How do you engineer a major overhaul to your web site to increase traffic and user satisfaction? Ask.com’s Senior User Experience Analyst describes the company’s careful yet bold approach to its recent redesign in this wide-ranging interview.
Editor’s note: Before Ask.com retired its iconic Jeeves character, the company did extensive research and testing on the potential effects of the brand change. In this interview with guest writer Rae Hoffman, Ask.com’s Senior User Experience Analyst Michael Ferguson offered extensive insights into the company’s approach to user research, usability testing and brand impact, offering lessons applicable to any brand change or site redesign that all webmasters,marketers and site owners can learn from.
Rae Hoffman: I heard you speak at the Toronto Search Engine Strategies conference and thought you had a great presentation on user interaction and usability—truly one of the best presentations on any topic I had seen in a long time, especially by a search engine representative.
Michael Ferguson: Thanks, that’s great to hear. I think we’ll be hearing more and more about user experience and brand in relation to search as the sites keep innovating and differentiating. There’s a lot of nuance in understanding search and the more we share, the better for the engines, users, and search marketers.
You made the decision last year to change your brand from Ask Jeeves to Ask and retire the butler. What were the thoughts in regards to why you needed to do it and how it would affect your users?
The butler part of the brand was strongly associated with “the web site where you can search by asking questions.” The range of products and experiences we now provide is far beyond that and will keep expanding–we needed to free up the “brandwidth” to allow current and new users to see what we’ve become and consider us as we evolve. We were a search engine, but we had a question-and-answer engine’s brand.
For a few years we tried to get people to consider Ask Jeeves an “everyday search engine.” Moving that perception as quickly as we wanted was proving difficult, so we bore down on why: the character and the user’s mental model of the site were tightly bound together.
Originally we marketed Ask Jeeves as easier-to-use because you could ask questions and get to information faster. At the time (1997-99), the major competition’s technology was relatively basic, looking at the text and structure of pages for the most part. When you’d type in “cheap Hawaii airfare” you might get pages about ‘cheap airfare’ but not ‘Hawaii,’ for example. Users of the other engines complained about relevance, and some tried to pick up basic Boolean tips—but it all seemed like too much work. In addition, you wouldn’t have to sort through page after page of spotty results—we (with human editors) built a database of reviewed sites that directly answered questions on your topic. The sites we chose were well-written, from authoritative resources, and easy-to-navigate.
The marketing worked well: we were differentiated, appealing, and the user experience was mostly satisfying. We knew the scalability of the large database of questions and answers we were maintaining was limited, so we offered metasearch at the bottom of the page. The matches from the Top 10 Yahoo, Excite, and AltaVista results appeared in dropdown lists, with a link to those engines’ full results pages.
Ask Jeeves went public and responded to the “make the quarter” market pressure by moving too many ads and ad formats to the page, while relaxing the focus on user experience. Others did this too, of course, and Google provided a tasty antidote. In addition to being ad-less, Google’s relevance for keyword searches was better than what users had experienced before.
We stuck around through the bubble burst because of the brand, its butler character, and the fixed Q&A role. Users returned to us when a search seemed easier to phrase as a question, or when they wanted a second, different set of results.
We had this differentiated site with a frequency problem—and for some an image problem. The user need served by phrasing searches as questions is only about 10% of overall search activity. So users kept coming, just in really wide arcs
And “just ask a question” set expectations too high—the technology couldn’t deliver consistently. We’d get questions like: “I grew up in Manhattan in the Fifties. There was a young nurse working at St. Mary’s named Beverly. How can I get in touch with her?” We promptly showed a banner ad for New York City hotels and some unsatisfying algorithmic matches.
We had to make a clean break.
Our marketing team dug deep on a host of ideas about where to go with it. Rather than tossing all of the equity we’d built by introducing an entirely new brand, we found a workable solution: retire the butler to break with the past, but keep the easy-to-remember three-letter word, logo, and URL Ask.com.
How did you prepare for the brand change, but more specifically the affect that a total redesign would have on your users? What did you find in regards to the effects a brand change would have on Ask users?
We use a wide variety of methodologies: ethnographic studies, eyetracking, interactive evaluations, focus groups, and usability sessions. It’s important to vary these and try new approaches, to get at different facets of user experience.
For the redesign, we talked with thousands of people in the US and UK, and iterated prototypes weekly over several months. We listened to everyone from current daily users to those who haven’t been by in 7 years to people new to the brand and site.
While we did find some current users that were resistant to change, most were excited to see the brand and site in a new light. From their perspective, we were addressing a whole new range of search needs they didn’t know we could. They saw it as a vast expansion of what we could do for them.
Did you learn any information on what some potential effects a brand change and site redesign have in general with any web property from your research?
You have to get close to your users and see the site experience from their point of view — and then the transition has to make sense, there needs to be a story that resonates.
Most current users picked up on the story: this isn’t just a restyling to “freshen up” a brand and a site — this is a declaration that we are changing to expand the range and deepen the quality of what we can do for you.
That said, there are always going to be customers that leave the new brand and site—you have to understand their value to your business, and the trade-offs of some short term loss for long-term gain.
How did you handle the day the new design was released? How did you let your users know about the change, how did you let the public know and for how long did you leave any notices up on the site in regards to the change (if at all)?
We knew there would be some users that would be disappointed by us dropping the butler, and others that would support the move, and then others who wouldn’t notice or care. So the key was to give the ones who would care a heads-up, an explanation of why he was leaving, and a way to say goodbye.
About a month before we relaunched, we linked from the home page to a site that explained why Jeeves was retiring and gave users opportunities to sign a farewell card, and vote for where Jeeves would go (my idea of having him retire to join a circus and perform an act called “Jumping the Shark” did not win, by the way). Thousands of users signed farewell cards and participated in the formal farewell to Jeeves. As the day approached, the Jeeves illustration on the home page gradually changed, showing him riding a horse into the sunset.
Our PR, product, and marketing teams gave previews to print, online, TV, radio, analysts and bloggers so they were ready when we revealed the new site at SES New York in February 2006. Jim Lanzone demoed the site as part of Barry Diller’s keynote.
On the home page that day was a prominent “Welcome to the new Ask.com” link which gave an overview of what was new.
We also made notes about Jeeves’ retirement available easily in our About section, and through Smart Answers to searches about him. Those are still there.
Was your suggestion about Jeeves’ retirement related in anyway to the time you spent impersonating him in the past?
Ha! No, I enjoyed playing the big-headed mascot, but that suit was way too hot.
Part two of this interview continues in tomorrow’s SearchDay, Rebranding Ask.com, Part Two.