A Shift in the Efficiency Frontier in Education

Earlier, we discussed the concept of the efficiency frontier. Firms face a trade-off between lowering their costs and increasing their customers’ willingness-to-pay by providing them with better or more convenient products or services. In chapter 2, we saw how Blue Apron and Uber shifted this frontier in their respective industries to raise the customers’ willingness-to-pay while paradoxically lowering costs.

What does the efficiency frontier look like in education? If you go back in history, private teachers educating the aristocratic elite through one-on-one tutoring seems to be one of the earliest forms of formal education. The power of one-on-one instruction is obvious: The teacher can spend all her effort and attention on the unique needs of one student. Content and speed of instruction can be customized. If the private teacher comes to the student’s home (or castle), convenience for the student is also maximized. But it is very costly and inefficient from the perspective of the teacher. In the modern era, the teacher would much rather explain the concept of quadratic equations once to a class of thirty rather than having thirty students individually try to learn this skill during her office hours. From an efficiency point of view, it would be even better to give the lecture of quadratic equations in a huge lecture hall, as is commonly done in introductory courses at universities. At the same time, the student’s happiness and effectiveness of learning is reduced. We can see this trade-off in figure 5-1. (Recall that willingness-to-pay reflects the benefits that a student receives, not the price the student is actually paying.)

FIGURE 5-1

The traditional efficiency frontier in education

Private teaching is what Khan provided to his cousin Nadia, much to her benefit. This was only possible because of the love and empathy of her uncle. Let’s do a quick back-of-the-envelope calculation of the costs of having this particular private tutor. As a Harvard MBA working at a hedge fund, Khan likely made somewhere between $500,000 and $1 million per year. Even if he worked many hours per day and rarely took a day off, his average hourly compensation must have been between $300 and $400 per hour. But in an educational setting, a cost of over $300 per hour per student is not a scalable model.

Now let us turn our attention to quality, or how much a student benefits from a particular way education is delivered. Most likely, these benefits are a function of the following factors:

The quality of the instructor

The customization of the content relative to the student’s interests, career ambitions, and learning style

The degree to which the speed of instruction is customized to the ability of the student

The convenience of the educational service in terms of the timing and location of classes

Because of economies of scale, teaching classes of one hundred is much more efficient than teaching classes of ten. This is the reason why educational institutions have long discussions about faculty-student ratios. But the trade-off is that, in classes of one hundred, it is hard to customize either the content or the speed of instruction, not to mention the timing and location of the class.

This brings us back to the idea of shifting the frontier. When the instructional videos produced by Khan were uploaded on YouTube, the cost of production, including his time, was amortized by many more students. On YouTube, EdX, or Coursera, many video lectures have been watched by tens of thousands. Even if we factor in the costs of production, including expenses such as video editing and production (which make it more expensive than simply lecturing in front of students), the cost per lecture per student is reduced to pennies.

But what about the student benefits derived from such video lectures? Aren’t they just as bad as Ford’s slogan, “Any customer can have a car painted any color that he wants so long as it is black”? The answer has been the biggest surprise to all those active in online teaching: no. To understand why, let’s go back to the drivers of student benefits mentioned earlier.

First, consider the quality of the instructor. Hundreds of years ago, those seeking entertainment and amusement would go to the local market to watch a clown or acrobat. Clowns were not making fortunes, but there was enough demand for this type of labor that pretty much any town would bring enough business to support its own clown. The clown profession, however, changed abruptly with the introduction of film technology. In times of movie theaters, the production of films was centralized, reducing the cost per laugh and limiting the demand for clown labor. For clowns, this was a sad story, but not so for the audience. Because the best clowns would star in the movies, the audience now could watch those who were really funny. Teachers and clowns have more in common than most in our profession would like to admit. With 3 million teachers in the United States, we have about 250,000 teachers per grade level. If we further break this up by subject, we end up with some 50,000 math teachers in eighth grade. Every one of these 50,000 teachers will explain the concept of quadratic equations each year. Most of them will do it well. Nevertheless, the idea of watching the very best teacher on video is increasingly appealing to students and parents alike.

Who teaches where has long limited the ability of schools to offer a wide range of topics. For instance, which foreign language you learned in elementary school, if you had the privilege of learning one at all, depended on which school you attended. Few elementary schools have the resources to teach French, Spanish, Mandarin, German, and Hebrew. In contrast, platforms like Rosetta Stone that serve a national and even global market have scale. They are thus in a much better position to provide students with the language instruction they desire.

The award for Most Popular Customization Tool in Online Education (we made that up) should be given to the pause button on the video player. In a lecture hall with one hundred students, there are only so many times a teacher or professor can pause and repeat herself. Online, there are no limits. If a student is distracted, the content is difficult, or the explanation of the professor is unclear, all it takes is a click and the video is paused, allowing the student to reflect and replay. As experienced online teachers, we also learned from our students that the second-most-popular tool is that for adjusting the video speed. Apparently, when watched at 1.5 times the normal speed, some of our most boring lectures become tolerable.

Finally, there is the effect of convenience. The new generation of learners, whom we as professors now teach, grew up with online devices and are accustomed to the “anytime, anywhere” paradigm of our society. From Khan Academy to smartbooks and from Rosetta Stone to Lynda.com, convenience is a key need and expectation of many users that we as educators might not welcome but must embrace.

Before we continue, a clarifying comment is in order. Being both parents and experienced online teachers, we by no means want to imply that kids should be educated by video instruction alone. Teachers will always play an important role in education. Nevertheless, technology has changed the way education is organized and has shifted the frontier, leading to bigger student benefits at lower costs. The following sections dive into greater detail on how the repeat dimension especially can shift the frontier in education and other industries.

Create Unified Customer Experiences across Episodes: Strengthen “Recognize”

Up to this point, we have discussed customer experiences with firms by looking at one episode or transaction at a time. But the greatest potential of connected strategies lies in creating deep, ongoing relationships with customers that weave together multiple experiences. The repeat dimension is thus fundamental in transforming stand-alone experiences into relationships. The first step to achieve this goal might sound trivial, but it is essential and turns out to be quite difficult: you need to be able to identify the customer and treat him or her as the same person whenever you interact, regardless of when and where this interaction takes place. Only if you keep track of your customers will you be able to learn more about them—that is, improve the recognize dimension of connected customer relationships.

Such a customer-centric view is remarkably uncommon. For instance, in the world of education, students traditionally interact with the school or university one course at a time, and it is up to the student to stitch together a coherent experience. A connected strategy approach, in contrast, focuses on the learner, not the course. This allows the aggregation of otherwise disjointed learning experiences into one unified learning journey. Teachers and counselors have access to the data of past student performance, and no student falls through the cracks, which increases the quality of the instruction. Costs also come down at the same time by saving teacher and counselor time that is otherwise spent trying to make sense of poor student performance that could have been predicted (and avoided) much earlier.

Similarly, in the world of health care, most of us have experienced the annoyance of checking in at a physician’s office. How many times must we as patients provide our medical history, our allergies, and our insurance information? Wouldn’t it be nice if, when our sleep pattern suddenly becomes abnormal, our physician is put into the loop? Chances are that if we are Apple Watch users, Apple now knows more about our health than our doctor, for whom we are patients when sitting in the exam room but strangers when we are not.

The problem of orchestrating all interactions and weaving them together into a unified customer experience is harder than it first appears. The reason lies in the fact that many companies now interact with each customer through multiple channels. This creates at least two problems.

The first is a technological one. Complex businesses with multiple product lines often do not use a single database or IT infrastructure. For instance, when a firm interacts with a customer through both traditional brick-and-mortar retailing and online channels (omni-channel retailing), it is quite problematic to track that customer across all her interactions with the firm’s various touch points.

This leads us to the second problem, an organizational one. The reason for the multiple IT systems is often historical. Different business units develop their own processes and systems, a problem that is exacerbated when divisions are added through mergers and acquisitions. Moreover, these units often fight for internal resources or compete for status and career slots. So, when a customer who was well advised by an employee in a retail store ends up buying a product from the online branch of the same retailer, the store manager might view him as a customer who was lost to another unit. Similarly, consider Disney. To generate the amazing guest experiences that we described previously, Disney had to overcome exactly these challenges:

The data related to a given customer was scattered among (for example) the Disney video games the customer had on her PlayStation, the retail store at which she purchased the last piece of Disney apparel, the Disney movie she saw on Netflix, the Disney theme park she visited last year, and the Disney Hotel she stayed in. Integrating this into a single customer relationship is not easy, but without integration, how could Bill, who was acting as Captain Jack Sparrow in Disney’s Anaheim park, remember that little Sydney had seen Bill’s colleague, François, in Disney’s Paris park last year?

Even though they are part of the same company, theme parks have to be profitable, and so do feature films. To move from a product-line-(channel)-based view of the world to one that puts the customer in the center of all transactions requires strong vision and leadership support from above. Traditionally, it was the customer who had to navigate Disney’s organizational chart in order to stitch together a seamless experience. As part of the MagicBand implementation, organizational changes had to be executed as well.

But Disney did it, and you already know the results from chapter 1. Not only did the guest experience improve, but in many cases, the costs also dropped. Where it was once necessary to manually weave together transactions across channels and time when handling special customer requests or complaints, now a seamless customer experience can be delivered with high efficiency.

Improve Customization Based on Past Interactions: Strengthen “Request”

While the first level of customization is all about keeping track of and getting to know the customer across individual transactions with the firm, the second level is about turning this information into actionable knowledge. The firm needs to use the information about a customer’s needs to translate it into a specific request for an appropriate product or service. To understand which product or service is most appropriate, the firm needs to understand which willingness-to-pay drivers are particularly important for a given customer.

In the last chapter, we introduced the concept of the customer journey (see figure 4-1). At each step of this customer journey are a number of possible willingness-to-pay drivers. Understanding these drivers is essential to customizing the customer’s experience. (We will guide you through this process in the next workshop chapter.) Most importantly, what the customer journey highlights is that your customer’s willingness-to-pay is driven not only by the product or service itself (the “what”) but also by how a customer interacts with you and how the customer can access your products.

For instance, convenience of access has become an ever more important element of customization. Again, the world of education provides an illustration. In the old world of brick-and-mortar education, physical campuses and class schedules created a rigid delivery system. In today’s world, “anywhere and anytime” has become the mantra of online education, especially in the market of educating busy professionals. Thus, physical campuses and fixed class schedules are inconveniences that negatively impact the customer willingness-to-pay.

Customization goes beyond the ability to access content anytime the learner wants. While it sounds great to have access to tens of thousands of educational videos around the clock, the options can be overwhelming. Maybe a learner wants to become a web developer. Coursera, EdX, and even YouTube have plenty of video material that the learner would benefit from. But where to start? As mentioned earlier, Lynda.com bundles videos so that they collectively correspond to a career track. It takes the expressed need of the learner (“I want to become a web developer”) and turns it into a solution (“Take JavaScript first, then take a course on interface design,” etc.). This is the idea of curation leading to customization. The smartbook at McGraw-Hill takes customization one step further still. Beyond reacting to explicitly expressed user needs, it also infers customer needs based on past interactions. Past reading and test-taking behaviors are analyzed and used for future curation.

This is the skill that Amazon has mastered so well. By observing our past browsing and buying behavior, Amazon is able to infer our needs. Moreover, it creates a virtuous cycle. The more a firm engages in business with somebody, the more it learns about the customer and the better it is able to customize future offerings. The better the firm customizes its offerings, the more delighted the customer becomes, bringing the customer back again and again, creating even more information for the firm. At some point, the customization becomes so good that customers get locked in and stop taking their business to competitors. Recent data shows that Amazon has more than 40 percent market share of online retail. This feedback loop is visualized in figure 5-2. Given one customer, a firm learns more and more about what that one customer needs. This creates a positive feedback loop: recognize, request, respond, and repeat, then recognize, request, and respond even better, and so on.

FIGURE 5-2

Learning at the level of the individual customer

Learn at the Population Level to Enhance Product Offerings: Strengthen “Respond”

We recently worked with a telecommunications executive who shared the following story. He was at the checkout counter at a large home-improvement store. The cashier asked for his zip code, to which he responded, “I will tell you my zip code if you give me a 5 percent discount. In fact, if you give me 10 percent off, I will tell you the street I live on.” The clerk called the manager, who took the deal!

It is said that, in this connected world, customers pay not just with their wallets but also with their data. We will discuss this theme further in chapter 8. For now, let us simply observe that knowing your zip code and your street address does not just allow the store to serve you better; the store can also transfer this knowledge to better serve other customers like you. Conversely, it can use the data on customers like you to help in predicting what you might need. This is the first advantage that comes from population-level learning. The firm can move beyond using an individual’s data to help that person by using aggregate data to make customized suggestions or decisions for each of its customers.

Population-level data makes even more powerful learning possible. By learning about its customer population, the firm can create a better product or service offering. After all, what good is it to have a deep understanding of your customers’ needs if you don’t have the products or services available to satisfy those needs? True customization requires not only understanding the customer deeply but also having the right product and services available. Thus, level 3 of customization is fundamentally about strengthening your ability to respond.

First, consider examples from the world of education. Learning analytics is emerging as a hot new field. If we can predict which students are likely to struggle in a course, we can take corrective action before problems occur. Teachers can learn where individual students or entire classes are likely to struggle, allowing them to proactively alter what they offer in their courses. The same can be true for authors like us. If, for example, we knew that learners love the worksheets in chapter 3 but rarely use the ones in chapter 10, we could improve this book. In fact, we hope to achieve exactly this through our website, connected-strategy.com.

A parallel trend is playing out in medicine. Under the label of personalized medicine or precision medicine, health care companies mine genomics data in the hope of finding predictive patterns for who will develop Alzheimer’s or cancer, among a range of illnesses. For example, the genetic testing firm 23andMe is establishing itself as a valuable partner to biotech companies as it amasses the genetic profiles of millions of people.

As a firm learns more about its customers, it can also broaden the set of customer experiences that it creates. Consider Square, a financial services provider founded in 2009. Square started out by providing small businesses with a lower-cost option for accepting payments via credit cards. Through its Square readers (small electronic devices for swiping cards), Square helps its clients to improve their respond-to-desire strategies. Over time, as Square learned more about the needs of its clients, it created curated offerings that included new features such as tailored dashboards providing information about the end customers and new services such as payroll systems. The information contained in the Square system also allows small businesses to provide more curated offerings to their customers—for instance, via targeted advertising. Lastly, Square has started to offer an automatic execution experience by automatically issuing lines of credit in real time based on the merchant’s cash flow.

As these examples illustrate, population-level learning allows firms to refine their product portfolio in two different ways. First, learning about demand allows a firm to better choose which products it should carry. The second type of portfolio adjustment is more radical. As a firm learns more about its customers, it might get deeper insights into them than any of its suppliers have. These insights might then allow the firm to backward integrate and produce (or direct suppliers to produce) brand-new products. Consider Zalando, one of the largest German online fashion retailers. Zalando started as a copycat of Zappos, the largest online retailer of footwear in the United States. Zalando initially focused on providing a respond-to-desire customer experience. Over time, as Zalando learned more about its customers, and customers were willing to share personal information and fashion preferences, Zalando was able to add curated offering activities, matching individual customers with selected items that are presented to them through the company’s website. Eventually, Zalando was also able to use the data it gathered to start a private-label brand. From searches on its website, Zalando had customer data for which price points and product categories customers rarely used a “brand” filter. Zalando realized that for these products, its customers didn’t care much about the brand name. Therefore, Zalando started to offer its own products in these categories. (See the sidebar for another example of repeat in action.)

Again, we can observe a virtuous cycle, a positive feedback loop. The larger the set of customers a firm serves, the more information it can gather to fine-tune its existing product portfolio through better assortment or the creation of new products. The better its product portfolio, the more likely it is that it can find a good match between the needs of a customer and its product offering. This good match, in turn, leads to customer satisfaction and expands the customer pool, again creating more data.

FIGURE 5-3

Population-level learning

Figure 5-3 illustrates this positive feedback loop. Unlike figure 5-2, which was all about finding what is best for one particular customer in order to provide a better curation, metadata enables learning about many customers.

Become a Trusted Partner to the Customer: Recognizing Deeper Needs

As a firm learns more about its customers, it also has the opportunity to move from addressing one or more narrow needs to focusing on more fundamental ones. A narrow need is to learn about compound interest rates. A more fundamental need is to be able to ascertain the value of an investment. More fundamental still is the desire to become an investment adviser. When a learner entrusts her career dreams to Lynda.com, customization can be done at a whole new level because Lynda.com can take a more active role in the connected relationship. Beyond curation, the firm might nudge the learner to keep on top of her homework (coach behavior) or even automatically sign her up for an important job fair.

The distinction between narrow needs and more fundamental ones is also relevant in the field of health care. If a patient feels some heart palpitations, the narrow need is to talk to a cardiologist. More broadly, what this patient wants is to have the health care provider deal with her cardiac problems. Actually, what the patient really wants is for her health care team to provide the right health care when needed. Most fundamentally, what the person wants is for her health care team to keep her healthy. Thus, we can identify a hierarchy of needs in which the current request is an expression of a higher-level, more general need. The promise of connected strategy is that through repeated interaction, a firm is able to move up this hierarchy of needs and embed each user experience in a deeper relationship between the firm and the customer. In doing so, firms can address more fundamental drivers of customer value, increasing a firm’s value proposition.

A helpful approach to discover such deeper relationships by addressing more fundamental needs is the why-how ladder. Figure 5-4 provides an illustration for our cardiology example. Each of the boxes in the why-how ladder corresponds to a specific problem definition. Problem definitions at the bottom of the ladder are more focused, addressing how a need could be fulfilled. We climb up the ladder by asking why. Why is that problem relevant in the first place? Why would it be good to fulfill this customer need?

Going up the why-how ladder accomplishes two goals. First, it aligns the search for solutions with what the customer really cares about. Again, patients don’t really care that much about their cardiologist; they just want to make sure that their heart is in good shape and, even more broadly, that they are healthy. That’s the most relevant problem for the patient, and whoever provides the solution to this problem is likely to win the competition to serve this patient.

Second, this understanding opens up alternative solution approaches: solving the problem of providing easy access to a busy cardiologist is hard. At this level in the why-how ladder, our solution space is limited to finding more cardiologists and making them work faster or longer hours. But as we go up to the problem of keeping a patient’s heart healthy, there exist many alternative solutions, ranging from changing exercise routines and nutrition to reinforcing medication adherence. For every dollar that we spend, we might be able to improve patient cardiac health by a lot more if we invest in methods to reinforce medication adherence or lifestyle management. This improves efficiency.

FIGURE 5-4

The why-how ladder for cardiology problems

In the eyes of the customer, the purpose of the relationship with our firm is to …

Researchers at the University of Pennsylvania conducted clinical studies on cardiac health and found some interesting results. Prior studies had shown that many of those discharged from the hospital after being treated for major cardiac problems were not willing or able to stay on their medication for longer than six months. Using pill bottles connected to the internet, the Penn research team could quickly detect when patients forgot to take their medications. By automatically hovering over the patient this way, deviations can be detected early and patients can be engaged and trained to form healthy behaviors. The researchers used small financial incentives and peer pressure built through social media to coach patient behavior, nudging them to stay on their medications and lead a healthier lifestyle.

Clearly, a firm needs to earn the trust of the customer before it is permitted to manage a more fundamental need. This is why we put this at the fourth and highest level of how the repeat dimension transforms customer experiences into customized, connected customer relationships. There is an interesting circularity here: Only if you have a deep connection with a customer—relying on intensive data exchange—will you be able to address more fundamental needs. At the same time, unless you are able to address more fundamental needs, customers likely won’t want to engage in a deep relationship with your firm in the first place. Deep, embedded connections can be intrusive; customers will have serious and justifiable privacy concerns. Unless the value delivered to the customer is high, customers will not want to engage deeply or may feel that their data is being exploited without their consent. Thus, you won’t be able to jump straight to level 4 of customization. Level 4 is achieved in stages. The customer allows you access to a certain amount of data. Once you have proven to the customer that this data enables you to make the customer’s life better, the customer might grant you access to the next slice of data.

As you can see, the repeat dimension at all four levels of customization helps a firm to shift the efficiency frontier. A better understanding of customer needs, a better ability to translate those needs into specific product requests, and a better assortment of products that fulfill those needs precisely all increase the willingness-to-pay of customers. At the same time, a better understanding of demand allows the firm to avoid inefficiencies. Table 5-1 summarizes the four levels and their impacts on willingness-to-pay and fulfillment costs.

The Importance of the Repeat Dimension for Creating Sustainable Competitive Advantage

The repeat dimension of connected strategies moves the relationship from episodic transactions to a continuous relationship. Once the individual transactions are woven together into a customer-centric, unified experience (level 1), a firm has set itself up to serve its customers better and more efficiently. This improvement, and the associated shift in the efficiency frontier, is made possible by two learning mechanisms, summarized in figure 5-5.

The first mechanism plays out at the level of the individual. As a firm engages in more interactions with that customer, the firm better understands the customer’s current needs and what products or services would best fulfill those needs. This is level 2 in our framework. For respond-to-desire customer experiences, the firm also can help the customer in understanding and expressing his or her needs more precisely. Thus, the first mechanism across levels 1 and 2 strengthens the dimensions of recognize and request.

While it is wonderful to have a deep understanding of your customer needs, this information is not very valuable unless you have the products or services available to satisfy those particular needs. The second learning mechanism operates at the level of the population (or the segment) by analyzing metadata. This learning creates a feedback into the assortment of products or even creation of new products in the first place: “Given what we have learned about customers of varying types, what would be the optimal assortment to carry or products to create?” In short, this learning mechanism improves the dimension of respond. This is level 3 in our framework.

FIGURE 5-5

The positive learning feedback loops created by the repeat dimension

Together these learning mechanisms allow a firm to enhance the personalization of its offering. The firm can create a better fit between the needs of the customer and the product (or service) that responds to this need. The more Netflix knows about Samantha’s viewing habits, the kinds of films her friends are tweeting about, or perhaps her upcoming vacation plans, the better Netflix is able to personalize her viewing recommendations (“Flying to Italy? Watch Tuscan Wedding to get into the mood!”). At the same time, as Netflix learns more about entire customer segments, it can optimize not only what kind of content to license but also what kind of content to produce. The data Netflix is able to gather from its more than one hundred million subscribers worldwide has allowed it to create more than twenty-seven thousand genres, including genres such as 20th-Century Period Pieces Based on Classic Literature, Absurd Opposites-Attract Comedies, and Biographical Fashion Documentaries. This fine-grained categorization, combined with viewer feedback and observed behavior at both the individual and the population levels, gives Netflix deeper insights into its audience than any movie studio could ever hope for.

Eventually, Netflix or other firms will be able to use this information to move up the hierarchy of needs of their customers and achieve level 4 of customization. Yes, a customer wants to watch a movie at certain times, but the deeper need might be entertainment. Once a firm understands a customer deeply, not only can it suggest movies, but it can also arrange for tickets to live concerts, automatically record sporting events, and play the customer’s favorite music in her house and car.

What makes the repeat dimension so powerful is that it involves positive feedback effects that over time can create a tremendous, sustainable competitive advantage for a firm. As we see in figure 5-5, the tight fit between customer needs and available products—that is, the high degree of personalization—leads to more value created by the firm, either in the form of higher willingness-to-pay by the customer or by higher efficiency. This allows the firm to provide more value to current customers, creating more future interactions with these customers, which increases the individual-level learning (the top feedback loop in figure 5-5). At the same time, the increased value allows the firm to attract new customers, thereby enhancing the population-level learning (the bottom feedback loop in figure 5-5). With more learning at the individual and population levels, the firm continuously improves the recognize, request, and respond dimensions, creating ever-increasing degrees of personalization. It is a process that feeds on itself and can allow a firm that gets ahead of its competitors to continue to expand its competitive advantage.

Moreover, as a firm is able to improve its knowledge about its customers’ needs and its ability to service these needs, it has the ability to move up the hierarchy of needs of its customers. Once the firm has transformed a series of customer experiences into a true relationship, customers will be much less likely switch to other firms. Firms with established connected relationships with their customers do not have to compete transaction by transaction for the business of their customers because they have created an effective lock-in. To woo customers away, competitors have to work much harder than simply offering an occasional better deal. As a matter of fact, if you are able to reach the status of a trusted partner, customers are quite likely to become advocates for you, telling their friends about the great service they receive.

In our foregoing discussion, we stress the learning feedback loops of the repeat dimension, as they have been most underappreciated and underexploited in our experience. As a firm gains more customers, three better-known positive feedback loops can also arise that will further strengthen a firm’s competitive advantage.

First, as a firm attracts more customers, it will enjoy economies of scale: fixed investments can be spread over a larger customer base. For instance, Amazon’s investments in recommendation engines, website design, and technological improvements in Alexa can all be spread over its millions of customers, giving it a cost advantage over firms with fewer customers. Economies of scale allow a firm either to offer increasingly better products without having to raise prices, or to lower its prices to its customers—or both.

Second, as firms attract more customers, network effects can arise. Network effects exist when the willingness-to-pay of customers increases with the number of other users. For instance, the more that people use Facebook, the more likely it is that the next user will pick Facebook because all his or her friends are on this platform. That, in turn, increases Facebook’s user base even more.

The third positive feedback loop is a two-sided network effect that exists when more participants on one side of a transaction increase the value for the participants on the other side of the transaction, and vice versa. For instance, the more customers Apple is able to attract to its App Store, the higher the incentives for software developers to write apps and post them in the store. At the same time, the more apps that are available, the more customers are attracted. Likewise, the more that customers use a ride-hailing service like Lyft, the easier it is to attract new drivers; and conversely, the more drivers a ride-hailing service has, the shorter the wait times and the more likely that a customer will choose this particular service. All of these positive feedback effects create ever-increasing advantages as a firm grows faster than its competitors.

As we noted at the end of chapter 2, creating new and superior connected customer experiences is only the first step of building a successful connected strategy. If you can utilize technological advances to create a better customer experience, so can your competitors. But if you can go through the recognize-request-respond loop more often and learn more than your competitors each time you repeat the cycle, you can indeed create a competitive advantage that is sustainable. While all the firms we use as examples in this book have been innovative in creating new connected customer experiences, only those that are able to utilize the repeat dimension thoroughly, and create and exploit the various positive feedback loops, will be successful in the long run.

The Data Trust Challenge of Connected Strategies

As we depicted in figure 5-5, two feedback loops are at the heart of a connected relationship: by repeatedly having interactions with one particular customer, the firm is able to better and more efficiently serve that customer; and by obtaining information about many customers, the firm can better position itself for the future.

The resulting competitive advantage can lead to market share and profitability. That is great for the firm, but is it great for the customer? As firms perfect their service to a particular customer (upper part of figure 5-5), two investments have to be made. One investment is the data-collection and analysis effort of the firm—listening carefully to the needs of the customer and learning from one episode to the next. The second investment is made by the customer, who has to share information with the firm, be it actively by answering questions and expressing preferences (“Siri, wake me up every Monday at seven o’clock and order my coffee from Starbucks”), or by permitting passive monitoring by the firm (e.g., allowing a fitness app to track sleep patterns). So, the value that is inherent in a successful connected relationship, the force that allows the firm to shift the efficiency frontier, is coproduced by firm and customer.

This coproduction concept is not just a matter of semantics; it raises customers’ expectations of how much value they should receive from the relationship. Unless customers think they are getting a fair share, they may want to quit the relationship and you will never reach level 4, becoming a trusted partner.

More generally, firms will not be able to maintain the repeat dimension if they lose the trust of their customers. Because a rich information flow from the customer to the firm is central to a successful connected strategy, data privacy, data security, and transparent data use are absolutely essential.

Both the regulatory space and user attitudes toward privacy are likely to change over time. As a result, guidelines will evolve. Still, privacy guidelines from the Organisation for Economic Co-operation and Development and the European Union’s General Data Protection Regulation are good starting points for your considerations. To build a connected strategy, you will have to have policies that address these guidelines, including the following:

1. Collection consent:  Whenever you collect data, it should only be with the knowledge and consent of the individuals you collect it from. Customers should have the right to withdraw this consent subsequently.
2. Data quality:  It is your responsibility to keep data accurate and up to date for the purposes for which it is to be used.
3. Purpose:  You need to state clearly the purpose for which data is collected before collection starts, and that purpose should not be changed unless you notify the customer.
4. Nondisclosure:  The data you have collected should not be disclosed or made available to others except with the consent of the individuals it is collected from.
5. Safety and breach notification:  It is your responsibility to protect data against unauthorized access or disclosure. Should a breach occur, it is your responsibility to notify your customers in a timely manner (within a few days).
6. Openness:  Your customers should be able to easily understand who is collecting data and for what purposes.
7. Access:  Your customers should have the right to access the data that you have collected and to have corrections made if the data is not accurate.
8. Data portability:  Your customers should have the right to receive their data in a commonly used and machine-readable format and have the right to transmit this data to another firm.
9. Data erasure:  Your customers should have the right to have their data erased and to stop further dissemination of their data.
10. Accountability:  You must commit to being held accountable for following the foregoing principles.

Four Levels of Customization to Become a Trusted Partner

The repeat element in a connected customer relationship can often raise a chicken-and-egg problem:

To provide a customer the level of customization that fulfills her deepest needs requires a strong connection, including large amounts of data from prior interactions.

But to obtain the permission of the customer to collect large amounts of data requires that the firm is capable of providing a high level of customization and fulfilling the deepest needs of the customer.

How do you break into this seemingly closed loop? In this chapter, we have proposed that connected customer relationships get deeper and deeper over time by moving through four levels of customization:

Level 1 is about creating a unified customer experience by weaving together previously unrelated episodes. Taking a customer-centric view, potentially across channels, can create efficiency gains by eliminating data reconciliation, is more convenient for the customer, and increases the amount of information that a firm has available on a particular customer.

Level 2 uses the data from past interactions to improve customization and to learn which products or services are the most important drivers of willingness-to-pay—that is, to determine what is truly requested by the customer.

Level 3 is about developing the capability of delivering on those drivers when and where desired by the customer. Responding to customer requests in an efficient manner requires the firm to aggregate information across many customers. This population-level learning improves its assortment of product or services.

Finally, level 4 corresponds to a move of the firm to tackle more fundamental needs, evolving from offering rental cars to becoming a mobility solution or from being a provider of accounting courses to becoming a source for business knowledge.

As a firm moves from one level to the next, it shifts the efficiency frontier and strengthens its relationship with its customers, thereby creating a competitive advantage. Even at the higher levels, the firm still needs to provide a more attractive option to its customers than its competitors, but it is freed from competing for every individual transaction. As we will see in chapter 8, this allows for revenue models that truly focus on long-term value creation.