Consider the following alternatives to selling the toothbrush with a fixed price:
You could sell the toothbrush for $300 and then make your money by selling the replacement heads with a healthy profit margin, a model familiar to anyone using Gillette razors.
You could offer a subscription model: for $10 a month, a new brush head is automatically shipped to the customer. In men’s facial care, this is the revenue model that led to the emergence of the Dollar Shave Club, a startup later purchased by Unilever for $1 billion.
Both revenue models might be innovative compared to just selling at a fixed price, but they have nothing to do with connectivity. As can be seen by the examples of Gillette and the Dollar Shave Club, both strategies are used by the (poorly connected) existing razor companies.
As a firm with a connected strategy, you have a long-term relationship with the customer, including a high-bandwidth information exchange. What revenue models can you design that could not be replicated by a company relying solely on episodic interactions? Consider the following options:
You could charge the customer ten cents per minute of brushing time. Because of the connectivity, your firm measures this so you can use this information as part of your revenue model. (That creates a financial disincentive to brush, an issue that can be handled by making your guarantee contingent on certain minimal use.)
You could launch an optional app that helps the customer in her brushing behavior for a one-time fee of ten dollars or a monthly subscription fee. Such a coach behavior connected experience might alert the customer when the toothbrush has not been used for twelve hours, or when the right-handed customer brushes too much on the left side of the mouth and too little on the right side, or when the customer exerts too much pressure (note that such apps already exist for Bluetooth-enabled electric toothbrushes from Oral-B, for instance).
You could have a sensor at the brush head that automatically detects its deterioration and reorders a new head as needed, resembling the printer toner example from chapter 4.
These revenue models allow you to appropriate some of the extra value that your product creates for the customer (beyond the value that the customer is currently getting from using normal toothbrushes). Are there more alternatives? So far, we’ve only looked at dividing the value between your company and the customers. But what about other parties? Let’s continue brainstorming other forms of revenue models:
Some dentists might not be happy about the Smart Connect XL3000, but others might pay you a referral fee if your toothbrush makes an appointment at their practices. How about insurers? You could give the toothbrush to the insurance companies for free and then ask to be paid a percentage of the savings relative to past patient expenses.
You could also collect data about brushing behavior, including what time your customer gets up in the morning and when (or what) he eats. You could sell this data to Starbucks (which could provide coffee just when customers are in their wake-up routine) or to the customer’s life insurance company, alerting it that its customer seems to be on an unhealthy diet or is smoking, creating a strong incentive for the customer not to engage in these activities.
Finally, through your connection to the customers’ bathrooms (not to mention their mouths), you could become a trusted partner in oral hygiene and have the Smart Connect XL3000 be the platform on which other oral care transactions are organized, earning referral fees when customers make purchases for toothpaste or dental floss.
The breadth of these models is already spreading across various industries, as the following real examples illustrate.
In the medication adherence domain, PillsyCap has developed a forty-nine-dollar pill bottle that reminds patients to take their medications or supplements. The bottle has a simple sensor to detect when it is opened and is connected to a cloud-based server. At AdhereTech, another startup with a similar technology, the pill bottles are given to the patient for free. Pharmaceutical companies and pharmacies benefit from the technology by selling more pills, and hospitals benefit from reduced readmission. Thus, a price of zero to the patient maximizes adoption and increases the value that can be shared among insurers, pharmacies, drug companies, and health care systems. But how can we be sure that the pill has been taken after the bottle is opened? Smart pill bottles have no answer to this question. The schizophrenia drug Abilify, a pill with an embedded sensor that tracks whether the medication has been ingested, has solved this problem. The pill’s sensor is connected to a wearable patch that feeds data to a mobile application.
Similarly, consider Fitbit. Fitbit has emerged as a powerful brand for its wearable devices. Given millions of Fitbit users, the company has access to remarkable data. For example, it has access to 105 billion hours of heart rate data, six billion nights of sleep, and two hundred billion minutes of exercise. Even though all this data is depersonalized, it is still of enormous value. Fitbit is preparing to launch digital health tools for the detection of atrial fibrillation, sleep apnea, and other conditions.
The purpose of this chapter is to discuss revenue models for connected products and services. As you likely have noticed, we are using the health care space as a case study. Entire books have been written about revenue models, so our focus is on the unique opportunities for connected strategies. We do this in four steps:
We first provide a brief overview of revenue models and point to some of the key limitations resulting from episodic interaction.
We then discuss what is unique about connected relationships that could be used as part of a revenue model. We point to the increased value that is created by connected relationships; the higher dimensionality of the pricing mechanisms, which reflects more data availability; and the different timing of payments that result from the longer relationship.
We then link these idiosyncrasies of connected relationships to the revenue models and propose a framework to explore alternative revenue models.
Based on this framework, we articulate a set of guiding principles for choosing a revenue model and illustrate those with examples from other industries.
We conclude the chapter by discussing challenges related to privacy, an issue that is particularly relevant in revenue models in which customers pay the firm not with money but rather with data.
Revenue Models: A Brief Overview
Consider the history of pricing in four episodes. The first episode is haggling, still common at many bazaars. The vendor does not preannounce a price and haggles with each potential customer.
The second episode is posted prices, such as those printed on items in a supermarket, listed in catalogs of mail-order companies, or displayed on billboards. Posted prices simplify transactions, increasing convenience and efficiency. However, they enforce uniformity across customers. If Selena is willing to pay $500 for a phone but Jackson is only willing to pay $300, price discrimination between the two is difficult. Similarly, if a retailer has only one phone left in stock, it might make sense to raise the price, but this is often impossible if the price is posted.
With the arrival of the online marketplace, we entered a third episode. It became feasible to adjust prices dynamically and intelligently. As consumers, we are most familiar with, and sometimes annoyed by, airlines doing this. A flight from Philadelphia to Boston might go from $99 one day to over $400 the next, reflecting seat availability and the airlines’ ability to identify us as a likely business traveler given our travel time. The internet also facilitates more complex pricing schemes, such as customer loyalty programs or group buying.
Yet despite these variations, from haggling at the bazaar to dynamic online pricing, traditional revenue models retain three limitations:
Limited information: Given the episodic nature of the traditional (nonconnected) business transaction, there comes a time when buyer and seller have to agree on a price, be it for a toothbrush or a medication. The problem is that the value the buyer will derive from that transaction is still unknown at that time. Will the new toothbrush r
eally reduce my need for dental services?
Limited trust: One solution to dealing with limited information is to delay the final pricing decision until more information is available. For example, the toothbrush manufacturer could require the customer to pay another $500 if his teeth remain healthy. The problem with that solution is that in the case of a cavity, the customer will blame it on the toothbrush and the manufacturer will blame it on poor brushing behavior. Short of any monitoring data, the conflicting interests of buyer and supplier will erode trust between them.
Transactional friction: Even if we could overcome the limited trust and find a way to determine whether the customer’s degradation of teeth is due to poor brushing or poor product performance, we still face the problem that the customer derives value from the toothbrush every day. Yet, traditionally, paying every day is a very costly practice. Every transaction requires an administrative overhead for payment processing, which likely will separate the customer’s timing of payment from the timing of deriving value.
With advancements in connectivity and the resulting emergence of connected relationships as discussed throughout this book, we have now entered a fourth episode of pricing.
What Is New with Connected Strategies?
In this fourth episode of pricing, the three limitations just discussed are overcome by longer-lasting, connected relationships facilitated by high-bandwidth information exchange. It is possible to use a whole range of additional variables as part of the revenue model. In other words, the price can now depend on factors that previously could not be used to influence the pricing decision. This includes information about the following:
When the product was used
Where it was used
Who used it
What benefits were derived from using it
What problems occurred while using it
In short, the resulting revenue models can now be tailored to the particular use case. As a result, the connected relationship allows the firm to eliminate the three limitations discussed earlier in the following ways:
The problem of limited information can be overcome by delaying payments until more information becomes available. The most common revenue model that results from this is referred to as pay-for-performance. Payments are delayed until more information about the user benefits are known.
The problem of limited trust can be overcome as constant information flow allows for the monitoring of actions taken by two parties with conflicting interests. Such verification is also necessary for the pay-for-performance model.
Because of low transaction costs, there is no reason to lump all financial transactions into a single payment, as is common in episodic relationships. Instead, we can use revenue models such as pay-per-use (pay every time the product is used).
Thus, connected strategies allow us to create completely new revenue models. For our toothbrush, we can make the price a function of how many minutes the brush was used per day, how many different customers used the brush (hopefully with different brush heads), or the degree to which cavities could be avoided. In other words, the constant connection and associated information flow allow us to increase the dimensionality of the pricing space. No longer is there a single price printed on the box; there now exist many different options for revenue models, including different ones for different segments.
The increased dimensionality is appealing at first because it provides us with many levers that we can pull to increase profits. But, just as would be the case for anybody going from the driver’s seat in a car to the pilot seat of an airplane, too many levers can be overwhelming. This raises the question, What are the general rules on how to form new revenue models, especially those that take advantage of connectivity? The following six guiding principles will help you answer this. Each is written as an action item and illustrated with our toothbrush example in the sections that follow, as well as with case studies from other industries:
Think value creation first.
Make pricing contingent on performance.
Remember the ecosystem is broader than the supply chain.
Get paid as value is created.
Reinvest some of the created value into the long-term relationship.
Be cautious when replacing cash payments with data payments from users.
Let’s look at each principle in turn.
Principle 1: Think Value Creation First
Consider the payments received by ophthalmologists for performing eye exams on patients with diabetes—a practice generally recommended annually to prevent diabetes-related blindness. Patients don’t particularly like these examinations because they require a time-consuming eye dilation for a retinal photograph to be interpreted by the ophthalmologist. The dilation can leave the patient with blurry vision for several hours. Patients who could otherwise get back and forth to the doctor on their own often need someone to take them home. Adherence rates for exams are low, no doubt contributing to the high incidence of preventable blindness among diabetics.
A recent study showed that a commercial insurer would typically reimburse $254 for an in-office examination involving retinal photographs—about $26 for the photographs plus some facility fees and the professional service. The same insurer would reimburse a total of only $16 for the photographs when the service was performed remotely, with no payment allowed for the interpretation of the images.
This example shows that many business relationships are destroying value because of poorly aligned incentives. For example, in health care, the payer is not the consumer. Insurance companies, short of information, are concerned that patients consume too much and physicians provide too much. In the diabetes case, insurers apparently liked the higher price, pain, and friction of the traditional exams because they deter patients from having them. (Whether such preference would backfire because of higher costs down the road is a matter for debate.) Thus, they were comfortable with high reimbursements for an office visit but would only pay a small fraction for an equally effective remote service.
Before we think about how to build a revenue model for any business, we should ask ourselves what actions maximize the value in the system. Once we know the desired actions, we can think about a revenue model that rewards people for those actions. In the foregoing cases, we want diabetics to get their eyes examined and everyone to brush their teeth.
Rather than just replicating the old relationships—in the eye case, the insurers being concerned about fraud by doctors and overuse by patients—we should use connectivity to ensure that every actor makes value-maximizing decisions. For example, in the construction industry, contractors and customers are often at odds from an incentive perspective, where contractors get paid through costs-plus pricing, including an allocation for time worked. This incentivizes contractors to take more time to complete a project, harming the customer both by delaying the project and by charging a higher rate. Because customers often interact with contractors only once, the reputational loss of such behavior for contractors is low. This hurts both customers and the contractors who are actually doing a good job. Now, connected market makers such as Angie’s List connect users via crowdsourced reviews to create transparency in a contractor’s practices. Contractors who repeatedly exploit their customers will receive lower ratings, which translates into fewer opportunities. This aligns contractors’ incentives more closely with the customers’ incentives because their future business depends on their reputation.
Principle 2: Make Pricing Contingent on Performance
When deciding to buy a product, customers face uncertainty about how good that product or service will be. Whether they are consumers or firms, customers don’t like uncertainty, and avoiding it can kill a value-creating transaction.
Pay-for-performance is one way to overcome the problem. Customers don’t pay for the product or service; they pay for (some of) the value it creates for them. This is possible in a connected world because we have good data about the customer.
I
n the toothbrush example, connectivity allows us to observe the health of the teeth. This enables us to give the customer a form of performance guarantee (“If your teeth are not in good health, you will not pay a penny”) while also aligning the customer’s incentive (“If you do not brush, our performance guarantee no longer holds”), thereby avoiding the incentive inefficiencies mentioned earlier.
In general, we define pay-for-performance revenue models as revenue models in which fixed transaction prices are replaced by payments contingent on achieving certain objectives. The following examples help illustrate their use in a wide array of industries:
As we noted in chapter 1, Rolls-Royce offers jet engines and accessory replacement services on a fixed-cost-per-flying-hour basis to airlines (“power-by-the-hour”), linking revenue with performance. This is aided by onboard sensors to track on-wing performance.
Power purchase agreements (PPAs) are widely used in the solar industry. Rather than asking customers to buy the solar panels, pay for installation upfront, and wait for energy savings to trickle in, PPAs allow customers to lock in the energy produced by the installation at a fixed cost per kilowatt-hour over the term of the PPA, without owning the equipment or paying upfront.
Some consulting firms are moving from a billable-hour model to a fees-at-risk model, where a portion of the consulting fee is linked to their clients’ results. This is in response to corporations’ increasing desire for impact and outcomes over pure insights.
Principle 3: Remember the Ecosystem Is Broader Than the Supply Chain
Connected Strategy Page 16