CHAPTER 8

Data Products as a Business

Business models based on data products are even better than the asset-less business models because they do not have to mobilize asset suppliers and service providers. Data products are sold like physical products but do not incur the cost of manufacturing, supply chain management, services and repair, and more to the same extent as physical products do. Thus, the organizational model becomes even lighter as no divisions are needed to manage these functions. Indeed, two of the big five asset-light companies that Buffet mentioned in his 2018 speech to shareholders—Google and Facebook—sell pure data products.1

As with other digital transformations, the transition to data products was not instantaneous. Many of the earlier products were offered for free to lure customers, and many companies like Google and Facebook had to create new ways to monetize data. A freemium model is unthinkable in the physical world because of the manufacturing costs. But even this is changing when data can make the utilization more valuable than the ownership of the asset, as we will discuss in the chapter on Products-as-a-Service business model.

Today, data products take a big share of our daily attention. Smartphone users spend 3 hours and 10 minutes per day on the phone of which 90 percent is spent in apps.2 What if we add the extra time spent on data products on desktops, tablets, smart watches, and other devices? The statistics are clear that today people spend more time interacting with data products than with physical products. My car dashboard tells me that I spend an hour and half per day in the car. But even then, my driving experience is infused with digital experiences. I look at the GPS, listen to streaming radio or audio books, and do things that drivers should not do—check Facebook, Instagram, or LinkedIn. The navigation system further engages me by delivering information that I do not want to ignore—traffic conditions, roadside attractions, and much more. Many drivers admit that they look more often at the car computer screen than at the rearview mirror. We also check the data apps while we bike and run to get feedback on our performance despite the dangers of swerving or stumbling.

In 2014 Forbes magazine defined three types of data products that were on the raise: (1) benchmarking data products used to compare performance, (2) recommendation engines that assist people in making choices, and (3) predictive data products that help users form expectations about the future.3 However, today those are more likely to be found as features of data products. A simple data product is the weather app. It repackages weather data from hard to access data sources and makes it conveniently available to consumers. It is so useful that people check it nearly as often as they check the time. And it provides you with weather comparisons between locations and weather predictions.

Data products are not only here to stay; they are winning over by replacing or augmenting the physical products, thus gaining more users and more usage. The online encyclopedias have made the print editions extinct. Location and mapping apps have won over paper maps, but the cute pop-up city maps invented by Stephen Van Dam are still hanging in as many tourists love them. The endurance of this palm-size origami paper map is largely due to its aesthetic appeal and perception as a novel artform. When these enduring physical products are in coopetition with similar data products, their utility is often augmented by the data products. If information is missing on the pop-up map, it can easily be obtained from Google, Yelp, or other apps that offer very relevant and up-to-date information.

All physical products are constrained by size, shape, weight, and time. Data products have the competitive advantage of being infinitely extensible without constraints. The origami map may fit in your back pocket, but the unfolding can never pack as much information as a data product can deliver. Printed information can only be relevant as of the day of the printing. But data products are like “breaking news,” offering the most up-to-date information. Checking Yelp can show you the hot restaurant that just opened. Whether we are talking about search apps, review apps, chat apps, or social media apps, they all make us feel in-the-know.

These examples point to the existential dynamics between the physical and digital products. As the transition from physical to digital unfolds, the product markets are full of competition, coopetition, creative destruction, and previously unseen innovations. Some physical products will become obsolete and others will be augmented with complementary experiences. New products are emerging to satisfy needs and wants for which no physical products ever existed. There are also physical products that theoretically can be replaced by digital products, but it is not practical or economical to do so. Many products can only be offered as physical products, but even they are being transformed through digital fusions that we will discuss separately later in this book.

By understanding the competitive dynamics between the physical and data products, companies can make better decisions about product launches and product life cycle management. Timing the transition to digital products too early may not attract customers; while timing it too late may result in sunk costs. The failure of Encyclical Britannica management to see the obsolescence of its print edition in the raise of online competitors has been made the premier case study for late change management. Why did they need a catastrophic event—the plummeting of sales from 120,000 to 30,000, to consider and accept the necessity for change?4 Missing on opportunities like we discussed earlier in this book is the side effect of past success. Emotional attachment and hardwired belief that habits do not die easily is why managers resist to acknowledge the apparent obsolescence of long-term successful physical products.

Product Adoption and Replacement Dynamics

There is a strong relationship between habits, needs, and convenience that allows us to better assess the competitive dynamics between physical and digital products. The Competitive Dynamics Matrix below illustrates the relationship between habits and convenience in satisfying a need with a product. The matrix can be used to determine the timing, longevity, and coexistence of physical and digital products.

The “Habit Stability” dimension shows the relationship and transition from stable and well-established habits to newly forming habits. Habits make behaviors automatic. Once we learn how to ride a bicycle, we stop thinking how to keep balance. Automatic habits keep people entrenched in their behaviors and product preferences. New habits are formed either because new needs arise or because existing needs can be satisfied in a new and more beneficial way. Replacing old habits with new habits is not effortless. Psychologists call this process “the habit replacement loop.”5 To learn how to use Google maps versus a paper map requires attention, focus, and purposeful repetition. People are easily convinced to try new ways of doing things. But they also become easily frustrated when their automatic actions do not yield immediate results. People subconsciously benchmark how quickly they get things done the old way versus the new way, and feel intimidated when things they used to do without much thinking become suddenly extra hard. Their pride of having mastered a task suffers when learning a new habit.

The second dimension—“Need to Use,” is the force for change in life. The more convenient it is to use a product, the more incentives people have to change their old habits. Data products have a distinct advantage in this aspect as they are portable. They fit on a phone, tablet, and laptop. They can be accessed from one’s own device or from any connected device. The physical products take space and have weight. Miniaturization, the process of making physical products smaller and less bulky, evolved precisely to alleviate this pain. While we can carry only a limited number of physical products, we can load or access unlimited number of data apps on an origami-size smartphone. The convenience of portability has created an insatiable demand for data products. Wherever we are and whenever we need information, we can get it through a data product that is or can be downloaded on the phone (Figure 8.1).

Figure 8.1 Competitive dynamics matrix: Data products

Based on the Competitive Dynamix Matrix we can see that it is easiest to launch a new product that requires the formation of new habits when there is not an existing physical product. Personal measurement apps such as the apps that track sports activities or health care status are just two examples. Strava (strava.com) allows people to record their sports activities and track or benchmark their training and progress. People who train know how important such data is. But aside from stop watches there were no physical products to do this easily. With Strava and other similar apps one can track the activity with unprecedented detail—speed, distance, route map and segments, elevation, heart rate, and much more. The benefits are obvious and desirable and thus drive instant adoption. The learning experience is not obstructed by the habit replacement loop.

It is also easy to see why location information apps are easy to adopt. The ability to locate and find a route to a destination wherever you are and whenever you need it gradually wins over even the most habitual individuals. It often happens through the force of necessity, as sooner or later every person finds himself in a situation where he needs directions but does not have a physical map in his briefcase. Once Google rescues you from such a situation, you become a convert to the new way of doing things. Because life exposes us to many such AHA moments, the data products keep finding new user segments. My sister was a staunch opponent to using Facebook until she went on a charitable mission in Kenya. When she left, she wanted to stay in touch and share stories and pictures with the local community. The convenience of Facebook was undeniable and now we all suffer picture overload from this new staunch Facebook user. Even though she still prints pictures and shows them to people, her old habit coexists with the new habit.

There are many digital products that call for the formation of new habits but do not replace or make obsolete the old habits. Applications like Kisi create digital keys that unlock the doors in office buildings. There are many variations of Kisi. Some hotels deliver your room in their branded mobile apps. Thus, you do not have to see the concierge. These apps are replacing the plastic key cards and the traditional metal keys. But the adoption is slow, and the old habits coexists with the new habits in clearly delineated use cases. While many people use digital keys to get in the office building and to their office floor, many still prefer metal keys for their personal offices. While some people use the hotels’ apps to open their rooms, others like to stop by and get their key card from the concierge along with recommendations for nearby restaurants and other amenities. And not many people are replacing their home keys with digital keys. This is driven by the Need-to-Use as there is no compelling reason to make a universal switch to digital keys.

Finally, there are physical products that can be replaced but it is not economical or practical to do so. Such products are typically used only on occasion and in very specific settings like the coffee table books and magazines that we see in waiting rooms and hotel lobbies. Even though people can find the same information online, they prefer to browse the physical books and magazines. The activity is not purposeful. The person just happened to be there, and something grabbed their attention—the cover image, the title, or just the colorful combination on the front cover. If they wanted to discover or find some particular information, they will do so on their phone using a data app. Coffee table books are part of the décor and are intended to distract us from thinking about time by arousing our curiosity. Neither the iPhone nor a tablet placed on the table can do that as they do not have a cover page. Devices can only alert us that something known to us or something that we are looking for is occurring. Hence, coffee table books seem indestructible at least for the moment.

There are many digital product failures because people try to replace old habits or create new habits without taking into account the nature of the habit and its relationship to the need to use a product. Some data product architects think that information is the sole driver to habit formation. But an app that tells you how many eggs are left in your refrigerator like the Quirky Egg Minder is not worth even downloading. Every time you take an egg, you see how many are left. And in the supermarket, you need a consolidated shopping list instead of a bunch of apps tracking separately each item in your refrigerator. The design starts with the understanding that the habit and the need that you want to reshape with a new data app is the entire shopping list. This may be doable or not because the idea that every habit can be digitized is false.

Indicators for Scale and Scope Potential

The Competitive Dynamics Matrix tells us how a data product will fit and exist in the physical world. The economic performance of the data product can be assessed by looking at the Monetization Opportunity Matrix. What are the key factors that affect the scale and scope potential of a new data product?

The monetization opportunity can be mapped to two dimensions that reflect the breadth and depth of the data product usage which ultimately determines its scale and scope. The “Specificity” dimension indicates whether a product is designed for a very broad or very narrow purpose and use cases. The “Frequency of Use” dimension defines the intensity of the habit. The four key factors from the asset-less business model are slightly modified to reflect the pure digital nature of the data products. The type of asset is replaced by the specificity of the product and the density of assets is replaced by the frequency of use. In other words, the physical and location characteristics of the assets in the asset-less business model are replaced with the digital characteristics of the data products. Even though the monetization opportunity is conceptually the same, the physical and digital products have fundamental differences that are reflected in the Monetization Opportunity Matrix in Figure 8.2:

Figure 8.2 Monetization opportunity matrix: Data products

Maps are typically narrowly defined. They depict a geographic area and some meaningful information about the area such as roads, buildings, and so on. But location information is a very broad category as maps can be overplayed with layers of useful information such as population statistics, historical information, commercial information, and much more. Hence, a location information product, such as Google maps, can have a very broad usage. Products like that can be overlaid with more and more information, but they reach a saturation point where more becomes overload. Hence, the broad location information apps will always coexist with location information products with narrow specificity. Data apps that provide information on parking locations such as Boxcar (http://boxcarapp.com) coexist in a co-opetition with Google maps. Data products with a broad specificity and high frequency of usage offer multiple monetization opportunities. Each layer of information can be monetized separately. They also drive the creation of large ecosystems because applications with narrow specificity can be built on top of them. A parking finder app can leverage Google maps.

Personal health apps have a broad use as people can connect multiple devices to track various vitals but low frequency of use. General purpose, less frequently used apps are harder to monetize, and thus the opportunity is limited to what the end users are willing to pay or to how much the device manufacturers are willing to subsidize the end users. Very often the device manufacturers own the apps. While this may work for large device manufacturers with many devices, smaller ones end up facing the problem of narrow specificity and low usage frequency. Users do not want to install and manage information in separate very specific but rarely used apps. The monetization opportunities increase significantly when such apps are used for continuous 24/7 monitoring. For people with chronic conditions continuous monitoring makes the app the most used and useful of all apps. Continuous monitoring can be appealing to healthy people, too, who want to control risks during exercise or on-the-job stressful activities.

We all are amazed by Amazon and Netflix’s ability to recommend products and movies that are very relevant to us. LinkedIn recommends new connections. There are many search and recommendation engines that work extremely well for very specific domains—products, movies, medical research, and so on. These utilities are data products. They have been trained and continuously learn about the behaviors and preferences of individuals to aid their choices by filtering out irrelevant information. They are close-minded experts very knowledgeable about a topic, but clueless about the rest of the world. Such data products are monetized within other applications. There are “services” supporting the core business model. There are many variants of such highly specialized data products—image recognition and classification, text classifications, and so on. Many artisanal software companies develop such highly specialized data products and sell and maintain them for other companies.

The monetization opportunities for such data products are in their ability to improve and drive incremental business. If more relevant recommendations drive more purchases, companies will be willing to invest and pay more for such data products. The growth of LinkedIn from 20 million members to 400 million is frequently attributed to the “connect” recommendation pioneered by the company. The frequency of use often determines not just the monetization but also the relevancy of the recommendations as such data products learn from past experiences.

Unfortunately, such data products cannot be generalized to expand their scope and move them to the upper right quadrant. Their value derives from their accuracy, which increases by constraining the learning and recommendations to narrow domains. It is also hard to sell them to competing companies as they provide a competitive edge. People visit and spend money on sites where products are easy to find and come across. Hence, companies keep such data products proprietary and often develop them in-house. Content classification data products do not have the same problem as they are utilities that support the business but are not revenue drivers.

Narrowly defined data products with low frequency of use irritate consumers but are necessities. “Why do I have to download an app just for that …” is the typical reaction when we have to get an app to do one thing. The NJ Transit app delivers train schedules, alerts, and digital tickets. But the schedules and alerts are not the reason why people download it. It is the digital tickets. If you want to avoid the fine for purchasing a ticket on the train, you must have the app. To overcome user frustrations, single purpose apps have to find a way to expand the breadth of useful features or merge with apps that allow the purchase of multiple types of tickets. Why not merge the NJTransit app with Amtrak or United apps? Google authenticator is also a single purpose app with low frequency of use, but because people can use it to authenticate with many sites and apps it is a must have.

Value to Consumers

Why do consumers love data products and spend so much time in apps? Data products satisfy the fundamental human needs to discover and learn new things, to seek entertainment, and to socialize. Data products have become part of all conversations as people check facts and seek information in real time during meetings and events. Apps allow us to fill or kill time with entertainment or socializing whenever and wherever we want. The less we can do this in person, the more we turn to data products and apps to fill the need. Data products have become as essential as physical products are. Many physical products are modern inventions that we learned how to use and that are now an indispensable part of our lifestyles. Apps are arguably even more indispensable. Some are useful tools that we cannot live without, whereas others provide deeper meaning as they connect and keep us connected to people.

Today and even more so in the future, many companies will rely on data products as a core source of revenues. As with physical products, managing the lifecycle of data products requires complete understanding of why, how, and what drives the adoption and the scale and scope of the data product. Because most of the data products do not start their lifecycles as obvious necessities like many physical products do, all data products have to inspire the desire to try and learn by appealing to fundamental human needs and thus drive the formation of new habits.

---