designed to address uncertainties in all parts of the business. There are eight types of experiments that I have seen in multiple contexts (see Figure 7.2).

1. Value creation experiments are designed to understand the customer value your solution can create as well as what must be done by the customer to realize that value. The first step is to identify the categories of customer value, which can emerge from observation and customer interviews. The next step is to identify metrics for each category and to estimate ranges of expected changes as a result of the innovation. This data is helpful both in understanding the basic magnitudes involved and in highlighting what needs to be measured and what baseline data is required. Finally, the data are measured and the value is calculated.
2. Willingness to pay experiments focus on customer-perceived value and pricing. They usually are conducted through simulated transactions since people cannot usually state what they would do in the abstract. The New Earth Tire case discussed in Chapter 2 is one example of such an experiment.
3. Channel effectiveness/efficiency experiments test the reach, effectiveness, and cost of proposed channels for selling your offering. Channel effectiveness experiments, again, simulate real-world transactions. The experiments measure the number of targeted prospects reached and the number who purchase, usually as a function of promotional approach and price.
4. Supply chain experiments test the ability of the supply chain to provide the inputs required for the success of the business.
5. Operational cost experiments measure the costs of providing particular capabilities. Estimates are often good enough, but it is important at times to understand the underlying drivers of cost.
6. Technology in use experiments measure the delivery of value in the real world. There is often some erosion of value from the ideal when a new concept is introduced into a customer environment. This may be because anomalies in operations were not accounted for, because the technology requires more of a learning curve than expected, or because it is not as reliable as anticipated. Experiments in the real world with real customers can help to uncover these issues.
7. Technology and human behavior experiments. People often do not adapt to new technology as you might expect them to do. They may be intimidated by it; they may feel threatened by it; it might require adaptations to their work that are difficult to make. Learning these issues early is important. Once understood, they can often be designed for.
8. Partner support level experiments. Many new ventures require working with partners. Until the partnership is tested in some experiments, the level of support that will be provided in practice is unclear. Conducting targeted trials with key partners can help illuminate differences in priorities, clarify who is to do what, and assure that incentives are aligned.

Business experiments will never eliminate risk, so how long do you continue to invest in this learning? Different companies have different risk profiles. Some companies might go to market with high uncertainty (low predictability) about the profit that will be generated, as long as profit itself is likely in the long term: Time to market may be more important to them than the risk of a short-term loss. Others will want to invest more time and money in learning about key uncertainties and narrowing the ranges of key uncertainties before going to market.

6. Incubate the Business at Small Scale

The final step in bringing a new business model to market is to incubate it at a small scale in order to learn all those things that can only be learned in the market. Incubation requires setting up a functioning business, acquiring customers, setting prices, delivering the offering, managing expenses, and building a capable organization.

It is also during incubation that many of the issues with the core business are negotiated. Decisions about product development, IT support, contracts, liability clauses, etc., are resolved during this phase (see discussion of Graduated Engagement in Chapter 5).

In addition to demonstrating the profitability (or the potential for profitability) of the business, the team must also develop alternate plans for scaling the business during the incubation period. Inside a corporation, there are often multiple paths forward. The most obvious is to invest in the organic growth of the business. This can be done either to minimize losses or to maximize the success of the business (and absorb losses as the business invests in growth). The company might also decide to acquire capabilities or assets to accelerate the growth of the business. Finally, it might reorganize some of its current assets in order to structure the new venture for success. Each alternative needs to be considered prior to the decision to bring the business to scale.

The Business Model Pyramid provides a systematic way of developing and implementing a new business. It takes time, but it avoids the pitfalls of leaping into the unknown. Again, different companies will have different comfort levels with the amount of analysis required before proceeding through any level of the pyramid. It is important, however, to consider each step deliberately, in a manner consistent with your risk profile. Doing so greatly increases the chance of eventual success.