How Whirlpool Raised Customer Loyalty by 68% by Revamping its Site Using Persona-Based Design

CHALLENGE When Jeff Davidoff, Whirlpool Brands Marketing Director, joined the company three years ago, he inherited a site that was heavy on data and low on delivery.

"It was a 'We're-talking-at-you,' flat execution online," he says.

So as he began to develop a "new and improved Whirlpool.com," he wanted to get to the heart of the user to drive website development decisions, from the overall feel of the site to the tactical issues surrounding functionality.

His goal -- to meet the needs of differing groups of individuals within the same site and even on the same page.

CAMPAIGN Pleasing more than one type of consumer demographic meant relying on a persona-based redesign process where the target demographic would be sliced into tightly defined segments and profiled via "personas."

Davidoff and Dan Cooke, Manager Interactive Marketing, explained to us how they created eight personas that drove the rest of their web development decisions.

-> Step #1. Look at overall target audience

From a brand perspective, they had to start with overall positioning or the team could have gone too far afield. "We have a target audience, the 'active balancer' who juggles the demands of a career and family life," says Cooke.

The 'active balancer' helped inform the tone and voice of marketing materials. "But we didn't know how we were going to address her needs, and that's where personas come in," Cooke says.

-> Step #2. Persona development

The personas did not necessarily fall under the heading of the 'active balancer.' That's simply "the platform that tells us what we stand for," says Cooke.

The team looked at the different kinds of consumers that come to the site, seeking their pain points and the reasons they were coming to the site. To understand consumers, the team studied existing qualitative and quantitative (conducted by Whirlpool or secondary sources); a series of interviews with real consumers; call center data; website activity statistics.

Based on the findings, they created four "core" personas and four secondary personas. The four main personas were:

--Purchaser under duress (someone whose refrigerator has just broken down and has ice melting all over the kitchen floor, for example) --Planned remodeler --New owner --Owner with repair

The personas were given ages, names, families, occupations and income. The team also created a grid which showed other considerations of the personas, such as their most valuable sources of information (Consumer Reports, sales people).

Each persona also had several paragraphs of text explaining in detail their mindset when they arrived at the site, the trigger that sent them there and their approach to a resolution. Each profile a photo of a real human being to help the creative team visualize the target audience.

Davidoff describes the personas as "the creative brief come to life, the eyes that guide you through the development process."

You can't always trust your own eyes because you're too close to the process, he adds. "We had [the photos] up all the time, all over the place," says Davidoff. Throughout the campaign, the personas' pictures reminded the team to use less "I think…" and more, "well, this persona thinks."

-> Step #3. Development of voice

While the personas are different people with different needs, the copy couldn't reflect too much disparity in tone. "So we established a tone for the whole site," Davidoff explains. "We refer to it as a breezy practicality."

Then, depending on the area of the site, that tone could be "very concise or a little more flowing," he notes.

For example, in the tips area of the site, the tone is more conversational, whereas in the product areas, where people want to simply get to the point of the purchase, the tone is more direct.

-> Step #4. Development of functionality

Websites have many paths to solutions. "People navigate and think in different ways," says Cooke. So once the personas were defined, the trick was to create navigation that would work for them all.

The team created two main ways to navigate. Tabs at the top allowed users to navigate by "all products," "parts and accessories," and "customer support." Or, they could go to one of three clickable buttons in the middle of the page: Kitchen, Laundry Room and Whole Home.

Then, different functionalities were created to appeal to different personas within the site. Two examples:

o Example 1. For a duress purchaser, Cooke explains, "Your ice cream is melting and you need something today. So, how do we help you make decisions quickly?"

On product pages, the team created a set of choices to help narrow the search in what otherwise is a very complex category. Users can narrow their search for refrigerators by type, by finish, by height and by width. Once an item in one of those subcategories has been selected, a user can go to the next subcategory to narrow the search further.

Users can also select up to four items to compare side by side. "Without even leaving the page you whittle down what might be hundreds of selling SKUs to a more manageable consideration set," Cooke says. While Cooke wouldn't share exact numbers, he did say, "We know that people in large volumes are using the side-by-side tracking."

o Example 2. For the remodeler, whose purchase is more considered and who therefore has more time to explore, the team included a couple of different functionalities. First, each of the three areas (kitchen, laundry room, whole home) includes a "showroom" with highlighted hotspots for users to note special areas of consideration.

There's also the ability to email a product. "So if you're shopping at work, you want to send it home to your inbox to talk about it with your spouse, you want to email it to your home builder, it helps you pass on this information," says Cooke.

-> Step #5. Measure

The team conducted a customer loyalty survey online, breaking satisfaction down through seven questions: --Would you purchase the same product again? --Would you buy another product from us? --Would you buy only my brand? --Would you ask for my brand? --Would you overrule a sales person? --Would you switch stores if they didn't have my brand? --Would you recommend the brand?

Davidoff's team was able to differentiate users who had purchased a product and visited both the old site and the new site, based on the dates they had registered the product. Those users received the survey via email.

The team also included a pop-up for new registrants. To be considered a "loyal" customer, they had to choose the "definitely would" answer on four out of the seven questions.

RESULTS
The new site increased page views by 42% (while decreasing the amount of time it took users to reach the correct page by half) and reduced the amount of pre-purchase email and phone questions by 35% and 25%, respectively.

The site also saw a 10% decrease in the number of visitors who abandoned the site directly from the home page.

Last but not least, Whirlpool's customer loyalty index increased by 68%.

Useful links related to this article

Past MarketingSherpa Case Study: 'How to Redesign Your Site to Lift Conversions (Exponentially) with Persona-based Copy & Navigation' http://library.marketingsherpa.com/sample.cfm?contentID=2817