These days, successful businesses are the ones who can break down silos - the barriers that form over time which prevent the flow of communication across departments, stand in the way of sales and even prevent people from navigating easily across a website.
In fact, according to the U.S. Department of Commerce, ecommerce businesses are losing about $4.8 billion per year because consumers can't get the answers they need when and where they need them.
The truth of the matter is that the average website visitor is confused, frustrated or uncertain over a dozen times per day while online. He or she buys products, subscribes to services, pays bills, ingests information and connects with friends despite regularly encountering cryptic error messages, unexpected situations, indecipherable instructions, and "where'd-that-go?" and "how do I..." circumstances. Pay attention to your emotions when online and you will be surprised at how often you encounter speed bumps that force you to change your behavior, taking you off your intended path.
Because it is so difficult to know how users will think and act, websites need some form of customer support. Unfortunately, traditional customer support is poorly suited to handling the millions of specific little questions that arise during website usage every day. Traditional customer support was designed around the call center, a heavyweight behemoth few people-especially younger people-are willing to endure. Modern versions of assisted service such as email support and live chat are just extensions of the same idea. All forms of assisted service are costly and time consuming, and often lead to more frustration than they resolve. By the time most website visitors get assistance, they have already been long on the road of frustration and even anger.
Enter website self-service. Its importance is clear. If your website visitors can help themselves, they can find their own answers more quickly, easily and (importantly for you) cheaply than if you have to assist them. Studies show that 75 percent of all website visitors facing a problem want to self-serve. People like to feel competent and in control. Web self-service, if successful, enables people to feel that way.
When people help themselves, they feel empowered and self-sufficient. We all know the adage about teaching a man to fish... Despite the enticing benefits of website self-service, studies show that 61 percent of self-service attempts fail. In other words, more often than not, website visitors who try to help themselves come up empty. That is an abysmal record that must not be tolerated.
Run-time content, is content that arises while your website is being used by actual visitors. User-generated content is a major form of run-time content, but so are click patterns or anything else that arises during use. Incorporating run-time content into your self-service solution allows you to reflect actual-not anticipated-use and avoid the perils of your expert blind spot. For example, the search results you serve in a knowledgebase should adapt based on where users arrive from and depart to. If your site uses logins, personalized content should be possible based on customers' usage patterns. Utilizing run-time content is an example of building the door, not the path, and then seeing where the grass gets worn. Any FAQ your site employs should be built after seeing your visitors' most commonly asked questions.
Understanding how your website self-service is used by your visitors is vital. I find it useful to think in terms of "helpfulness audits." If an auditor were examining your self-service solution, what questions would he or she ask? Taking a critical look at your self-service solution to understand exactly how helpful it is (or is not) is an important practice. Unfortunately, most self-service solutions cannot fully support helpfulness audits.
At its most basic, a helpfulness audit must be able to identify what was helpful, what was not, how often and for whom. Such questions are easier to answer for assisted solutions than for self-service solutions because the latter lack direct involvement with the user. For example, in a searchable knowledgebase, which search results were used and which were not? Which knowledgebase articles were read to the end and which were not read past the first paragraph? For a FAQ or forum, which questions were viewed and which answers solved visitors' problems?
The psychology of tool use is the same with modern websites as it was with ancient stone: Users are focused on achieving their goals, not the tools themselves, and successful tools must therefore remain invisible. The moment a tool leaps to the fore and calls attention to itself, it has failed, consuming users' attention like a black hole consumes light. Users forced to deal with their tools at the expense of their goals will soon move to other tools, meaning they will find other websites that make it easier to accomplish their goals. Not surprisingly, studies by Forrester show that 57 percent of website visitors abandon websites the moment answers are not available. If it is more than five clicks or taps, or requires substantial text entry, only the most motivated website visitors will bother.
Similar to keeping your visitors close to their critical path, you must also reduce the amount of work your visitors must do to get the help they need. Requiring lots of explicit input from your visitors before they can get help makes an already frustrating experience even more so. Examples are the many form fields that must be filled prior to submitting an email from a corporate website or using a virtual agent. On mobile platforms especially, the requirement for explicit input must be kept to a bare minimum.
Instead, use implicit input to inform your self-service solution as to your visitors' needs. Implicit input is the input your visitors are already providing in the course of using your website. Implicit input includes what your visitors click on, where they go, what they view, and whatever text they have typed. Because implicit input is given prior to the self-service experience, it relieves the self-service solution from having to add additional frustration to an already frustrating experience.
Successful website self-service must recognize that on occasion, higher-touch assisted-service solutions will be necessary. The most successful self-service solutions will answer most questions for most visitors most of the time. But the "long tail" questions usually require a live person. The key is for the self-service solution to know when and how to escalate an extremely frustrated user from self-service to assisted-service. For example, searchable knowledgebases may track a user's session so as to offer live chat, an email form or a phone number after repeated failed searches. A virtual agent must know how to gracefully hand off to a live agent without interruption or delay. A FAQ must know whether departing visitors get what they need, and if not, give them opportunities for assistance.
My research over the last decade into the psychology and technology of websites has led to these five "must haves" for successful and satisfying website self-service. A surprising result is that many of the features on which technology companies compete actually contribute to broken user experiences from users' points-of-view. For example, proactive interruption by live chat is a touted technology feature that increases usage but annoys 99 visitors for every one it engages. Instead, self-service solutions that leave users feeling in control are more successful than advanced solutions only the most needy users tolerate.