A smartly selected stream of structured data and a seamless rewards system comprise Cardlytics Inc.’s recipe for success. The financial data processor’s commercial Big Data stream consists of payment and  purchase data, translated from the “simple act of swiping your [credit] card,” which generates an immense amount of information. Cardlytics’ Chief Strategy Officer Craig Snodgrass explained in a live interview at Hewlett-Packard Co.’s HP Vertica conference with theCUBE co-hosts Dave Vellante and Jeff Kelly that the company’s business model revolves around taking that data from banks, then using it to generate insights that help advertisers bring in new customers. He shares the importance of real-time data analytics, and why Cardlytics’ use of HP Vertica enabled the company to ditch traditional methods for processing sample and survey data.

How Cardlytics works

“We partner with banks and offer rewards through the bank to visit an advertiser. No coupon needed, you swipe your card and the money goes into account,” Snodgrass said. The insights they provide focus on determining where people like to shop and seek out relevant deals.

To reach consumers with their campaigns, Cardlytics goes where the customers are: “banking has moved online. People log on their banking account nine times a month. That is where we meet them. We are in the mobile apps as well,” Snodgrass said. Of the 400 banks they partner with, each has an app. “So we get them online and on mobile.”

Explaining how the system works, Snodgrass  said “the bank buys the hardware and puts our software behind the wall. We have all these installations everywhere. We have a stream of data coming back to us. “ Based on that influx of data and the insights it generates, “we create campaigns for advertisers. Advertisers pay us.” To monitor their campaigns, “they have a portal where they look at how their campaign is running. They get to see the purchases that these impressions are driving. They can see that real time.”

Asked how they provide value to merchants and keep the business model crisp, Snodgrass explained advertisers will “never going to be able to buy this data, we are protecting the data through our business model. What we can give out is insights. If they want to see what’s happening in a geography, we can aggregate that.” The insights focus on identifying the best 10 percent of an advertiser’s customers and the places they shop. They are extremely useful when acquisitions are made and two stores are close to each other, one of them needing to be closed. The insights help them find out “which one has more loyal customers.” Another popular use case is deciding where to build a new store to maximize sales and profit.

As far as the pricing model is concerned, advertisers “pay by people coming into the store. It is a pure performance driven model.  They can truly measure effectiveness” and the return on their advertising spend. “We guarantee an incremental revenue. They know what the ROI is” right from the start.

How tech has evolved the financial data business

Asked how the technology evolution has helped the business, Snodgrass said “the technology journey has been quite the ride.” In the beginning, the company targeted mid-tier banks. With the first large bank, “the tech was not going to keep up. We were using a competitor to Vertica at the time. We could not get the analytics we wanted,” As they moved to Vertica, “what we were looking for was rapid cycle insights.” Vertica allows them to deliver same day/next day insights for customers.

The benefit is that Vertica allows them to use the “entire full data set and get insights in seconds.” This makes sampling and traditional surveys antiquated tools. Every one of the swipes being tracked is in fact a survey.

As far as future technology projects go, Snodgrass said he was “on a new quest. Vertica has fulfilled  our first quest – rapid cycle insights. Now my quest is, I need the simplest path to productize the insight. What is the easiest, slickest, most beautiful way to productize our data? It’s a scale and ease-of-use challenge.” His vision for the perfect solution would be something similar to Google  – “you sit in front of it, write the question and get  the answer. Google for business intelligence.”