eWEEK content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.
Startup WibiData, whose software uses data analytics and real-time algorithms atop Hadoop storage and processing systems to help retailers sell more of their wares online, has come up with a new sandbox for potential customers to use.
The new addition to its frontline WibiRetail platform, called Experiments by Wibi, enables retailers to analyze various big data streams through machine learning in near-real time without any data preparation. Retailers then can quickly test out several machine-learning models before deciding which ones bring the best results.
Ostensibly, the result is a timely, accurate customer experience that is more likely to result in a sale.
Previously, there was no way to do such calculations in such a timely fashion. With fickle potential buyers and trends that can be quick to change, retailers often find themselves several steps behind what their customers and potential customers are thinking.
Getting Out Ahead of Customers’ Thoughts
Wibidata, which launched in 2010 and already has customers that include Macy’s and Neiman-Marcus, aims to get out in front of a buyer’s thought processes.
“Wibidata can personalize product recommendations, a la Amazon—everything you know about a customer up to that very moment; you can personalize search results, a la Google; landing pages; editorial content; even the imagery associated with a product,” Rob Seaman, Vice President of Product at WibiData, told eWEEK.
“At its core, Wibidata takes all the information you have about your users—every click, tap, swipe, hover, email from you that they’ve opened, search keywords terms they’ve entered (on your site), and so on. It’s using all of that information in real time in scoring machine-learning models that make predictions about their propensities to churn, to convert—their propensity to do something given what you’re about to do for them.”
Using Experiments by Wibi, which was released Jan. 12, retailers can gain better control and transparency over the performance of their models and experiments as they prepare to go to market. Retailers can use these experiments to improve their understanding of which models yield the greatest key performance indicator (KPI) results, including conversion rates, customer engagement, top line revenue and reduced churn, Seaman said.
Wibidata has an intuitive interface that requires no coding or script-writing, making it usable by line-of-business employees. Eliminating coding inherently accelerates collecting KPI results; it also allows retailers to create new intellectual property by customizing models based on their unique data and business needs and training models over their data and nobody else’s, Seaman said.
Real-time Processes Have Been a Long Time Coming
“Digital marketers have been trying to deliver real-time personalized customer experiences for a long time,” Seaman said. “But existing solutions have fallen short in their attempts to personalize customer experiences because they test only a fraction of customer interactions with incomplete data, then deploy changes weeks or months after testing.”
Wibidata is going to market in the retail sector first, but nothing it has built thus far is retail-specific, Seaman said.
The San Francisco-based company, named after a favorite sushi restaurant of the founders, has aimed its product at any enterprise or organization that wants to enable a more personalized relationship with its customers or users—a very important trend in new-gen IT, given continually increasing competition among online marketers to gain loyal customers.