The successful development of a new medication can take ten years or more, with much of that time consumed with clinical trials.1 But despite all that due diligence, medications raise a whole new set of questions when they finally hit the market: how many patients fail to take their medication properly, or even take it at all? Did the intended patient take the pill or did someone else? How many patients split pills? To make those determinations, should a doctor accept a patient’s word, or look at how often prescriptions get refilled and then draw conclusions? Or should the doctor and the health-care system ignore those questions as long as the patient shows therapeutic improvement?
These unanswered questions have already led to the development of a breakthrough pharmaceutical product: Abilify MyCite. This drug, approved by the U.S. Food and Drug Administration (FDA) in 2017, is a pill containing an ingestible sensor that digitally tracks whether patients have taken their medication.2 An FDA press release explains that the MyCite system “works by sending a message from the pill’s sensor to a wearable patch. The patch transmits the information to a mobile application so that patients can track the ingestion of the medication on their smart phone. Patients can also permit their caregivers and physician to access the information through a web-based portal.”3 What the Abilify pill makes available is what we call impact data.
Impact data shed light on when and how customers consume products and services, and how well these offerings actually perform. Impact data are the missing link in understanding the value a customer ultimately derives from a given purchase. Customer focus, as managers commonly understand it, emerged from the quest by businesses to identify their customers’ needs and wants. Over the last decade, companies have made progress in mapping these motivations as well as customers’ purchase processes (the so-called decision journey) and experiences. However, prior to the widespread availability of twenty-first-century information technologies, a company could not observe post-purchase behavior directly, completely, and in real time.
Now companies need to collect and use impact data to close the loop on customer focus. Every company has similar sets of questions that are impossible to address decisively without impact data: How often and how well do our customers use our solution? Where and why do they use it? How satisfied are customers?
Impact data add transparency, allowing organizations to pinpoint changes in behavioral patterns and draw more reliable conclusions about why they are happening. Organizations can improve their products and services to generate more value for themselves and their customers. The collection and interpretation of impact data create vast opportunities to improve exchanges, because the resulting insights allow an organization to adopt a revenue model that is more efficient.
At the same time, impact data create a brand new set of obligations for an organization, because they are extensive, personal, and may even expose behaviors that customers prefer to keep private. Will the company use impact data to exploit the customer relationship? Or will it use them to make commerce more efficient? These obligations create a mutual demand for accountability. Customers can demand that companies use impact data in their individual interests. Knowing they can make in-depth comparisons more easily than ever before, customers will naturally gravitate to sellers that adopt a revenue model best aligned with the value they derive. To paraphrase the old saying, customers can finally get precisely what they pay for, no more and no less. Companies in turn can demand accountability from customers to ensure that they use the product or service in a way that achieves the best outcome. Finally, and most importantly, companies can (and should) demand accountability from themselves, making a commitment to leverage impact data only to help their exchanges with customers.
Think about the car odometer. “How many miles does it have on it?” is one of the first questions a mechanic will ask when someone brings in a vehicle for service. It is also one of the most important questions a potential used-car buyer will ask. It is obvious that all miles driven are not created equal. What is not obvious, and what until recently was impossible to understand, is how different each mile driven actually is. An aggregate number of miles does not reveal who had access to the car, the conditions under which those miles were driven, and how well the vehicle performed for each individual mile. An aggregate number of miles also offers no insights into miles not driven because the car had some issue that made it temporarily inaccessible.
Nonetheless, both the mechanic and the potential buyer are likely to endow a car’s odometer reading with many layers of meaning. That single number sets expectations on wear and tear, repair needs, residual value, and the presumed intensity of usage relative to the car’s age. Car manufacturers even build assumptions about mileage data over time into their warranty offers. For example, in the United Kingdom the Volkswagen Golf, one of the best-selling vehicles in the world, currently qualifies for a three-year warranty consisting of a two-year unlimited mileage warranty and a third-year warranty with a 60,000-mile limitation.4
Now imagine that cars do not have odometers. How difficult would it be to make the same judgments about car usage and conditions, never mind draw useful conclusions from them? In lieu of a mileage reading, one could ask the current owner for impressions about his or her car usage, which in turn could serve as input to some back-of-the-envelope mathematics. One could examine the car for obvious signs of intensive or prolonged use, such as rust, worn parts, or unusual sounds. But no matter how much intuition and guesswork one applies, the truth is that no one could know anything about the car’s usage or performance with certainty. Once the vehicle left the dealer’s lot for the first time, its true ongoing history would be a mystery.
What solves the mystery is impact data. The individual customer in possession of the product may have some insight into how the product was used, but only impact data—collected independently of the customer’s memory and beliefs—can provide a complete, precise, and objective understanding. In fact, the absence of impact data can trigger vast inefficiencies: how many consumption opportunities are forgone because someone needed a product or service but could not access it? Beyond that, how many goods are used far beyond their intended or useful life spans? How many individuals or businesses fail to fully amortize their monthly or yearly subscriptions? How many customers buy items they rarely or never use? And finally, how many consumers invest in products or services that do not create the outcomes they were supposed to?
Impact data replace anecdote and guesswork. As we described in chapter 1, organizations have numerous proven means at their disposal to decipher customers’ needs, wants, and decision journeys. But they could not determine what truly happened beyond the purchase. Who was the ultimate user of the product or service? What did a customer really do after gaining access to it? How well did that product or service perform relative to its promised benefits?
Without impact data, companies need to draw inferences from measurements of repeat business, such as renewal rates, and measurements of customer loyalty, such as the Net Promoter Score (NPS), which is based on answers to one single question: “How likely is it that you would recommend [company X] to a friend or colleague?” The father of the NPS, Bain partner Fred Reichheld, explained its rationale in the Harvard Business Review in 2003: “By substituting a single question for the complex black box of the typical customer satisfaction survey, companies can actually put consumer survey results to use and focus employees on the task of stimulating growth.”5 Such proxies may have served their purpose well, but companies no longer need to make assumptions and draw inferences when data on consumption and performance are readily available.
To tell the full story of how technology creates this newfound transparency, we need to understand the technological advancements of the last two decades, and especially the last five years. These advancements have fundamentally altered the flows of information between organizations and their customers and, at the same time, altered the very nature of products and services.
These technologies fall into three broad categories: hardware, connectivity, and intelligence. The most basic hardware includes the physical objects that collect and transmit the data. Common examples are sensors and scanners, which detect and measure changes in an environment (temperature, moisture, motion, heart rates, speed, pressure, and volume, to name only a few.) Other major hardware components are communications devices that allow the transmission or receipt of data or both. Products that contain such hardware are often referred to as “smart products.”
Connectivity refers to the networks that facilitate the exchange of large amounts of data. These include the bandwidth provided by telecommunications networks (3G, 4G, 5G), cable networks, and the access to storage and applications provided by virtual or cloud computing.
Intelligence refers to how someone or something that receives impact data can transform, analyze, interpret, and apply them. It includes not only applications that make data useful to a customer but also more advanced areas such as artificial intelligence (AI) and machine learning. AI in its most basic sense is the use of computers and algorithms to emulate higher functions that humans normally perform, such as facial recognition, pattern matching, and decision making. Machine learning, fundamentally, is the process that enables these algorithms to improve themselves as they process more data.
Fitness enthusiasts show how people can benefit from this combination of hardware, connectivity, and intelligence. The heart-rate monitor strap houses the hardware (sensor and transmitter). During the workout, the smartphone—using a telecom, local area, or WiFi network—picks up the impact data transmitted by the monitor. The smartphone has an application (app) that presents the data in a useful form and in real time to the user. Most apps allow the user to store the data after a workout and send the data to other parties either manually or by opting into real-time transmission.
An AI app in this case could recommend future workouts based on impact data as well as on personal information such as weight changes, diet, fitness objectives, health parameters, and the performance records of the user and others. It could also detect anomalies in performance and generate incentives for improvement. In an analog world, a human being would have needed a stethoscope, a scale, plenty of paper, a very active imagination, and lots of patience to accomplish the tasks that AI performs instantaneously and far more reliably. Machine learning would enable that AI algorithm to improve itself and self-correct as it receives more data and better understands the relationships between inputs and outputs.
The Internet of Things (IoT) refers to the communication conducted by and between nonhuman devices. The technology consulting firm Gartner predicts that from 2019 to 2021 the number of connected “things” will almost double to 25 billion.6 These devices generate huge amounts of data, including where they are located (tracked by GPS technology), how they are performing, what they are experiencing (tracked by a wide variety of sensors on or near the devices), and who is communicating with them (tracked by anything from a mobile connection to voice or facial recognition.)
An example of how these technologies work together is the combination of a modern car and the “intelligent pavement,” which consists of concrete slabs “embedded with an array of sensors, processors, and antennae.”7 In 2018 the state of Colorado planned to test this duo on a stretch of highway. Thanks to the sensors and the communication technology, the pavement and all the cars traversing it can “read” each other and “talk” to each other, transferring data on vehicle weight, type of vehicle, speed, road conditions, and many other factors such as information on the driver, passengers, and cargo. All these interactions can be transmitted to and stored in the cloud, where they can be used to identify real-time situations such as traffic density and potential risks. Parties with access to the data can apply algorithms to identify underlying patterns, spotting everything from whether automated road signs need to inform drivers of hazards, impending weather, or sudden changes in traffic conditions to whether an individual driver may be impaired or engaged in dangerous behavior such as weaving, driving too fast or too slow, or driving on improperly inflated tires (inferred either from sensors in the tires or the weight distribution of the vehicle on the pavement).
These technologies are so pervasive, and the data they collect and transmit so extensive, that the preceding example could apply to the interaction of any device with any individual provided the requisite hardware, connectivity, and intelligence are present and engaged. Impact data not only drive transparency between organizations and customers, but also democratize it. The challenge to companies is whether they will use that better understanding of consumption and performance to make commerce more efficient. This is a question of accountability.
How organizations employ this newfound power is the essence of the Ends Game. For the first time ever, technology makes it possible to measure and understand consumption, product and service performance, and their interrelated patterns in real time, at a detailed personal level, and at scale. The technologies described above enable the tracking, storage, and analysis of immense amounts of impact data, which ultimately enables companies to bring their revenue model in closer alignment with the way customers derive value from their purchases. Companies are holding themselves accountable when they shift their emphasis away from selling ownership to charging for access, consumption, and possibly performance itself.
Think back to the odometer example. A mileage reading of, say, 55,000 miles says nothing about the patterns of consumption that led to this point, how they have changed over time, and the outcomes that driving enabled. How many trips were long vs. short in terms of distance and time? What were their purposes? Then there is the question of which individuals drove the vehicle, and when. Every driver has idiosyncrasies that can affect a vehicle’s wear and tear and performance. Tesla has recognized that fact, and now offers drivers the ability to store their personal settings or “driver profile” in the cloud, then download it into any Tesla vehicle they may drive.8 The car will automatically change the seat, suspension, mirror positions, and other features. Finally, the rich contextual detail matters. A six-hour trip of 300 miles in the winter, with four passengers and a full load of luggage, over rough, hilly, winding roads will have a vastly different effect on a vehicle than a similar trip on a perfect summer day with one passenger, better roads, and moderate traffic. All miles consumed are not created equal.
Think of the music industry as another straightforward example of how much customer behavior a company can now track, and how recent this phenomenon is. Years ago, when most people purchased compact discs, no one except perhaps the buyer knew what happened to the product once it left the music store. Even then, we think it is a safe bet that absolutely no one can say with accuracy how often he or she has played the individual songs on any given CD, nor where or when. Record labels had no way to tell what customers did with a CD, once they bought it. They didn’t even know if the shrink wrap was removed. The customers’ listening behaviors were a complete mystery. Nor could anyone know how much people enjoyed a given song. Artists and record labels could conduct surveys or trust anecdotal evidence, but they had no rich, dynamic, large-scale factual basis for making decisions. They had no impact data whatsoever.
Compare that to the knowledge that a streaming service such as Spotify currently collects. In late 2016, it launched an advertising campaign with the tagline “Thanks, 2016. It’s been weird” featuring provocative comments such as “Dear person who played ‘Sorry’ 42 times on Valentine’s Day: What did you do?”9 Spotify collects a large amount of personal data on its users—including information on the nature of their streams (what, where, when, and how)—and stores it on the Google Cloud Platform. Spotify’s proprietary algorithms then allow it to infer the “why.” As Spotify describes it: “Our system for predicting user music preferences and selecting music tailored to our users’ individual music tastes is based on advanced data analytics systems and our proprietary algorithms. The effectiveness of our ability to predict user music preferences and select music tailored to our users’ individual music tastes depends in part on our ability to gather and effectively analyze large amounts of user data.”10
That description may sound self-serving on Spotify’s part, but it encapsulates what a company can do when it has access to impact data. By observing actual usage patterns, a company can understand better than ever before why a customer makes certain choices, and therefore determine what needs and wants the customer is trying to fulfill and what outcomes they are trying to achieve. In the case of a music streaming service, a company can use impact data to get a step closer to understanding the value of a song choice—at a given time, place, and context—to an individual listener. That would have been a purely hypothetical exercise in the eras of vinyl, cassettes, or CDs. Real-time information and communications technology offer businesses unprecedented insights at a very detailed level into access, consumption, and outcomes from the customers’ perspective. Thanks to these ongoing advancements, some of which are still in their infancy, companies can measure, quantify, and communicate those benefits, often in real time at the individual level, and then design and implement a more efficient revenue model.
The Ends Game is best played in the singular, not the plural. Individual impact data allow companies to achieve efficiencies by pursuing the elusive “segment of one.” This term is not new. But with the emergence of advanced information technologies, the segment of one is no longer a theoretical aspiration—something to be approached but never achieved. It is now tantalizingly close to becoming mainstream reality.
“Marketers must leverage the power of insight-driven personalization and use a predictive or prescriptive approach to understand the needs and desires of customers,” CapGemini wrote in a report in 2018.11 Forbes described how that might work: “From now on, retailers won’t use the buyer’s segmentation where an audience acts similarly. They can create detailed digital customer profiles (DCP) and personal segments for each user. Machine learning-based algorithms (decision trees) show the retailer with a certain probability whether the buyer will perform the desired action at the suggested price or not. This is the future of retail: not to compete for a customer with price wars, but to fight for data to make the buyer’s experience as personal and unique as possible.”12
Commerce is moving toward this singular world. It is no longer about many customers’ needs, wants, and actions. Commerce is now about one customer’s needs, wants, and actions. As consumer research pioneer George Gallup apparently never tired of saying, there are billions of ways to live a life, and each one should be studied.13 Thanks to today’s scalable technology, impact data are finally available to study any individual life in all its rich detail. The challenge lies in accountability, which means cultivating the relationship between organization and individual in a manner that is sustainable and mutually beneficial.
The right revenue model is what sustains that relationship. When a company knows whether and how a customer—be it an individual or a business—uses a product or service, and how these uses relate to actual outcomes, the company has the opportunity to implement meaningful changes. It can easily charge in smaller, more manageable recurring amounts, thus opening up access to customers who otherwise could not find or afford their goods or services. This reduces access waste. Organizations can also cut consumption waste, because they can now track and influence usage in real time. For the first time, they can also conceptualize and measure the actual outcomes that customers derive, which can lead to revenue models that eliminate performance waste.
Give the same customer the same product twice, and there is no guarantee that you will see the same level of consumption, or even the same level of product performance from equal consumption. The difference is “context,” a much richer term than “occasion” or “event.” Context is the richest possible description of all the conditions that can influence access, consumption, and performance, positively or negatively. Such conditions can be external, such as the weather, time of day, and location, and can also be internal, such as a customer’s state of mind, motivation, or current level of preparedness.
Companies are now in a position to understand context in real time and respond to customers accordingly. Automotive safety is a good example. The World Health Organization estimates that road accidents kill more than a million people every year and cause some 50 million injuries.14 When cars can communicate with one another about traffic, road conditions, or other hazards, they can help drivers avoid accidents. At a personal level, drivers can also receive prompts from their own car about whether they are speeding and about other steps they might take to reduce their risk. The Swedish insurance company Folksam offers drivers such as an option, with the “long-term goal to save lives and reduce the number of traffic accidents.” Folksam’s incentives could reduce a driver’s insurance premiums by as much as 20 percent.15 The U.S. insurance company Progressive offers a similar program using either a separate device or the vehicle’s own telematics.
The powerful combination of real-time consumption patterns, personalization, and rich contextual data—all at scale—provides companies with a basis to establish and reinforce trust with their customers, one by one. That is the desirable outcome of today’s technological changes. We see no reason why the power of this combination won’t grow as new technologies emerge. But the combination is a dangerous cocktail if companies do not use it in an accountable manner.
1. PhRMA, “Biopharmaceutical Research & Development: The Process behind New Medicines,” 2015, http://phrma-docs.phrma.org/sites/default/files/pdf/rd_brochure_022307.pdf.
2. U.S. Food and Drug Administration, “FDA Approves Pill with Sensor That Digitally Tracks if Patients Have Ingested Their Medication,” news release, November 13, 2017, https://www.fda.gov/newsevents/newsroom/pressannouncements/ucm584933.htm.
3. U.S. Food and Drug Administration, “FDA Approves Pill with Sensor.”
4. Volkswagon UK, “Warranty,” https://www.volkswagen.co.uk/owners/warranty.com, accessed October 16, 2019.
5. F. Reichheld, “The One Number You Need to Grow,” Harvard Business Review, December 2003, https://hbr.org/2003/12/the-one-number-you-need-to-grow.
6. Gartner, “Gartner Identifies Top 10 Strategic IoT Technologies and Trends,” press release, November 7, 2018, https://www.gartner.com/en/newsroom/press-releases/2018-11-07-gartner-identifies-top-10-strategic-iot-technologies-and-trends.
7. J. Aguilar, “‘Data Is the New Asphalt’: High-tech Colorado Road Test to Be First of Its Kind in the U.S., May Improve Traffic and Save Lives,” Denver Post, May 30, 2018, https://www.denverpost.com/2018/05/30/us-285-smart-pavement-technology/.
8. J. B. Snyder, “New Tesla Features Make Car Sharing Easier,” Autoblog, August 21, 2017, https://www.autoblog.com/2017/08/21/new-tesla-model-3-features-carsharing-smartphone-app-key-card/.
9. A. Jardine, ““Thanks 2016, It’s Been Weird,” Says Spotify in Biggest-Ever Global Campaign,” AdAge, November 28, 2016, https://adage.com/creativity/work/thanks-2016/50063.
10. United States Security and Exchanges Commission, Spotify Form 20-F filing, 2018, https://www.sec.gov/Archives/edgar/data/1639920/000156459019002688/ck0001639920-20f_20181231.htm.
11. Capgemini, “Marketers Must Leverage the Power of Insight-driven Personalization and Use a Predictive or Prescriptive Approach to Understand the Needs and Desires of Customers,” Creating a Segment of One, June 20, 2018, https://www.capgemini.com/2018/06/creating-a-segment-of-one/.
12. A. Galkin, “Retail Switch: From Generalization to Hyper-Personalization,” Forbes, June 25, 2018, https://www.forbes.com/sites/forbestechcouncil/2018/06/25/retail-switch-from-generalization-to-hyper-personalization/#3a0682736bc0.
13. Gallup, “Recognized as One of the World’s Most Influential Americans. George H. Gallup, Founder | 1901–1984,” https://www.gallup.com/corporate/178136/george-gallup.aspx, accessed October 16, 2019.
14. B. Harper, “A Look at Ford’s Vision of the Future: Cars Connected to Everything,” Driving, August 16, 2018, https://driving.ca/ford/auto-news/news/a-look-at-fords-vision-of-the-future-cars-connected-to-everything.
15. Ericsson, “Digital Transformation and the Connected Car,” Next Stop: Smarter Cars, November 2016, https://www.ericsson.com/en/mobility-report/digital-transformation-and-the-connected-car.