Market research from the Web

Elemental forces shape the world we live in: oceans struggle with continents, wind erodes mountains, and marketing departments conflict with customers. Marketers need to know about customers in order to do their jobs, yet the methods that they use to get this knowledge can interfere with customers' desires to achieve their goals quickly and easily. And businesses fail when customers leave.

If you're thinking about doing market research through your Web site, think carefully. The last thing you want to do is turn customers off by turning market research on. You need to know how to conduct research without making your site user unfriendly.

A real-world experience

Do you ever have days when it seems like everyone wants a piece of you? We recently witnessed the competition for consumer feedback escalate at one of the worst possible times: April 15, the deadline for mailing tax returns in the United States. Those last-minute taxpayers--and anyone else trying to mail a letter that day--had to run a gauntlet to get to the post office. To take advantage of the foot traffic, several eager information-gatherers were strategically located along the path to the post office door. People had to dodge survey gatherers, petition lobbyists, and even radio station employees giving away T-shirts in exchange for information about music preferences. Mailing a letter that day escalated from a minor problem to a major hassle.

Web users experience similar hassles every day. We are constantly bombarded with information-gathering techniques that make completing tasks online more time-consuming and difficult. The problem is that marketers' goals (to obtain information about customers) are often at odds with customers' goals on the Web. A recent study by the Pew Research Center indicated that people typically go online to use e-mail, find product information, and locate info related to their hobbies. In other words, Web users are very goal oriented. They want to accomplish something, and anything that gets in their way is annoying. For companies that want Web users' business, annoying these people is dangerous. We'll begin by describing some common marketing research pitfalls.

Common mistakes made

No reward
Most Web users also don't like to provide personal information unless they get a meaningful reward. It's the age-old "What's in it for me?" (WIIFM) question that must be answered before a user is willing to take time and energy to share what is often considered private. Providing a compelling answer to WIIFM can be tricky. What's meaningful to one audience may not be so to another.

Poor communication of security policy
Maintaining privacy is another concern that many Web users have; they see market research as intrusive. The Personalization Consortium, a group of companies dedicated to the personalization of technology for business and customer relationships, recently reported that 84 percent of users have refused to provide information to a site if they were unsure of how it would be used.

Poor integration with users' task flow
Did your mother ever tell you it's not polite to interrupt? It's amazing how many sites could benefit from more thoughtful market research--that is, market research that doesn't interrupt users' task flow. Many e-commerce sites, for example, still require users to register with the site before they can make a purchase. Why not integrate site registration with the purchase process instead?

Long registration forms and surveys
Most goal-minded Web users simply want to access information in the shortest, most efficient way possible. Impeding their progress is a great way to drive them off. If you want to keep customers, don't ask them to complete long registration forms or surveys. Getting users to answer even one question is a challenge; getting them to answer several questions may be nearly impossible.

Updating sites based on surveys
Usability research studies by renowned experts such as Jakob Nielsen ("Voodoo usability") and Jared Spool ("As the page scrolls") have shown that what users say they do and what they actually do are often quite different. Therefore, updating your site based only on what users verbalize in surveys or questionnaires can be a big step in the wrong direction. Even focus groups, which aren't usually conducted on the Web, typically reveal only users' thoughts and opinions rather than actual behavior. Therefore, it's important to be aware of all the options for gathering user information before settling on a particular method.

Let users opt in

If you want to reduce the chance of annoying customers with information-gathering techniques, allow them to opt in as much as possible. For example, if you must have a registration process, make as many fields optional as possible. Specify which fields are required by bolding the field labels and including a short statement at the top, such as "Required fields are indicated in boldface."

If you're planning to integrate a survey into your Web site, avoid the "pop=up" method. Instead, post teasers within the Web pages so that users can choose to enter the survey if they desire. Offering incentives up front will improve your click-through rate. If you're thinking about sending surveys via e-mail, allow users to opt in by checking a blank box indicating that they are willing to accept e-mail on specific topics.

Offer meaningful incentives

Sometimes the little extras make all the difference. By offering customers incentives to complete a survey, you are basically telling them that you value their time. The incentive need not (and perhaps should not) be extravagant. Incentives could include coupons for products or services, cash rewards, or perhaps a donation to a favorite charity.

Karen Franks, senior director of marketing at MarketTools says, "In a world where consumers are increasingly bombarded by information and solicitations, effective, incentive-based loyalty programs are critical for driving survey response rates." Her employer partners with a private label-loyalty program company to distribute cash rewards to survey respondents for clients using MarketTools' zTelligence product. Incentives that range from $4 to $10--and sometimes as high as $100--are yielding response rates of 30 percent to 50 percent. MarketTools' survey technology is used at Procter & Gamble's Consumer Corner site, where survey-takers earn cash or prizes for completing surveys about P&G products or product ideas.

Keep forms and surveys brief
With your competition only a click away, can you afford to burden customers with long registration forms or surveys? Of course not. Keep registration and survey questions to the bare minimum. For every question you think you might need answers to, ask yourself, "How am I going to use this information?" If you don't know the answer, you probably don't need the question.
Other techniques for successful registration forms and surveys include:

• Ask closed-ended questions that require users to select from predefined answers. Examples include yes/no, true/false, multiple-choice, and sliding scale (1=strongly agree; 5=strongly disagree) questions.
• Write questions that are clear and concise. If you're not a technical writer, find one to help you minimize verbiage.
• Limit registration forms and surveys to one page; users will be more likely to complete the form or survey if they can see the light at the end of the tunnel.
• Put the most important questions first. If users get tired of answering questions, at least you will have captured the most essential information.
• Provide space for optional, additional comments at the bottom. Open-ended remarks by more conscientious respondents can provide valuable information that wasn't captured in previous questions.

Consider the one-question survey One-question, context-sensitive Web surveys generally ask for feedback about a specific page or feature within a site. The advantage of this approach would seem to be that users would find one question less burdensome than a complete set of questions about all aspects of a site.
OpinionLab, a Web survey technology company, offers unique, page-by-page user feedback systems for the Web. OpinionLab clients (which have included Adobe, Netscape, and Dell) get a small [+] icon strategically placed on their screens. When users roll over the icon with their mouse, they can answer a feedback question related to the page being viewed. If the user clicks the icon, they can answer the same question in greater detail.

Participate in discussion groups and e-mail lists A more discreet approach to market research is to participate in discussion groups and e-mail lists that are of interest to your customers. You can discover a great deal of information about your customers by searching discussion group and e-mail list archives or by posting occasional, strategic questions.

1. Begin by identifying your target audience.


2. Identify topics that your audience might be interested in discussing.


3. Search the Web to find related discussion groups and e-mail lists. (Many Web sites claim to be "the source" for discussion group and e-mail list directories. Start with your favorite search engine and search with keywords such as discussion groups, newsgroups, e-mail lists, and listservs to find out what's available.)


4. Read the group or list guidelines to make sure that you don't break any rules. If you think your participation might be offensive, check with the group or list moderator before posting a message. Be extremely careful not to use these resources as advertising venues, as that can get you kicked out.


5. Search the archives of the group or list to see if you can find out what you need through existing information. You may annoy participants if you ask questions that have been asked before.


6. If you can't find answers in the archives, post strategic, relevant questions to the group or list. In some cases, a survey may even be acceptable.

Choose the right research method
If you want to know what customers think about your Web site before it launches, conduct a focus group. A focus group should include six to eight representative customers and a neutral facilitator. The focus group results should tell you if customers understand your business, agree with the value of your products and services, and understand and appreciate your design ideas.
If you want to know what customers think about your Web site after it launches, consider using Web survey technology. If you implement surveys in a nonintrusive way and add meaningful incentives to drive up your return ratio, you should obtain information that's useful for making design decisions.
Keep in mind, however, that focus groups and surveys do not capture customers' actual behavior on Web sites. If you want to know more about what customers actually do on your Web site, you need to conduct iterative usability tests. In his recent book, Don't Make Me Think!: A Common Sense Approach to Web Usability , usability specialist Steve Krug explains that usability tests can be cheap and easy to implement. Krug says all you need are:

• three or four users
• a conference room or office
• a patient human being to conduct the test
• a plan outlining what you're going to test

Krug also suggests a $50 stipend for each user. This may not be necessary if you can recruit representative users from your workplace.
Take advantage of existing market research
Market research can be a time-consuming and expensive endeavor. Why reinvent the wheel if someone else has already done your research? There are numerous free or relatively inexpensive sources of market research on the Web, although locating Asia-specific research may be difficult. A few resources include:

NUA Internet Statistics
eMarketer
AC Nielson eRatings
Test your method
Before launching a Web-based market research campaign, you should test your method for usability. For example, if you decide that you must have site registration or a Web-based survey in order to gather customer information, conduct iterative usability tests to make sure that customers don't encounter any stumbling blocks or perceive your approach as an intrusion. In the iterative review process, users identify problems, and you fix them and retest the site to make sure that the problems have been removed and that the fixes have not introduced any new glitches. Fortunately, you have options for performing marketing research that will keep your site user-friendly. Here are some of the best ways to conduct research without turning off customers.

Take advantage of tacit knowledge on the Web

In their book "Digital Marketing: Global Strategies from the World's Leading Experts", professors Vijay Mahajan and Jerry Wind point out that technology makes tacit knowledge about consumers more accessible. They write, "Some of the most important knowledge about customers is tacit knowledge, such as their shopping patterns, interest and activities profiles, payment transactions, and chat room discussions."

For example, good Web analysis software will tell you what sites customers are coming from and which pages within your site are most popular. If you incorporate search functionality within your site, your analysis software can provide clues about customers' interests by telling you what keywords they use. Here is a list of some free and fee-based analysis tools that can be useful: Free:
Nedstat Basic by Nedstat
http-analyze by Netstore
TulipStats by Tulips and Bears
Fee-based
WebTrends by NetIQ Corporation
AccessWatch 2.0 by Maher Consulting Corporation
Accrue Insight by Accrue Software

Leverage customization

You can learn a great deal about your customers by letting them--or even helping them--customize their Web experience. It's a great win-win situation: customers get more of what they want from your Web site, and you get to find out their major interests. Two examples of customizable sites are Excite and Yahoo. Both sites allow users to indicate their content preferences for a customized view of the home page.

The Personalization Consortium also recently reported that 63 percent of U.S. consumers are more likely to register at a Web site that accommodates content customization and offers personalization features. The study also found that 53 percent of consumers are more likely to purchase goods from a Web site with personalization features.

So not only will customization provide you with information about your customers, it may get you more customers!