Rue La La harnesses the power of analytics to improve customer personalization


Like most retailers, Rue La La wants to know a lot about its customers — what they like, what they don’t like, and most importantly, what they are likely to want to buy.

The membership-based online retailer used a data warehouse for about eight years, storing customer information gathered through member sign-ups and purchase history. But executives at Rue La La wanted to know even more.

The members-only shopping site, which offers discounts of up to 70 percent on a wide range of products including fashions for women, men and children as well as home décor items and travel experiences, boasts millions of members. With so many products and so many members, it was difficult to determine what each customer was likely to want to buy.

In February 2016, Rue La La shifted to Snowflake Computing’s cloud-based data warehouse, where retailers and other businesses can store and analyze large amounts of sophisticated customer data.

“We did not have to replace what we had previously,” says Erick Roesch, Rue La La’s director of business intelligence and data warehousing, “but in moving to a cloud-based system, we were able to extend what we were doing in a more elastic manner.”

Personalized emails

Beyond knowing what customers purchased, the new data warehouse knows what products they looked at and didn’t purchase, as well as what web pages shoppers visited.

Using the Snowflake system, Rue La La can track exactly how customers got on to its website, what products they looked at and where they moved around on the site.

“This gives us a 360-degree view of our customers,” Roesch says.

Right now, Rue La La is using this information to customize the daily email offers sent to members — a membership of 16 million households means the company sends out millions of emails every day.

“Every day we email a list of sales items to our members,” says Peter Connolly, principal software engineer with Rue La La. “We used to manually select what items were listed on that email. Now we can use the data from our warehouse to personalize each email so that the offerings are the items we think that member will be interested in buying.”

In addition to the behavior data that Rue La La gathers, customers can also contribute information that will determine what products are shown, such as expressing what brands and products they like.

The sale emails typically include about 20 products each day, chosen from the tens of thousands that Rue La La offers. The company typically offers limited quantities, so it can be urgent for customers to make purchasing decisions quickly before a particular product is gone.

Although Rue La La does not reveal specifics, executives say there has been a significant increase in the number of products purchased as a result of these emails. “By utilizing real statistics, we have seen a big lift in sales and the results have been overwhelmingly positive,” says Ben Wilson, lead data scientist.

Marketing teams can utilize data from the platform to evaluate the performance of each email campaign and fine-tune efforts if necessary; while analyzing customer behavior can determine what current brands shoppers are likely to want to buy, Rue La La can also use customer behavior to determine which new brands to promote.

“We look at customers who have shown interest in similar brands based on price, product features and quality,” Connolly says. “For example, if they like luxury items or if they prefer low-priced items. Do they like modern or traditional items?”

Predicting desires

While Rue La La is using customer data to personalize emails today, it plans to do even more in the future. The company wants to use the data to personalize what customers see when they log on to the website.

The analysis could determine product recommendations when customers are on a web page or at checkout. While many retailers offer “products you might like” based on items customers have already purchased, Rue La La is looking to use a lot more data about customer behavior in determining these recommendations, as well as other ways to personalize shopping experiences.

“We will do some exploratory testing for future features,” Wilson says.

The ability to make good predictions about what customers will want comes from powerful analytics that not only predict likely customer desires based on prior behavior, but also can evaluate the results, Roesch says.

Beyond Rue La La’s own customer data, Snowflake was able to contribute best practices information from other clients, Roesch says. “Their engineers continue to work with us. Every company has different business needs and they are able to adapt to our abilities and desires,” he says.

Making the transition from Rue La La’s traditional data warehouse to that at Snowflake was relatively simple. No software or hardware was required and the transition was completed in a few days.

“Now we have all our data in one place, and the power to conceptualize what we want and get results in seconds,” Roesch says.

Increased revenue

Snowflake works with a wide range of businesses that want to understand their customers better. “Any retailer that has an online presence can benefit from our service,” says CEO Bob Muglia.

Most of Snowflake’s customers are small to mid-size businesses that want to do some of the same customer behavior analysis that the online titans can do.

“Companies like Netflix and Amazon have a huge number of technical experts who can help them provide extensive customer analysis,” Muglia says. “But there are only about 10 to 20 companies that operate at that level by themselves. We provide the same kind of experience to smaller retailers at a fraction of the cost and a fraction of the required resources that large retailers can use.”

One type of business with a lot of need for this type of analysis is gaming sites. Many generate revenue from special offers and tools that are sold to the gamers on the sites. By knowing what type of tools and added features users want, gaming sites can increase their revenue, Muglia says.

What retailers do with all the data and analysis that Snowflake provides varies. “Every company has a different strategy,” he says. “Emails are a common way to use this data, while others use it to structure their website.”

Retailers do not need to purchase any additional servers or software to convert to the Snowflake system. Most retailers can get started on the service within days — some within hours, Muglia says.

Retailers can work with a Snowflake systems integrator to decide how to structure the program and choose the tools. The expert will help build the reports users need and figure out how to analyze the data they are receiving.

Lauri Giesen is a Libertyville, Ill.-based business writer with extensive experience in covering payment and finance issues.


Comments are closed.