In advance of the Web Strategy Summit coming up May 4-5 in Calgary, Matthew Nish-Lapidus sat down with one of the presenters, Dianne Howie to get a preview of her presentation. In this interview, she shows how Logitech used qualitative research to dig deeper into issues highlighted by Logitech’s web analytics and how the qualiative actually allowed Logitech to identify and solve specific consumer problems.
2009 Web Strategy Summit – A word with Dianne Howie
by Matthew Nish-Lapidus
April 1, 2009
At the upcoming Web Strategy Summit Dianne Howie and Heather Searl will speak about the role of research in their session “Going Beyond Web Analytics to Dig into the Minds of Your Users.” I was lucky enough to get a few minutes of Dianne’s time to ask her a few questions about her session.
Matt: In your Web Strategy Summit session you are going to talk about using qualitative research to back up web analytics quantitative data. When was the first time you used this technique, and what kind of benefits have you seen?
Dianne: Both Heather and I have used qualitative research for years. The Logitech remote controls division began using qualitative research web research when Heather joined the company in 2006. Logitech’s Harmony remote controls have web based setup software, allowing the company to collect a lot of quantitative data including: the time to set up devices, which pages they visited, and where people quit the setup process. Previous projects had tried to improve the setup process based on this information, but it didn’t work was well as they hoped. In addition, the quantitative data and marketing data didn’t match. They had high return rates. They could tell when a returned remote wasn’t fully set up and where people dropped off. But that wasn’t enough to know how to improve the design.
It makes a world of difference to see someone struggling with the setup process in person. You see things that could never be revealed by web analytics. For example, people tried to use the remote with the (very realistic) store sticker still on top of the LCD display. It wasn’t obvious that the sticker had to be peeled off! Another example… People didn’t know where to look on their devices for the model numbers. Just providing some simple tips and pictures during the setup made a huge difference.
The qualitative research made a compelling case to get the funding to do redesign of the setup process properly. Videos from visits made the points well: for example one person spent 12 minutes going around and around trying to fix a mistake and not understanding what he had done wrong. After applying the findings from the qualitative research, Logitech measured significant improvements in customer satisfaction for remote setup.
Matt: In your research, you use in-home contextual inquiries to better understand your customers. How did that level of research help you make sense of the quantitative data form your web analytics?
Dianne: Logitech knew that there were problems with the remote setup process, but they didn’t know how the issues interacted or how to prioritize them. Qualitative research helped to prioritize issues, rather than just fixing the easy or fun to fix issues. Having a new perspective from qualitative research helped the team to refocus and fix the things that had the most impact on the customers.
For example, we did some in home visits setting up remotes with radio frequency (RF) instead of the typical IR), allowing the remote to communicate to components in cabinets. Initially, the team knew that customers didn’t like to see wires hanging around, but this wasn’t thought to be a big issue. After watching customers struggle with the wires and voice their displeasure with the appearance, dealing with the wiring became a significant project goal.
In addition, RF remotes had a large customer service call rate. Watching customers set up RF remotes, we realized how difficult it was for customers to understand what was going wrong when they had to deal with both remote setup and RF setup all at once. The customer service call rates made much more sense in this context, and the solution was more obvious.
Another example was a function called Button mapping. Web analytics indicated that 50% of people went into the button mapping page, but most people did nothing. It was unclear why. After watching people, it became clear that people that were lost were just trying everything and ending up on this page not knowing what to do.
Matt: How do your methods work on an ongoing basis? As more and more analytics data is collected do you have to revisit your qualitative research?
Dianne: Qualitative research is ongoing process. Over time, in home studies can narrow down to emphasize specific points such as RF setup. There also becomes an increasing focus on usability testing of the proposed solutions before they go live. Logitech has used a variety of approaches for usability testing from paper prototypes, to PowerPoint prototypes, to testing with the actual remote in a “simulated living room”. This iterative approach incorporating customer feedback at each stage helps to produce the best end product. Logitech’s latest remote, the Harmony One, has already garnered praise based on its design and ease of use.
You can see Dianne and Heather’s session at the 2009 Web Strategy Summit in Calgary.