2: Nine Things a Marketer Needs to Know About User-Experience Strategy
This chapter intends to explain some of the key concepts related to successful user experience, including a brief overview of topics that will be discussed in detail in later chapters.
The author recommends including UX strategy in the business case for the project, which ensures that it is prominently documented and that all stakeholders "sign off" on the UX goals, so they are given appropriate weight (rather than being forgotten or subordinated to other goals).
The UX strategy should declare that UX is critical to building relationships that maximize long-term success; focus on customer retention; identify usability as a key factor in successful engagement; consider the opportunity cost of "bad" UX; and assigns ownership of UX to marketing.
(EN: Some of this seems arbitrary and self-evident, but I'd agree that documenting UX will help to prioritize it, and ensuring it's integrated into the plan and development process is necessary, as it's often overlooked.)
1. Capture What Customers Expect in the User Experience
The expectations of the user are the first concern of UX strategy: if the site is not accepted by users, it will fail regardless of how well it is suited to the needs of the business.
In short, any Web project is created to enable people to perform a task, as well and efficiently as possible. Whereas the business may want a catalog that will "make users aware of the full range of products we offer," the user wants "to find the product I need quickly and easily."
Gathering this information means market research: it cannot be discovered by analyzing the business goals, nor can it be accurately identified by asking the business what it supposes customers expect - you must ask the customers directly.
It's also noted that customers are heterogeneous: not all of them have the same expectations, so you must consider all input to determine common expectations. Also, not all expectations are to be given equal weight: some are critical, others are "nice to have." These distinctions are important to make in order to come up with a coherent plan that negotiates conflicting priorities.
Leveraging pre-existing models from other channels (phone, brick-and-mortar) can be useful for envisioning the process. Users are accustomed to certain processes, and will make the transition to the online medium more easily if these processes model previous experience.
(EN: this was once considered good advice, but is a bit outdated. Not only do users have more experience in the online channel - for anyone under thirty, buying a book in a store is less familiar that using amazon.com - but it also neglects some of the capabilities of the online medium that are not present in other channels, and not all practices transfer elegantly from one channel to the next. More current advice is to "design for the channel.")
Once customer expectations are gathered, the UX representative must ensure that they are included and considered in the planning process, and should steer conversations toward considering the task that the user must perform, especially when the discussion becomes abstract and system-focused.
2. Link Usability and Profitability
Users turn to businesses in order to provide a solution to a problem, and seek to do business with a provider who is easy to work with. This is true of any medium, but especially so of the Web, where an alternative provider is just a few clicks away.
It's also noted that there is a "mismatch" between the processes that are easy for the business versus those that are easy for the customer. (EN: If a design is difficult for the business, it will negatively impact profit by increasing the cost per transaction. If a design is difficult for the user, it will negatively impact profit by decreasing the number of transactions.)
For e-commerce sites, "success" can be measured by the number of sales generated, but it's not the only measure. An example given is a brokerage: revenue is made on trading commissions, but there are a number of ancillary elements (investment guidance, quote generation, billing statements, and the like) that will contribute to the overall experience - if you do these poorly, getting the trading transaction "right" doesn't matter.
The author suggests that "new metrics will be needed" and names a couple of commercial products that propose to do this, and a few anecdotes about the way companies have put a dollar-value on something such as eliminating an error message.
3. Architect for Navigability and Scalability
The visual impact of the site, while important, has far less impact on UX than the way the user interacts with the site - specifically, how they find what they need.
As such, two critical components are navigability (the ability to "click" from the home page or other entry point to the destination within the site) and scalability (the development of an information architecture that can accommodate features that are added in future without having to rearrange the entire site).
(EN: The author pays a bit too much attention to scalability. My sense is it was a hot topic at one time, as many business would start with a small budget and then bolt on additional features over time, eventually becoming a labyrinth. This is no longer a significant problem: most companies have mature sites. Moreover, the advice about stubbing out sections with inactive links and coming soon pages has been found to be a bad practice.)
4. Make the "First Use" Frictionless
The author provides advice for considering the experience of the new visitor - one who has never visited the Web site before and does not have experience to guide them. The chief problem is that the company knows itself, and develops a site based on an intimate knowledge of the organization and its functions. The first-time user knows none of this - so when the site is designed, you must "think like a user" rather than an employee or subject-matter expert.
(EN: I'm skipping the rest of the advice here, as it's really not that good. It's more along the lines of imagining yourself as a user, trying to guess what the user knows. This is a very bad practice because people tend to rely on their assumptions, and often get it very wrong - either imagining the user to be much dumber or smarter than they actually are. The right practice is not to guess, but to involve users - conduct focus groups beforehand, do usability testing on prototypes. Anything else is sloppy and inaccurate guesswork that can do more harm than good.)
5. Thread Ease of Use Throughout the Experience
Ease of use is a core concept of user experience. At any point where a user does not know how to proceed in order to accomplish his goal, some percentage of users will quit. The author compares this to the "suspension of disbelief" in entertainment, a fragile state of mind that, once broken, is difficult to regain.
6. Execution: To Outsource or Not
The author considers the prospect of outsourcing: for cost and time efficiency, it seems like a good solution to outsource user experience. But given that user experience is critical to the success of your site, it is unwise to do so.
The author cites the example of the 1999 Christmas shopping season, in which many e-commerce retailers suffered the consequences of disastrous consequences. What followed was chaos: design shops were overwhelmed with demand and elevated prices, hired unqualified staff to cope with the workload, whose work did not justify the prices charged for it. And ultimately, many firms fell into the model of recursive engagement - hiring a revolving door of consultants, each to fix the problems created by the last.
The "takeaway" is that a company should "own" its customer experience and maintain an internal competency. The labor to develop the site can be outsourced to heads-down workers, but the firm should seek to control the UX and maintain in-house expertise to direct the work.
Organizationally, UX must be accepted as a core competence, rather than an afterthought. The user experience impacts the customer, and is a significant factor in the present and future revenues of the enterprise.
7. Fixing Broken user experiences
Especially since switching cost is low on the online medium, customers will leave if the UX is confusing or difficult. When this happens, the customer relationship is damaged, revenues suffer, and the organization suffers.
At yet, there is some argument over what constitutes a "broken" user experience. The conversation becomes more objective when you can demonstrate the problem using server log analysis to find where customers are dropping off. (EN: additional metrics can refine this, to find problems on the field-level, and to measure not only whether the user was ultimately able to proceed, but where there were delays.)
With this data, user experience can be triaged:
- A "fatal" problem prevents users from completing a task
- A "critical" problem exists when users experience difficulty, but can slog through
- A "minor" problem exists where there are minor cosmetic defects
The problems with a site should be listed and prioritized - not only in terms of the cost and time to develop a solution, but also in terms of the loss to the organization that will perpetuate until the problem is corrected. Use an ROI approach, and prioritizing problems according to which will have the highest return on the cost of development, suggesting that 80% of loss can be eliminated by fixing 20% of the problems. (EN: Pareto is trotted out when people lack research or experience - my sense is it's not so simple, and that many of the larger problems also require larger investments to correct.)
Also, you should track "lessons learned" from UX problems to prevent repeating the same mistakes, and avoiding them when the next major development effort is undertaken.
8. Channel Integration
For many firms, their foray into the Internet began as a side effort that was not managed in a coordinated fashion with its existing customer communications. As such, there were glaring inconsistencies between information communicated to the customer by the Web site that confuse and frustrate customers (e.g., an online price that is not honored at a store).
There is also the problem of integrating operations across channels: a person who places an order online is frustrated when they place a phone call and the operator can provide no assistance (or vice-versa).
The company's goal should be "harmonization" across all channels to provide a consistent user experience - primarily in functional ways, but also in non-functional ones (e.g., a printed catalog whose design, writing tone, and color scheme do not match the Web site creates a sense of confusion, even if the functional capabilities are equivalent).
(EN: this problem pre-dates the Internet, especially for companies that do both in-store and catalog sales, but customers were less likely to notice or pay attention. But with the Internet, and especially mobile computing, attention has been called to the problem, and a customer is far more likely to "look up" a price on their cell phone while standing in a store.)
9. Invisible Ubiquity and Transactional Intelligence
In traditional-commerce, customers were far more likely to be segmented and homogeneous: by virtue of physical location, collective attraction, and other factors, a store would attract a very specific market segment (in terms of demographics). We sites are more ubiquitous and attract a larger and more heterogeneous audience.
It's also noted that users' experience with the online medium is increasing and they are expecting a higher degree of sophistication in transactional experiences. There is greater likelihood of their drawing parallels, and expecting you to be able to provide the same capabilities as the most advanced of your competitors.
Specifically, users are expecting "transactional intelligence." They expect your Web site, and in fact your entire company, to be knowledgeable about their past interactions with you, and act accordingly to reduce the amount of interaction necessary in subsequent transactions. If you fail to do so, you are no better than an unknown merchant who may offer a better price or slightly different products.