Chapter 4. Product/Service Definition and Strategy
We all aspire to create the killer product or service that people want to buy and love using. The key to this is ensuring that the product solves an actual problem that people have, in a way that appeals to them. At a pinch, it might provide them with something new and wonderful that they never knew they needed. It sounds simple and obvious, but it can be remarkably difficult to get this right. Right now, the IoT market is skewed toward innovators and early adopters. There’s huge potential to create great new products for consumers, but they may have to contend with new types of complexity.
This chapter introduces:
Productization as part of IoT design (see “Making Good Products”)
Moving from innovation to mass-market products (see “From Innovation to Mass Market”)
How products differ from tools (see “Tools Versus Products”)
What makes a good product (see “What Makes a Good Product?”)
Building service offerings around products (see “The Product is Pleasing to Use”)
Business models in IoT (see “Business Models”)
This chapter addresses the following issues:
Why a clear value proposition is a prerequisite to great UX design (see “Why is This in a UX Book?”)
The different types of market for consumer IoT products (see “Value Propositions for IoT”)
Why consumers want products, not tools (see “Tools Versus Products”)
Why it’s important to design the service offering around a product (see “Building a Service Offering”)
How business models can shape UX (see “What is a Business Model?”)
How digital business models may start to appear in real-world products (see “Bringing Digital Business Models to Physical Products”)
Making Good Products
In this section, we’ll look at why creating a compelling product proposition is the necessary foundation for a good UX, and why it’s a particular challenge in IoT right now. We’ll also consider how an IoT product need not be a physical device: it can also be partly (or completely) a service.
What is Productization?
Productization is the extent to which the supplier makes the user value of the product explicit and easy to understand. Compelling products don’t just look good or otherwise fuel some underlying need for status (although those things are often important). They make it immediately apparent to their intended audience that they do a thing of real value for them: preferably something new that serves a previously unmet need.
Nest is probably the most famous IoT productization success story. Consumers were resigned to thermostats and smoke alarms being ugly, annoying boxes with usability flaws. It hadn’t occurred to most people that they could be better. Nest products promise to do the job better than most of the competition, in the form of attractive and desirable hardware that users are happy to have on show at home (see Figure 4-1). Of course, they are premium products with a premium price tag. The point here is not that all products should be expensive, but that a good product should fulfill a clear need for the target audience, with a usable and appealing design. This is the product’s value proposition: the user’s understanding of what the product does for them and why they might want it.
As Denise Wilton, designer (and former creative director of design agency BERG), puts it:
Never underestimate the power of a simple explanation, or a product that looks nice. If people can understand it, they can want one for themselves. They’re not scared of it. It stops being a weird thing that geeks do.1
Why is This in a UX Book?
To some of you, this may seem outside the remit you normally associate with UX design. You may work in a company where productization is handled by product management, or perhaps marketing. In others, it might be considered strategic design. UX is not always involved in identifying the opportunity and framing the solution. But most UX designers would walk over hot coals to be involved from the start, especially if they have firsthand knowledge of user needs from conducting research.
Whoever is responsible for it in your organization, it provides the strategic foundation for UX design. It’s not possible to design a great product or service experience if users don’t want, or understand, the service in the first place.
Value propositions help sell products. But they also drive UX. A clear proposition helps users decide whether to buy it in the first place, but also helps frame their mental model of the system and what it does (see Figure 4-2). When users are confident that they understand what the system does for them, they have a good basis for figuring out how it works (the conceptual model), and then how to use it (the interaction model). All the clever design in the world can’t overcome a murky or unappealing value proposition.
Why is This in an IoT Book?
Productization is, of course, not a challenge that is unique to IoT. It is included in this book as it is a particular challenge for the consumer IoT field right now. Many products and services aren’t yet offering good, practical solutions for proven consumer problems. Even where they are, the value isn’t always apparent from the product itself or clearly stated in terms target users would understand.
This isn’t a criticism of the many clever and talented people working in this field. Most of them are aware that consumer experience is a challenge.
It’s a result of the novelty and inherent complexity of the products and services. We’re still figuring out what we can do with the technology, and we’re asking users to wrap their brains around some novel devices and capabilities.
It also reflects that new technology products and services are often conceived and developed by people with an engineering mindset who value highly configurable functionality. These initiatives can often seem complex and unclear in purpose to consumers, because in trying to do so much, they fail to communicate a clear value for using the service.
There is, of course, a market for products developed to meet the needs of highly technical users. There’s also great value in products and services that help a wider range of people move beyond passive consumption of technology and learn how to construct their own solutions. For example, If This Then That offers an accessible way to coordinate different web services and even connected devices (see Figure 4-3). This is functionality that would previously only have been available to those with good programming skills.
Products Can Be Services
When we talk about IoT, we tend to focus on the edge devices: the activity monitors, thermostats, connected pet feeders, and more. This is especially true when the devices themselves look novel (such as the Nabaztag rabbit) or striking (such as the Nest thermostat).
But while the devices are a key part of the UX, they are not the whole picture. They are all dependent on an Internet service. This makes the user’s relationship with the product much more dynamic. Instead of the traditional one-off purchase of a traditional physical product, the user interacts with the provider on an ongoing basis. The user’s experience isn’t just shaped by the device—it’s shaped by the whole service. There might not even be a physical product at all: just as you can now pay for Dropbox storage or personal fitness training, so you may pay for software or storage to help you make the most of connected devices, or personalized health or energy-saving advice based around data gathered from your devices.
In this book, we use the term “product” loosely to refer to a packaged set of functionalities that solves a problem for people or fits neatly into their lives. That could be a physical device, a service, or frequently a combination of both.
But the bigger challenge is in creating products and services that work for mass-market consumers. For this audience, the functionality—what the system does and how to use it—should be transparent. The underlying technology should be invisible. The user should be able to focus on getting the benefit from the product that they were promised, not on configuring it and maintaining it.
From Innovation to Mass Market
The primary focus of this book is on creating consumer IoT products and services. In this section, we take a brief look at how technological innovations cross over into the mass market and consider what lessons there may be in here for IoT.
Innovators Are Not Consumers
In 1962, the sociologist Everett Rogers introduced the idea of the technology lifecycle adoption curve, based on studies in agriculture.2 Rogers proposed that technologies are adopted in successive phases by different audience groups, based on a bell curve (see Figure 4-4). This theory has gained wide traction in the technology industry. Successive thinkers have built upon it, such as the organizational consultant Geoffrey Moore in his book Crossing the Chasm.3
In Rogers’s model, the early market for a product is composed of innovators (or technology enthusiasts) and early adopters. These people are inherently interested in the technology and willing to invest a lot of effort in getting the product to work for them. Innovators, especially, may be willing to accept a product with flaws as long as it represents a significant or interesting new idea.
The next two groups—the early and late majority—represent the mainstream market. Early majority users may take a chance on a new product if they have seen it used successfully by others whom they know personally. Late majority users are skeptical and will adopt a product only after seeing that the majority of other people are already doing so. Both groups are primarily interested in what the product can do for them, unwilling to invest significant time or effort in getting it to work, and intolerant of flaws. Different individuals can be in different groups for different types of product. A consumer could be an early adopter of video game consoles, but a late majority customer for microwave ovens.
Geoffrey Moore identified a “chasm” between the early adopter and early majority market (which he called visionaries and pragmatists). These groups have different needs and different buying habits. Mainstream customers don’t buy products for the same reasons as early adopters. They don’t perceive early adopters as having the same needs as themselves. Mainstream customers may be aware that early adopters are using the product. But this will not convince them to try it out themselves unless they see it as meeting their own, different, needs. So products can be successful with an early market, yet fail to find a mainstream audience.
An example of this in the IoT space is the home automation market. Systems such as those based on the power line protocol X10 have been around for close to 40 years. (Early examples ran over electrical power lines and analog phone lines.) The example in Figure 4-5, from 1986, shows a system that allowed users to program and remotely control their heating, lighting, and appliances over a (landline) phone. These are all applications that still seem novel and innovative to us; they would have excited the innovators of the 1980s even more.
However, home automation remained a niche market. It was expensive. It required significant technical skill to set up and maintain. Even those mainstream consumers who had heard of home automation did not see much value in programming their heating, lighting, and appliances. Had it been more affordable or easier to use, more people might have been willing to try it out. But only now are consumers starting to see the utility of connected home products. This is arguably driven by the rise of the smartphone, giving us a metaphor for the “remote control for your life.”
What’s Different About Consumers?
Mainstream consumers are now more aware of connected devices, but they need to be convinced that these products will actually do something valuable for them. A product that appeals to an audience that loves technology for its own sake cannot simply be made easier to use or better looking. To appeal to a mass-market audience, it may need to serve a different set of needs with a different value proposition.
Mass-market product propositions have to spell out the value very clearly. Users will be subconsciously trying to estimate the benefit they’d get from your product as offset by the cost/effort involved in acquiring, setting up, and using it, and you need to be realistic about the amount of effort they will be prepared to invest in your product. The further along the curve they are, the more users need products with a clear and specific value proposition, which require little effort to understand or use. And they have a very low tolerance for unreliability. Your product has made a promise to do something for them, and it must deliver on that promise.
This is not simply a question of lacking technical knowledge, and certainly not of users being dumb. That 10-step configuration process to set the heating schedule might seem trivial in the context of your single product. But it can feel overwhelmingly complex in the context of a busy life with many other more pressing concerns. For this reason, consumers tend to be most attracted by products that seem as if they will fit into their existing patterns of behavior and don’t require extra effort. For example, ATM cards and mobile phones were arguably successful because they reduced the need to plan ahead in daily activities (getting cash from the bank, or arranging to meet).
Value Propositions for IoT
The guidelines just outlined can of course be applied to any type of product or service. But connected products can be complex and often do novel things that are hard to communicate succinctly.
Core value propositions should be straightforward—for example, a company offering smart meters may promise to “tell you where your energy spend is going,” which is relatively simple. A good test of an IoT product proposition is that end users should not need to focus on its connectivity or onboard computing: it should just make sense.
But there may be complicating factors that users need to understand before buying. You may have to explain which other systems can interoperate with yours, or who owns the user’s data and what they can do with it. You might have to guarantee how far into the future you will maintain the Internet service (if your company is acquired, goes bust, or discontinues the product).
A new product in a new market
Embedded connectivity and intelligence will fuel the appearance of new classes of product and new markets. In consumer terms, the challenge is often to convince users of your vision. You have solved a problem they didn’t realize they had, or had just accepted as “the way things were.” The Glowcaps pill bottle top reminds users to take their medication and helps the patient’s doctor track how frequently it is taken.
A new type of product in an existing market
Here, the challenge is to convince users that your product is the best solution to the problem. Perhaps it has better features or better performance. In IoT, these products may be familiar physical devices newly enhanced with sensing or connectivity (e.g., the Withings bathroom scales). Users need to understand the value that is added by the enhancements, such as easier weight tracking. They need to decide whether it’s something they want, especially if it costs extra.
It might also be a technology that offers a step change in experience design. For example, airport terminals can be large and confusing. You would normally rely on signage to find your way around, but this isn’t always clear, consistent, or guaranteed to tell you what you need when you need it. You don’t want to miss your flight, but nor do you want to end up sitting around at the gate for too long because you were cautious and got there too early. Beacons offers precise indoor location. Several airlines have been trialing the use of beacons to provide passengers with in-context information and directions (see Figure 4-7). Passengers can be directed to the correct gate more easily, based on their current location in the airport. If a passenger is running late but is very close to the gate, knowing his location might help the crew decide to wait. And if their plane is delayed, the app could provide them with a voucher to a nearby restaurant or café.
A low-cost entrant to an existing market
The falling cost of embedded computing enables cheaper alternatives to systems that used to be prohibitively expensive. For example, Lowes Iris (see Figure 4-8) and SmartThings offer DIY home automation kits at a far lower cost than professionally installed systems. You may be aiming the system at people who could not previously afford this category of device, or trying to convince those who could that you’re offering a worthwhile saving. Either way, it’s important to convince users that the system performs the basic functionality just as well as more expensive options. Any compromise needs to be something that doesn’t matter too much. You need to be clear up front how you have achieved the cost saving: is the hardware cheaper? Does the system involve more work from the user (e.g., DIY setup)? Does it provide them with less personal (e.g., automated or lower bandwidth) customer service?
A niche entrant to an existing market
Augmenting an existing product type with connectivity and potentially intelligence can create opportunities to address previously unmet user needs in an existing market. It may target a niche with specialist interests: for example, an energy monitoring system designed for those who generate their own electricity and may sell it back to the grid. Or it may introduce a premium product for those willing to pay more. The Nest thermostat offered the first intelligent heating solution with high-end hardware and polished UX design in a market previously dominated by ugly, unusable plastic boxes. This reshaped consumer expectations of what a heating controller could be, even in the part of the market that couldn’t or didn’t want to pay extra for a Nest.
Tools Versus Products
For some specific connected devices, like a heating controller, there’s a close mapping between function and value. It’s easy for people to understand what it does. That’s not enough to make it a good heating controller. But it’s pretty clear what it does, and why you might want it. It will keep the house at a comfortable temperature, and perhaps save money. Devices that are enhanced versions of preexisting product types, like bathroom scales or baby monitors, have the advantage of being recognizable as things that meet a defined, familiar set of needs. You may have to convince customers as to why that product benefits from connectivity. And you may have to address concerns they have about adding connectivity or technical complexity to the product, such as security, privacy, or usability. But at least the product is familiar.
Mass-market consumers, in areas in which they do not have deep technical or domain knowledge, generally expect a product to come designed and engineered to fulfill a specific need. The Nest Protect smoke detector and carbon monoxide alarm is a good example of a product.5 The marketing website focuses on the ways in which it is a better safety alarm (see Figure 4-9). Connectivity is only briefly mentioned, to say you’ll be alerted on your phone if there’s a problem when you’re away from home.
But many IoT services and devices can be configured to meet all kinds of needs. The onus is on the user to define their own needs and configure the device (or service) to achieve them. These are not products, but tools. Tools are often general-purpose devices, such as contact or motion sensors. The device has no inherent value to the user. The value comes when they are applied to solve a particular need, such as detecting intruders in the home, or warning you that you left a window open. The Belkin WeMo Switch (see Figure 4-10) is a tool. It can be used to turn power to any appliance on and off remotely from a smartphone, or using an automated schedule. But it’s up to users to define their own problem, realize that a smart plug could help, and configure it to solve the problem. An imaginative leap is required. In reality, many smart plugs end up being used on lamps. In our own research, users struggled to think of other uses for them (although ensuring hair straighteners/curling irons were turned off was a popular suggestion).
Services can be tools as well. The aforementioned If This Then That (which can also be used to control WeMo Switches), aims to make it easier for nontechnical users to link up and program devices and services.
Tools aren’t bad. They can be very powerful for users with technical or domain knowledge. Users who have the time and motivation to configure a system to meet their own very specific needs and aren’t daunted by the need to learn the system may really enjoy this process. This could be the home brewer who enjoys rigging his or her own fermentation chamber out of an old fridge (see Figure 4-11). Or a horticulturalist might be motivated to learn about the technology to configure a remotely controlled plant monitoring, watering, and feeding system. Tools give us the possibility to be creative and take control of our environment.
The IoT market, to date, has tended to create tools for innovators and early adopters. In an immature market that is exploring possibilities, that’s fine. But it has tended to assume that the way to reach a mass audience is to make better-designed tools.
You can’t turn a tool into a million-selling product just by making it usable. The WeMo Switch comes with a well-designed smartphone app that walks users through the setup process fairly clearly and makes it easy to set up rules to control the plug. But the onus is still on the user to use the plug creatively. It’s not actually the plug they want to control: it’s the appliance. Controllable plugs are simply a first step in the journey toward controllable appliances.
In spring 2014, Belkin and appliance manufacturer Crock-Pot released a controllable appliance: the Crock-Pot Slow Cooker with WeMo (see Figure 4-12). This allows the user to control the temperature and cooking time of a Crock-Pot remotely from a smartphone app. The context of use is a perfect fit for connectivity and remote control. Their value proposition is convenience: the meal that cooks itself while you’re out all day. Remote control increases that convenience by allowing you to adjust the timing if you’re home late. And being able to keep an eye on the device alleviates any anxiety about leaving a hot thing unattended in an empty house. It may be a niche appliance, but it’s a well-formed product solution.
Mass-market consumers don’t necessarily lack the knowledge, skill, or imagination to solve their own problems. They may be perfectly capable of doing so, but simply lack the time or have other priorities. At best, they might only have time to solve a few of them.
There is a rich market for products that solve their problems for them!
What Makes a Good Product?
Good products seem to appeal to common sense, and new good products are often greeted with the reaction “I can’t believe no one thought of that before!” But developing good products can be far harder than our 20/20 hindsight might lead us to think. This section looks at the general qualities of a good consumer product before considering what features come with IoT.
The Product Solves a Real Problem (and Makes This Clear)
Most products are acquired in order to solve a problem for the user. A good definition of the problem, and the audience, is essential to creating a clear value proposition. This is the definition of what your product does for people, and why they might want it.
A clearly communicated value proposition is fundamental to user experience. When people come across a product (or service), they try to form a quick judgment about its purpose, and whom it is for. If it’s not immediately clear what the value proposition is, it may be dismissed: either because it is too hard to figure out, or because it does not appear to do anything of value for that person at that time. Worse, potential users may wrongly assume it is able to fulfill a purpose for which it is not really suited and waste time and/or money on a fruitless endeavor. (You may be happy to take their money in the short term, but over time too many unhappy customers will damage your reputation!)
It’s all too easy to end up with a poor or unclear value proposition despite good intentions. This is often the result of failing to identify the right problem for the right audience. You might have added features to show off what the system can do, or because they are simple to build, dictated too much by the capabilities of the technology at the expense of the original purpose and user needs. Or maybe there are competing interests involved in feature scoping. It’s common for systems to try to do lots of things. That may create a great tool for early adopters who like to tinker and customize, but it risks muddying the value proposition for a mass-market audience. Imagine you’re making a wrist-top device for outdoor pursuits like hiking or climbing. The core features are an altimeter, barometer, compass, and perhaps GPS. It might be quite straightforward software-wise to add on a calendar, to do list, and maybe even games. You can probably imagine a situation in which someone, somewhere, might use those features. But you’ll be at risk of obscuring the key purpose of the device: helping users find their way and stay safe. Too much flippant functionality might even undermine the perception that the device offers good quality in its core functionality. And it will make it harder for users to access the key features they most want and need.
If your device can fulfill multiple purposes for the user, you’ll have to invest extra effort in helping users understand its value. A home contact sensor is a generic piece of hardware with no inherent value to the user. The value is in the function it enables: used to detect when an intruder has forced a door open, or when a medicine cabinet has been opened. Early adopters may love the flexibility to use the sensor as a tool that can do all kinds of things. But you’ll have to help mass-market users understand what it could be for. For example, your app might offer specific window or cupboard alarm functionality to go with the device, even if these do much the same thing under the hood.
Connected products intended for the mass market need to demonstrate a clear advantage over any predecessors. Connected things are not inherently better than nonconnected things, just because they are connected. Despite being demo-ed at consumer electronics fairs year after year, the much-maligned Internet fridge concept has so far felt like a solution in search of a problem. Research shows that people can imagine using intelligent fridges that provide information about their contents, nutrition, and health, but this has not translated into demand.7 Tasks such as managing shopping lists and looking up recipes simply don’t feel as if they require a new, fridge-based screen. The idea of the fridge that automatically orders more shopping when goods run out is fraught with potential for irritating errors. If you have to make the fridge sync with your calendar or heating thermostat to see when you’re on holiday in order to stop your regular milk order, maybe it’s just simpler to buy your own milk after all.
Connected sensors enable many kinds of data in the world to be captured, quantified, and made visible. Fitness tracking and energy monitoring (see Figure 4-13) are obvious consumer examples of this. But beware you’re not just counting things. Data should be used to provide genuine insights that users can act on.
Connectivity can enable remote control of devices. The core value of connected sockets and door locks is usually remote control (see Figure 4-14).
Connected home systems that allow automated rules to be created are examples of products whose main value is in automation (see Figure 4-15). Intelligent systems such as the Nest thermostat may promise to do the job (such as setting a heating schedule that best fits home occupancy) better than a human.
Tags or sensors embedded in objects allow them to be trackable and identifiable. The FedEx SenseAware service (Figure 4-16) embeds a multisensor device inside sensitive shipments (such as medical supplies), allowing the sender to track the location of a parcel and the temperature, light levels, humidity, and atmospheric pressure to which it has been exposed. If any of these fall outside a set range, a replacement parcel can be dispatched.
It goes almost without saying that your system needs to be reliable enough to fulfill its promise. Glitches and outages are inevitable in most systems and early adopters will forgive these more readily. But if there are contexts of use in which you cannot afford failure, the product must be 100% reliable. For example, emergency alarms for elderly or vulnerable people must always work. You’ll need a backup power supply and connectivity (see Figure 4-17), and regular checks to ensure these work.
The Product Comes at a Cost Proportional to the Perceived Value
A good product needs to balance the cost and effort required from the user against the value it delivers. If the value is very high, users may be prepared to pay more, or invest more time in configuration.
Determining a price point is a tricky matter in itself. You’ll have to consider manufacturing costs, competition and market conditions, and what users are prepared to pay.
You’ll also need to consider the cost to the user of switching from whatever they were using previously. Household technology, like heating and alarm systems, tends to last years and users won’t want to replace working boilers, sensors, or other kit at great expense without a significant benefit.8 If you can support retrofit—new technology that can easily be integrated into old systems—without greatly increasing the cost of your product, you’ll increase the potential market for the product.
In the context of UX, the perceived cognitive effort to use your product and the time it will take to get it set up and working affect who will buy it, and why. Be careful in your judgment here. In the thick of a project when you are excited about your idea, it’s easy to overestimate how motivated users are to invest time in your product.
Smart homes are a typical example here. It’s been possible to connect up lighting, heating, appliances, and entertainment systems for around 40 years, as we saw earlier. But you needed to be an enthusiast to set it up and program it (or wealthy enough to pay someone else to do that). A niche of users has taken great pride in their automated homes, but others have found them fraught with support issues, technology failures, and a poor fit with the needs of other guests and residents. Mass-market users often view home automation with suspicion: home is a very personal context, and one in which we are often loath to introduce novel technologies that might break our established routines. Most of us don’t want to have to do a load of programming just so we can turn lights on and off. We manage that well enough already and it’s an effort to switch unless the benefits are really evident.
Adding extra cognitive effort to everyday tasks is a common risk. The designer Scott Jenson proposes the idea of the “surprise package”: the mature consumer product that is “enhanced” technologically, turning it back into an early adopter product. As Jenson puts it:
Companies take product concepts that are now far into the laggard range of stability and established behavior, and they change the product significantly. ... The new product is effectively repositioned ‘back to the front’ of the curve, creating a high-tech product that can only be used or appreciated by the forgiving and accomplished early-adopter group of consumers. This is where much of the consumer backlash appears, as safely mature and benign products such as TVs, radios, thermostats, home phones and even cars are turned back into early adopter products, and then sold to an unsuspecting laggard audience.9
TV is a great example. TV used to be an instant-on experience. We may have had less choice of channels and no on-demand services, but you could be watching something within a second or two of turning it on. It can now take minutes. You may be faced with software updates for your set-top box and/or connected TV (perhaps for apps you don’t even want but can’t delete), then minutes of navigating around a program guide or on-demand service using a cheap remote control poorly equipped for the job.
If your product is replacing an existing consumer product or way of getting something done, pay attention to what was good about the old way of doing things. Try to preserve that and enhance the experience, rather than adding new complexity.
The Product is Pleasing to Use
The hard-headed cost/benefit analysis is important for any product, but the best products speak to us on an emotional level, too. This is partly about aesthetics, but it’s not just about bolting pretty design on top of functionality. We form an integrated impression of the functionality and design of the product, and how well that fulfills our practical and emotional needs and fits (and perhaps communicates) our sense of who we are.10 Figuring out the right experience is about design as well as product strategy.
Services in IoT
In this section, we’ll consider the role of services in creating a viable IoT product. Why are services important, how do users think about products that have device and service components, and what kind of services should you consider offering?
IoT Products Combine Devices and Services
At the start of the chapter, we set out that an IoT “product” is frequently a hybrid of physical device(s) and service provision. At the very least, Internet connectivity is a form of service, and there may be other service components that your product needs to work properly and ensure a good experience.
When people buy a product, they expect to have the right to use it for as long as they like. When the product is dependent on an Internet service, there is a reasonable expectation that that service will continue to be available, for taking it away would render the product at worst useless and at best limited. After all, you would not expect your home heating or lighting to cease to function because the company that produced the original system had gone out of business, or no longer wished to support you. The service forms part of that experience. The relationship between the device and service can vary.
There is an inherent tension here between the old world of physical products, and the new world of Internet/web services. On the Web, new services appear and old ones are “sunsetted” on a regular basis. This is acclaimed as progress. But a physical product is likely to come with expectations that it will last for at least a few years. If the service stops working, the lifespan of the device is shortened, creating landfill (and unhappy users). Service providers have a responsibility to ensure that they are able to maintain and improve their Internet services, so that the product has a reasonable lifespan.
How Users Understand Devices and Services
Connected heating controllers and door locks are examples of systems where the device is likely to be the focus: we might call them service-enabled devices. Users view the device as the most salient part of the system. For example, Nest users are likely to say “I have a Nest,” referring to the thermostat to represent the entire system. (It’s far less likely you’d hear someone saying “I use Nest,” “I have Nest,” or “I have a Nest system.”) Because the device is so central to the UX, users will have high expectations of its design and functionality. The service enables remote access and smarter functioning, but in the user’s mind it is a way to control the device (see Figure 4-18).
A security alarm is an example of a system where the service is the focus: we might call it a device-enabled service. The alarm service is what users care about. The sensors and other devices are generally low profile and most of the intelligence sits in the Internet service or gateway software. You could add or swap out devices without affecting the core functions of the service.
Interactions are distributed across multiple devices, so no single device is the center of attention
Most functionality lives in the cloud service or gateway software (perhaps because local devices don’t have much computing power)
Devices can be added, removed, or swapped without changing the core functioning of the system
As the UX expert Mike Kuniavsky describes it, the device is an avatar for the service.11
The Oyster travel card (see Figure 4-19) is a stored value contactless smart card used on London public transport. It can hold various types of tickets or a credit balance for travel on the underground, trains, buses, trams, and boat services. (“Stored value” means that the credit is notionally held on the card itself, rather than in a separate account, as with a debit card.)
Passengers add tickets or money to the card itself via online purchase, ticket machines at stations, or by setting up regular debits from their bank accounts. They swipe the card on a reader at the start and end of journeys to validate their tickets or deduct credit. The Oyster saves time and money processing ticket office transactions and reduces the number of paper tickets. To encourage use, fares are substantially cheaper than paper tickets.
The Oyster card itself is not much of a smart object. It’s just a piece of plastic containing an RFID chip and a small amount of memory. The RFID chip passes a unique ID to a reader when a passenger swipes in. The memory holds information about the tickets or money stored on it, so the reader does not need to contact the back office service in real time every time the user swipes the card. This speeds up the rate at which passengers can pass through ticket barriers, which is vital during rush hour. Readers transmit transactions to the back office in batches.
The Oyster card is an icon of London life, but it is really just an avatar for the service. Without the card readers or the ability to top it up it wouldn’t be much use to you. The Oyster service involves smooth coordination between many different channels, such as the Transport for London website, the ticket machines, ticket offices and shops that sell top ups, the readers themselves, and the back office systems (see “Service Ecosystems”).
Technically, the Oyster card itself is not even an essential part of the service: Oyster can also be used via NFC-enabled phones and bankcards. In the future, the dedicated Oyster card might even disappear, but the service will remain. But services are intangible, and avatars can provide a concrete, tangible focus that helps us understand the service.
Right now, IoT systems are still pretty novel and not well understood, at least by consumers. It’s easy to look to individual devices as a handle to understand the system, whether or not this is accurate. (We’ve heard smart meter users refer to the in-home counter-top display as the “smart meter,” and the actual smart meter as the “computer under the stairs”; see Figure 4-20). You might need to play up the role of the devices in communicating what your system does (presenting it as a service-enabled device), just to help consumers understand it.
Over time, as we all become more accustomed to the products around us having intelligence and connectivity, our ability to understand connected products as services without depending on physical manifestations may become more sophisticated. The idea of a heating system without a visible controller, or a door lock without a visible lock may seem strange right now, but in time, as long as they work, we might be more open to such things.
It will probably always be appropriate for some systems to have highly visible devices, and for some to focus more on service design (see Table 4-1). The key is to pitch your UX to best suit your product, and the needs of your users.
If your service is the focus of the UX, you can still make beautiful devices but make sure the service design is at least as good. And if the device is the hero of your UX, make sure it’s attractive, usable, and does what it needs to do, elegantly.
|Device is the Focus of the UX||Service is the Focus of the UX|
Industrial design of devices is striking
Industrial design of devices is low profile
Many interactions handled via the core device (e.g., a watch)
Interactions are distributed across multiple devices
Most functionality is handled on the core device itself (e.g., a washing machine)
Most functionality lives in the cloud service or gateway software
User views service as a way to control the device (e.g., “I have an app for my thermostat”)
User views device as a means to enable the service (e.g., “The intruder alarm uses motion sensors”)
Swapping the core device for a different make or model would radically change the user experience of the system (e.g., changing to a different brand of thermostat or make of car)
Devices can be added, removed, or swapped without changing the core functioning of the system (e.g., paying for travel with an NFC phone instead of an RFID card)
Services are delivered through the interactions of networks of people, organizations, infrastructure, and physical components. The devices, and even the digital components, are only part of the experience. A part of the Oyster experience is the interactions you may have with station staff when buying a ticket or asking for help. In order to help you, they will have been trained to provide good customer service, but they will also need access to good information about the transactions on your card and system information. Making this whole system work smoothly is a lot more complicated than just making cards, machines, and a website: it requires someone to take a holistic view of how the service is experienced, and make sure all the components work reasonably smoothly together (see Figure 4-21).
Complex IoT services, such as a connected home, require the coordination of multiple devices working together, perhaps even using a degree of intelligence to automate some functions without explicit user instructions (e.g., turning off the electricity supply if a gas leak is detected). There may be multiple digital interfaces. There may even be coordination between multiple digital and physical services—for example, a heating system may use data from a third-party weather service, and the user might have the option of taking out a service and repair contract.
This is called an ecosystem.
If you’re designing a service, you’ll need to take a more complete view of all the parts of your service, and the relationships between them. For example, you may need to think about handling software and component upgrades across your devices. Ovens may require new controller boards, and washing machines may demand firmware upgrades (see Figure 4-22). You’ll need to design processes for handling these issues with consumers.
There may well be customer support, marketing, sales, and perhaps professional installation and maintenance too. Mark Kawano, founder of Storehouse and former Apple User Experience Evangelist, describes how creating a well-designed product is about getting the entire experience right, not just about the UI layer we typically associate with design:
You see companies that have poached Apple designers, and they come up with sexy interfaces or something interesting, but it doesn’t necessarily move the needle for their business or their product. That’s because all the designer did was work on an interface piece, but to have a really well-designed product in the way Steve would say, this “holistic” thing, is everything. It’s not just the interface piece. It’s designing the right business model into it. Designing the right marketing and the copy, and the way to distribute it. All of those pieces are critical.12
Building a Service Offering
Thinking at the service experience level encourages us to take a broader look at user needs, not just the interactions with the website, mobile app, or embedded devices. In this chapter, we’ll consider how this might create new opportunities to look at the wider service package you may offer customers.
For example, take home security. We typically think of alarm systems as something that makes a loud noise and acts as a visible deterrent on the home. A connected system can tell you when someone is breaking in, and perhaps film them. But if you’re not able to get home, and no one else can respond, you can’t do anything about it. Connectivity has alerted you to the problem but also enabled you to feel powerless to act.
In this case, you might have the option of paying for a professional monitoring service. The security firm (with your permission) can view your cameras before sending someone out.
IoT services often provide opportunities to capture data, which can be used to improve the service offered, perhaps through better customer service or smarter support. For example, diagnostic data about the functioning of a boiler could be used to identify systems at risk of breakdown and schedule an engineer visit before they fail. You might even package the cost of the service contract with the monthly fee for maintaining the service. Users may be happier to pay for something like this than to cover the cost of maintaining your Internet servers!
Professional installation or configuration may also be an opportunity for complex systems such as home automation technology. Time It Right,13 designed for the needs of the Orthodox Jewish community, is a culturally specific example (Figure 4-23). There may in the future even be a role for independent IoT “plumbers” who specialize in helping consumers install, maintain, and repair connected home systems: a kind of role the technology blogger Anil Dash refers to as “blue collar coders.”14
Another service opportunity may involve helping users secure lower prices or better service from third parties, or otherwise benefit from the data that comes from your system. In the sustainable housing development of Little Kelham in Sheffield, northern England, residents have smart meters to track energy use and band together to bulk buy electricity.
Personal services, such as installation or intensive customer support, are not necessary for all services and may not be practical within your business model. But it’s worth considering the bigger context of user’s expectations around the service you’re offering, and how their needs will change over time, to make sure your service isn’t missing something they think they need, or to spot opportunities to improve the overall experience.
Establishing the relationship between your system and the surrounding service is, in part, deciding on your business model. Put crudely, this can be summarized in two questions: what will people pay for? And what do you need for production to be sustainable?
What is a Business Model?
A business model is the blueprint for how a business creates value for customers, and makes money. For example, a classic business model is the “bait and hook” one used by printer manufacturers. They charge a relatively low price for the initial hardware but make money on toner cartridges.
The model maps out how the business will make money, either from increasing revenue (selling more) or decreasing costs. Increasing revenue can be approached by:
Generating new business from new customers
Generating more business from existing customers
And even a not-for-profit organization needs a sustainable business model in order to survive.
How Do Business Models Affect UX?
Business models shape the way users perceive the value of the service and the fairness of pricing. This can make the product proposition more or less appealing. Users will approach the product or service with a positive, trusting mindset, or a more skeptical or even negative one. This sets the tone for the rest of their interaction.
For example, the major energy companies in the UK have recently come under pressure for perceived unfair pricing practices. All are rolling out smart energy meters, which generate data that can be used to offer customers tips on saving energy and therefore money. But customers who feel that prices have not been set fairly treat this money saving advice with skepticism. This has a knock-on effect on the perceived UX of the energy saving service. Issues of trust need to be tackled up front in the design, perhaps through presenting pricing in more transparent ways.
Device and Service Models
In the traditional product business model, the provider charges once, up front, for hardware. This is how we are used to buying, for example, cars or household appliances. For a long time, this was also how we bought software.
In service models (of which many digital models are examples), the provider charges for ongoing service provision. Music subscription services such as Spotify or storage like Dropbox are examples of service business models. A nonconnected product can be supplied as a service; for example, renting a car through a traditional rental company is a service. The customer pays for use of a car, not for ownership of a particular car. Adding connectivity to objects increases the potential for service models.
The choice of business model is a balancing act between where the customer perceives the value to be and what they expect to pay for, and what it costs you to provide. IoT is new, which means that what the customer expects to pay for is not always a reflection of the costs you may incur!
For example, your connected heating controller requires an Internet service to provide you with remote control via your smartphone app. That costs money to maintain, especially when you consider that the lifespan of a heating controller might be 10 years or more. The provider might factor that cost into the price they charge you for the product up front, but that might make it more expensive compared to competitor products. Or they might choose to charge you a regular fee for ongoing service provision. But they might find that customers aren’t used to paying an ongoing fee to keep their heating working! In that case, the company might try to add value to this service, by turning it into a platform that also supports other devices (like connected lighting or energy monitoring), or adding extra service components, like a maintenance contract.
As already discussed, the perceived value may rest in different places for different types of connected device/IoT system. Is it the overall service users think is most valuable, or one or more of the devices that deliver it?
IoT is immature and there’s perhaps a tendency for users to focus more on devices, as these are novel (as we suggested earlier). Service providers may focus on more on services, as that’s where the potential for long-term customer engagement sits. You may think that you are offering a health management service, but if your users see beautiful bathroom scales, then persuading them to pay for an ongoing service may be tough. The key for a business model is that customers feel they are paying a fair price (whether in terms of money or sharing their data) for the value they receive, and you are making the money you need.
Bringing Digital Business Models to Physical Products
Building digital services around physical products enables suppliers to apply novel business models, more commonly found in the digital realm, to physical products.
Combining physical devices and services is likely to lead to some interesting, novel, and disruptive business models that challenge our preconceived ideas of what it is to own and use a product.
Business models we are accustomed to in the digital realm might make their way into the physical world. For example, we are accustomed to using websites that are free at point of use but make money from selling eyeballs to advertisers. It’s not a huge stretch to imagine that physical devices might be given away in exchange for advertising or user data. The ubiquitous computing researcher Pertti Huuskonen jokes about the freemium fridge. Your supermarket might give you a free fridge with a screen that forces you to watch an advert every time you open it. Or the fridge might track the products you put in it and eat, and where you bought them, and share that data with the supermarket, and other advertisers. Users could buy their own fridges, essentially paying for ad-free experiences or privacy, or get free or cheaper appliances in exchange for their eyeballs and data. In the future, privacy may be a rich person’s luxury.15
More positively, there are benefits to digital service business models. There are opportunities to develop ongoing relationships with customers, understand more about the people who buy your products and what they do with them, and tailor services better to their needs. Users are accustomed to (and mostly comfortable with) web-based services that capture and store information about them to provide a better service. The sensing and processing capabilities of IoT devices open up potential to extend these personalized services to the physical world.
You may be able to capture user behavior that wasn’t previously visible, perhaps in real time—for example, how people are using energy via smart metering, identifying and tracking people in a physical space, or monitoring traffic levels via aggregated data about the density and speed of movement of drivers’ smartphones. Knowing how often or intensively a product is used, and what for, enables you to tailor the service, sell supplies, improve the next version, or offer additional services. For example, some Nespresso coffee machines now come with a SIM card, allowing Nespresso automatically to send more capsules when the customer is running low (see Figure 4-24).
You could charge users based on their behavior. For example, some car insurance companies, such as Insure the Box in the UK,16 base pricing on actual driver behavior rather than demographics. This may benefit responsible drivers who are in a demographic category considered to be at high risk of accidents, such as under 25s.
Products can proactively maintain themselves or provide data to enable smarter support: as discussed earlier, boilers could identify when they are developing a fault and be serviced before they break down; appliances could share fault data with the manufacturer so customer support can help users diagnose and fix more of their own problems. It is possible to vary pricing based on time of use (e.g., charging more for electricity at certain times of day to manage demand).
A good UX design needs to balance the needs of the business with the needs of the user. Even if you are not shaping the business model of the product you are working on, you at least need to understand it. The best business models serve the interests of customers as well as the business. In the car insurance example, cheaper insurance is granted to drivers who would otherwise be penalized on grounds of age, but the model also allows the company to reduce costs and uncertainty through more accurate risk profiling. User and business needs can be in tension: for example, a connected home service might rely on heavy upselling of new products. Here, design can make the difference between the upsell advertising being either useful or at least minimally intrusive, or downright irritating or something that causes users to stop using the system altogether.
A clear value proposition ensures users understand what your product does, and whether they want it. This is essential in order for them to understand how it works, and how to use it.
Innovators and early adopters are inherently interested in technology and forgiving of imperfections. Mass-market consumers often have different needs.
Many IoT systems are tools: they require the user to frame their own problem and configure the system to solve it. Consumers tend to look for products that promise to solve a particular problem for them and come already configured to do that. They expect the cost and effort of using the product to be in proportion to the value it brings them.
IoT creates new opportunities for information gathering, sharing, remote control, and automation. But there are common pitfalls that can limit a product to early adopter markets. In particular, be careful of introducing new complexity to mature consumer products.
The UX of an IoT product might be focused around the device or the service. All IoT systems depend on some kind of digital service, and perhaps offline service components too, like professional installation, maintenance, or customer support helplines. Ensuring these work well together is an important part of the overall UX.
Business models shape the way users perceive the value of the service and thus the UX. Bringing connectivity and intelligence into devices may lead to digital business models appearing in the physical world.
1 From a talk at UX Brighton, November 2012.
2 Everett M. Rogers, Diffusion of Innovations, Fifth Edition (New York: Simon & Schuster, 2003).
3 Geoffrey Moore, Crossing the Chasm (New York: HarperBusiness, 1991).
4 Steve Blank, The Four Steps to the Epiphany: Successful Strategies for Products that Win (K&S Ranch Press, 2005).
5 Nest Protect has suffered from some interaction design issues related to the ability to disable false alarms. But the Protect is still a good example of a clear product concept.
6 At the time of writing, the Arduino model is being phased out for a newer version based on the Spark Core development board.
7 Matthias Rothensee, “User Acceptance of the Intelligent Fridge: Empirical Results from a Simulation,” IOT’08 Proceedings of the 1st International Conference on the Internet of Things, 123–139.
8 C.F. the model of shearing layers, which describes buildings as a set of components that evolve and obsolesce over different timescales. “Services,” like HVAC and plumbing, are expected to last 7–15 years. This concept originates from architect Frank Duffy and was developed by Stewart Brand in his book How Buildings Learn: What Happens After They’re Built (1994, Viking Press).
9 Scott Jenson,, The Simplicity Shift (Cambridge, UK: Cambridge University Press, 2002). Available from http://www.jensondesign.com/The-Simplicity-Shift.pdf.
10 Lionel Tiger’s The Pursuit of Pleasure is an interesting viewpoint on the anthropology of what makes products appeal (Boston: Little, Brown, 1992).
11 Mike Kuniavsky, Smart Things: Ubiquitous Computing User Experience Design (Burlington, MA: Morgan Kaufmann, 2010).
15 Personal communication with Pertti Huuskonen; Pertti Huuskonen, “Run to the Hills! Ubiquitous Computing Meltdown,” Proceedings of the 2007 Conference on Advances in Ambient Intelligence, 157–172, http://www.cs.swan.ac.uk/~csmax/csrsRG.pdf.