FAQs on the Web

A Study of "Frequently Asked Question" Lists as a Method of User Support

By Craig Noeldner

Frequently asked questions, or FAQs (see Figure 1), are used to provide support and to explain many kinds of topics—both technical and non-technical. FAQs have been used for many years in print publications as a method for answering questions and solving problems. Now that web sites are developing more functionality and turning into web applications, there is an increased need to provide support to users. FAQs are one of the ways web sites can provide this support.

Figure 1: A section of a FAQ from Schwab.com.

Little data has been published regarding the impact of FAQs in web design and their effectiveness in providing user assistance. Some web sites use FAQs along with other forms of user assistance; others sites use FAQs as the only form of support. To try and better understand online FAQs, I conducted a study of FAQs and their implementation. The focus of the study was to answer the following questions:

• How prevalent are FAQs on the Web?
• How are FAQs implemented?
• Are there any common techniques used in FAQ design?

This article describes the methodology used in the study and provides a discussion of the results.

Methodology

Selecting Web Sites

With millions of web sites available to study, the first issue was what sites to review and how many to select. The focus was to be on sites that provide services or sell products. The sites were also to be of interest to mainstream web users. The selection of these web sites needed to be done in a manner that eliminated personal bias. PC Magazine provided a solution with its top 100 web sites. The PC Magazine 100 comprises 20 different categories, including family, shopping, and web development sites. Each site on the list was reviewed for this study. They are listed in the Appendix.

Defining a FAQ

For the purpose of this study, a "FAQ" is defined as a section of a web site that attempts to support users by delivering solutions using questions and answers. The questions represent typical questions asked by actual customers as well as questions that web designers believe their users might formulate. Despite a few differences in implementation, there are two key characteristics of a FAQ. First, a FAQ is in a question and answer form and second, a FAQ is labeled using the term "FAQ" or similar term. It is easy to identify which sites use FAQs based on these criteria.

For example, a section in Britannica.com labeled "Comments & Questions" is considered a FAQ because the primary content is in question and answer form.

Chumbo.com has a section on its web site labeled "Help/FAQs" that is not considered a FAQ because there is no specific FAQ section (see Figure 2).

Figure 2: Chumbo.com did not have a FAQ as defined here.

There are also several sites where the border between a FAQ and Web site Help is blurred and other sites that include both FAQs and a Help section. Some sites use only a FAQ as support and have many topics. AnyDay.com has well over 200 FAQs.

See the Appendix for a list of the sites studied, data gathered on the sites, and comments relating to borderline cases.

Results and Discussion

How prevalent are FAQs on the Web? A large number of the sites studied, 65% as shown in Figure 3, use a FAQ section. This indicates a significant presence of FAQs in web sites. It is not clear if the presence of a FAQ is related to a certain type of web site. There is no distinguishable relationship between web sites that use FAQs and those that did not, so the use of FAQs is not related to a specific web site category. Therefore, it is important to focus on how and why FAQs are being used in order to determine their importance on a web site.

Figure 3: Presence of FAQs in the studied web sites.

The significant presence of FAQs on the Web raises an important question: "What are the reasons behind a site's decision to use a FAQ?" Although only a survey of the decision makers for the web sites can properly answer this question, I have a few hypotheses. First, FAQs are considered easily identified due to the history and offline presence of FAQs, the existence of FAQs online, or both. Essentially, the term "FAQ" has good branding. But the branding does not guarantee that FAQs are the most efficient and best way to deliver support. Second, the FAQ format is the most natural way to deliver content such as answers to customer service questions. In the cases where both Help and FAQs are used, the FAQs aid in highlighting crucial questions.

Although the purpose of the study is not to include persuasive FAQs, marketing related questions are present in many of the support FAQs. A good example is the question "Why should I choose Schwab?" in Figure 1. It is difficult to believe that Charles Schwab gets a large volume of users asking this question to justify its use as a support FAQ. User assistance is generally more informational-based and not persuasive.

How Are FAQs Implemented?

Although a majority of the sites studied use FAQs, significant difference exists among sites as to how the FAQs are presented. One example is the FAQ on Hifi.com. The FAQs on most of the sites studied directly answered a question (as seen in Figure 1). The questions on Hifi.com link to pages with separate headings and include content that is unrelated to the question. For example, one question asked, "What if I have a question about ordering?" That links to a separate page with the heading "Forms of Payment," and subheadings "Confirming Price" and "When Will My Credit Card Be Changed?" Hifi.com does not directly answer the original question. Instead, it requires the user to do more work.

Hifi.com is only one example. Many of the sites studied vary in other ways. This variance suggests that users may not be able to use what they have learned from working with one FAQ and apply it to another.

FAQs are a very helpful way to deliver specific information to users. "Frequently Asked Questions" implies the presence of useful information. It also can imply that the section labeled FAQs is not exhaustive. Perhaps more infrequently asked questions exist and are not included.

Most of the web sites studied did not approach FAQs as a collection of only the top most frequently asked questions. A handful of these web sites have over 100 questions. This indicates that there are no current conventions that limit the number of questions in a FAQ, and therefore no precise definition of what "frequently" means in a FAQ.

Without a standard for FAQs, focus on site-specific implementation becomes more important. Can a user efficiently navigate through a FAQ to find an answer? If a user looks for an answer in a FAQ, the user must first locate the correct question. There are several prominent techniques used to facilitate the scanning and navigation of a FAQ section.

Are There Any Common Techniques Used in FAQ Design?

In most support situations, a user has a specific problem to solve. The mental process used to solve the problem often goes something like this:

1. A user scans the questions in the FAQ, not the answers, for words relating to the specific problem.
2. If the user finds a word relating to the problem, then the user spends more time analyzing the question.
3. If the question seems related to the problem, then the user reads the answer. If not, then the user continues to scan the questions.
4. If the answer did not solve the problem, then the user continues to scan the questions.

Because most FAQs are question-based, they are usually written using the Kipling Needs Analysis (who, what, when, where, why, and how). If a user is searching for a "who" and the question is formatted as a "why," the user may have difficulty locating the correct question. A user may also scan a sentence looking for certain keywords, which may be synonyms of words in a question.

The highly iterative process of finding the appropriate FAQ to solve a specific problem makes finding an efficient way to deliver FAQs very desirable. The web sites studied use a number of techniques to attempt to reduce the amount of scanning required for a user to find the solution. As shown in Figure 4, the following techniques were found frequently:

• Separating FAQ topics into separate sections.
• Using multiple pages rather than a single page to deliver FAQs.
• Listing the questions first with links to the complete FAQ.
• Using a "Back to the Top" link for smoother navigation within a page.

Figure 4: Four popular techniques used to display FAQs.

Separating FAQ Topics into Separate Sections

Many of the web sites in the study have over 100 questions and answers. Most of the sites have a dozen or more questions. Of the web sites with a large number of questions, many separate the questions into sections. Dividing the questions into sections saves users the task of scanning every question and hopefully reduces the amount of time it takes to find the appropriate FAQ. Of the sites studied, 42 (or 65% of the sites using a FAQ) divide the FAQs into sections.

Although NexTag.com divides its FAQ into sections, the sections are not obvious to the user. The sections are listed on one page and there is no navigation between sections. A user might not scroll through a page to look for additional sections.

Other web sites using a single page for a FAQ section include section navigation. eCircles.com divides its FAQ into a number of sections (see Figure 5). The sections link to different areas on the page.

Figure 5: eCircles.com's FAQs are divided into sections.

Many of the section titles do not divide FAQs into unique sections to help a user search for a solution. AltaVista's FAQ has five sections: (in order) Searching, My Site, Technical, Miscellaneous, and Business Related. The question, "Does your engine detect and penalize for word stacking, spoofing, multiple page submissions, and other kinds of spamming?" is found under the My Site section. The question could also fall under Searching, Technical, or Miscellaneous, but does not. One solution to the problem is to create unique and descriptive sections (such as a title "Miscellaneous").

eBay's FAQ is divided into more intuitive sections for the user: Registration, Bidding, Selling, and others. These sections are related to a user's task and can increase a user's ability to successfully navigate a FAQ. In some cases, a user may expect to find a specific FAQ under different sections, so another way to solve the problem is to include a question under multiple, relevant sections. This method should be used cautiously.

As sections create an additional level of navigation, it is important for the navigation to be apparent and intuitive. The FAQs studied use simple and often one-word terms to separate FAQ topics, and most sections are only one-level deep. AnyDay.com, with over 200 questions, divides its FAQs into only one level of sections. The limited hierarchy depth may be due to the fact that the questions cannot be as easily arranged in a hierarchy. FAQs cannot be easily organized from general to specific information. "How can I use the feature?" is not an intuitive subset of "How can I use the product?" The question format of FAQs restricts the flexibility of layering the information.

Using Multiple Pages Rather Than a Single Page to Deliver FAQs

Another technique that web sites use when dividing FAQs into sections is to use multiple pages. Multiple pages are easily adapted to existing site navigation and can be a helpful way to organize large amounts of information. Of the sites studied, 41 (or 63% of sites using a FAQ) chose to use only one page for the FAQ.

A disadvantage to using only one page is that the page can become extremely long, forcing the user to scroll extensively. And if FAQ questions are not in plain view (when a user loads a page), a user may not realize that the answer exists or may not want to take the time to scroll. A technique to solve this problem is listing the questions first with links to the coupled questions and answers.

Multiple pages allow flexibility in implementing a large FAQ, making additional features easier to implement. Web server logs can provide detailed information on how multiple pages in a FAQ are being accessed. Adding a feature such as a FAQ search, like AnyDay.com, is useful if FAQs are on separate pages.

However, large FAQs disregard the notion of frequency in frequently asked questions. FAQs that contain a large number of topics are more like a collection of all the questions asked, and not the most frequently asked questions. Amazon.com does not have this problem and uses only five top questions with web site Help topics as the basis of its online support.

Why do so many FAQs use only one page? From a web development perspective, one page is much easier to update than multiple pages, especially if the updates involve changes to navigation elements. Many FAQs could truly live up to the term and dynamically update FAQ content based on the number of questions for a topic. Perhaps the print-based history of FAQs makes it more difficult to adapt FAQs to the Web's hypertext ability.

Listing the Questions First with Links to the Complete FAQ

As with FAQ sections, proper implementation of FAQs on each page can have a positive effect on the user's success in finding the answer to a question. The sites studied tended to use two techniques to deliver FAQs on the page level. One technique groups all the questions together in a block and links the questions to the associated answer below (see Figure 6 for an example). Of the sites studied, 53 of them (or 82% of the sites using a FAQ) group questions at the top of the page. A second technique uses "Back to the Top" links for smoother navigation, discussed in the next section.

Figure 6: The Monster.com FAQ lists all questions first.

Using lists is helpful in a similar way to listing sections. This technique allows a user to scan all of the questions at once without scrolling. Listing all of the questions also displays the number of topics on the page, giving a user a sense of page size and cueing a user to scroll to the appropriate question. A user might not know that the appropriate question is further down in a page, which makes cues to scroll down a page helpful. A disadvantage to listing the questions first is that if an answer does not solve the user's problem the user must scroll to the top of the page and scan the questions again. (A "Back to Top" solution is described in the next section.)

Even if questions are listed first, it can be difficult to locate keywords in FAQ questions. For example, Monster.com lists 16 questions in one of its FAQ sections (see Figure 6). One of the difficulties in finding a keyword is that all of the words are the same color, font, and size. The lack of distinguishing characteristics between words and questions makes it more difficult to help a user scan the list. Also, a common characteristic of links is that they are underlined, which makes reading the questions harder.

If the success of scanning a FAQ relies on a user's ability to find keywords in a question, then listing the questions creates a single block of text to scan. There should be focus on the selection of keywords and the presentation of the keywords. Could keywords that are visually separated from the rest of the question decrease the time it takes to find the right answer?

A technique like bolding keywords could enhance their detection (see Figure 7 for an example), but can also make it harder if keywords were incorrectly chosen and/or labeled. Of course, too much bolding, like any heavy use of specific styles, will overload a user's ability to scan the keywords.

Figure 7: A sample FAQ before and after bolding keywords.

Another possible route is to use an indexing technique, such as those found traditionally in the back of books or in a concordance.

Using "Back to the Top" Links for Smoother Navigation

A popular technique used to minimize scrolling is to use a "Back to the Top" link (see Figure 1 for an example). A "Back to the Top" link takes a user from the specific location in the middle of the FAQ back to the top of the page. These links are only used on pages that list the questions first. This technique is used primarily after each answer, although many sites use a link after each section or group of questions. Of the sites studied, 37 (or 68% of the sites FAQs using links to separate questions from answers) use a link that would take the user back to the top of the page. The number of questions on a page could have influenced the decision to use this technique.

Conclusions

The primary finding of this study is that about two-thirds (65%) of the sites examined use some form of FAQ. The study results also show a significant lack of consistency between FAQ presentations: labels are different, navigation is different, and answers are presented differently.

Jakob's Law of the Web User Experience states that users will spend most of their time on other sites. This means that users will have the knowledge and experience from navigating other sites as they navigate through yours. FAQ designers can take advantage of both the user's experience with other web sites and techniques to simplify the search for an answer by following a number of guidelines:

• Separate long lists of FAQs into sections. Make the sections unique to aid navigation.
• Use multiple pages to further separate long lists of FAQs and decrease the amount of scrolling required.
• List questions at the top of the page and have the question text link to the complete question and answer. Use a "Back to the Top" link so users can easily go back to the questions at the top of the page.
• Create questions that can be quickly scanned by a user. Focus on the presentation and choice of keywords to allow efficient scanning.

The goal of any product support should be to deliver the best, most efficient way to solve a user's problem. The study results show a very common way to do this is using a FAQ. Before deciding to use a FAQ, weigh the advantages and disadvantages of using FAQs as support. This becomes even more important if it is the only form of support being used. Is a FAQ going to be the most effective method? Would a non-question-oriented Web site Help work better? Could both methods be used together? Answers to these questions and using proper design techniques can turn a frustrated user into a happy one.

Craig Noeldner currently works as the Senior Technical Writer and Manager of Communications at iSolute.com, a 100-person software company based in Bellevue, Washington. Prior to iSolute.com, he has held multiple Web development positions as a graphic designer, editor, and software engineer. His experience includes work for USWeb/CKS (now MarchFirst), eHome, and Premera Blue Cross. Craig holds a Bachelor’s of Science degree in Technical Communications from the University of Washington. He can be reached by e-mail at craig@noeldner.net.