This picture is intended to illustrate the flexibility of our WARP survey technology. It shows four different types of surveys running concurrently on the same site. Each survey serves a different purpose and is presented under different conditions to visitors exhibiting different behaviors. But no visitor would ever see more than one invitation. And all of this happens without any IT involvement (beyond the original tag deployment). WARP is deployed via the site’s CMS as a single, universal tag that requires no customization and sits in the footer of each page and is the last element to load.
Here are typical business scenarios behind these kinds of surveys. In the top left corner, we see the invitation to a longitudinal survey – the kind that most websites run to monitor the user experience and drive continuous site improvement processes.
In the top right-hand corner, we see the Multi-condition Behavioral survey. In this scenario, Zara wants feedback on a dress it is considering manufacturing in larger sizes than the original design. The survey tells the WARP tags to look for visitors who meet all the required criteria, which are: navigate to the ‘Women’ section; then to the ‘Dresses’ category; then use either the ‘Large (L)’ or ‘Extra Large (XL)’ filter to narrow product selection. This sequence yields a high probability that the visitor is a female meeting the target audience for the proposed dress. Once the tags detect a visitor whose behaviors match the criteria, WARP triggers the survey – which in this case has two quick questions and a photograph.
In the bottom left corner, the survey is triggered more simply – anyone who navigates to the Stores page. Those who do so can reasonably be expected to be multi-channel shoppers, since they are online and looking for a physical store. Zara, in this scenario, wants to understand if the recessionary economy is likely to shift shopping patterns away from the stores (their primary channel) to their website. The results would be used to inform their short-term channel investment allocation.
In the bottom right corner, we see a Site Search survey. Less than 20% of visitors typically use keyword search on a retail site, so it could take Zara weeks or even months to gather actionable, statistically significant, search-related feedback through their longitudinal survey, which might invite one in every 20 visitors. Using WARP, Zara can shorten collection time dramatically by identifying every visitor who uses site search and asking questions only of them and specific to their actual searches.
In these scenarios, the Zara market research department is serving four different internal customers – the web manager for the longitudinal survey; the manufacturing manager for the dress survey; the Chief Strategy Officer for the channel survey; and the Search team for the site search survey. Using WARP, the research team can get answers from separate, behaviorally qualified samples in the shortest possible time. When a research department has to go through its IT team to get surveys deployed, the lead time often renders the answers moot by the time the research is deployed and the responses collected. To use an outside panel vendor would be costly and might yield less reliable samples.
Not with WARP. One major brand has used WARP to run 16 concurrent surveys across several sites in a variety of languages.
By transforming the website into an “always on” research channel, WARP helps the research team provide a level of responsiveness and value to their internal customers that would not be cost-effective through any other means.
That, we believe, makes the story told by the picture pretty compelling for any research unit.
What do you think?
–Roger Beynon, CSO, Usability Sciences