Blog Post

How Intuit uses big data to ‘delight’ you

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

piggybankIf you’ve ever wondered what big data means at an individual level, this realization about sums it up: “I could either keep dying my hair or retire a year earlier.”

That, Intuit (s intu) Senior VP of Big Data, Social Design and Marketing Nora Denzel told me, was the reaction of one of her co-workers after getting a view of her personal finances using Company bias aside, the story is telling of how individuals might expect to directly benefit from analytics today. Large companies use big data techniques as methods to simultaneously increase revenue and save operating costs, but the most-direct benefit to consumers is usually a targeted advertisement or a list of possible social networking connections. That’s pretty lame. A consumer service that actually cares about your bottom line…? Well, that’s something.

“Data for delight”

That’s what Denzel says Mint is, and it’s able to be that because Intuit collects a lot of data on more than 5 million Mint users. Intuit calls it “data for delight.” The service becomes about more than just tracking your spending, but becomes a tool for comparison. Users can compare their financial situations to others who are similarly situated by demographic, geography or other factors. The service can tell users how much money they’d save by switching to a new credit card or refinancing their mortgages.

Nora Denzel

“The [alerts users] really dig is when we give them personalized information,” Denzel said, comparing Mint to her ink-ordering printer that saves her trips to the office supply store. “At least someone in this house is helping me.”

And let’s say you’re more than one person — let’s say five people running a small business — Intuit has you covered there, as well. Small business owners often have little insight  into the broader market, because they lack the tools, knowledge, and time to collect and analyze the relevant data. “You’re running a race and you have no idea where are,” is how Denzel explains the situation.

However, while many small business owners might never have heard of Hadoop, Voldemort or other big data technologies, Intuit has. And it uses them, Denzel said. Its data store is growing exponentially as it collects more data on users from a growing number of sources. “We call it big data for the little guy,” Denzel said.

For example, a flower shop that uses QuickBooks or any of Intuit’s small-business services can compare its finances to other flower shops in the same city or elsewhere based data gathered from Intuit’s more than 4 million small-business customers. Or the flower shop could be alerted in real-time that it’s about to pay too much for new packing supplies because Intuit’s system has spotted other businesses buying the same thing for less elsewhere.

“Data for decision-making”

But before you can delight anyone with data, you first have to attract and maintain customers. That’s why, Denzel said, about half of Intuit’s big data effort goes toward marketing and social media engagement, and why she has such a unique job title. Marketing especially is a blend of art and science, she explained: “No offense to Mad Men, but it’s all tested now.”

Although “the last mile is always humans,” Denzel said Intuit continuously tests everything from ads to product design to find out what’s working best for customers. After decades in the industry, one of the biggest — and most-humbling — changes she’s seen thanks to big data is that “we’ve outsourced product management to consumers.” If your experience tells you something should work but customers don’t respond, you’re wrong.

Notice the fourth tweet down.

On the social media front, Denzel said Intuit is adamant about monitoring and analyzing social data, as well as about engaging with customers. Sentiment analysis and data mining can tell a company a lot what consumers are saying or feeling about the brand, but they can also help the company meet customer expectations. Denzel said if a customer complains on Twitter, that person expects a response from the company (it’s “their Batphone to your company”). And if they complain on Twitter and then are forced to call, they expect the customer service agent to know how the situation has escalated.

“The chief communications officer is now the chief listening officer,” Denzel said. (I assume Intuit is doing plenty of listening thanks to Friday morning’s outage for QuickBooks Online.)

And Intuit is listening. Last week, I explained how LivePerson (s lpsn) uses predictive models to help decide when website visitors might need to engage in live chat with a customer service representative. It turns out Intuit is a LivePerson customer and actually uses text from chats to make Intuit customer service even better.

Denzel said Intuit feeds that data into Hadoop, which lets it determine certain words and phrases that suggest a customer is going to stop using an Intuit product. When those phrases pop up in future chats, a customer might get transferred to the best-available agent; if they’re used on some public forum, that customer might get a call. As with most things big data, an unexpected phone call might be a little bit creepy at first, but it’s probably less so if the problem actually gets resolved and both parties get what they want.

Feature image courtesy of Shutterstock user val lawless.

One Response to “How Intuit uses big data to ‘delight’ you”

  1. Derrick does Mint succeed at “data for delight” for you or people close to you?

    I’d love to hear the specific examples from Intuit of the most successful big data based wins for customers. I don’t think the “save by switching” I’ve seen appear to be big data driven, but benefits based on external factors like credit card rewards and affiliations.

    I have two Mint accounts, one Canadian and now one American, as we’ve recently used to the US.

    The Canadian one has helped with some relevant large amount alerts, but there is no big data or even smarts needed there, just simple automated alerts.

    The budgeted alerts continue to be far more noise than signal — example almost alternating months depending on how many days are were in the month to seems, I receive alerts that my bi-weekly mortgage payments have resulted in “Unusual Spending on Mortgage & Rent”.

    Now, as I start the US account I’m drowning in Unusual Spending alerts.

    Back to my Canadian account which I’ve had much longer. The “save by switching” are few and have always seems out right bad advice for the same few products. Intuit has not taken the time to analysis the small pool of Canadian credit cards.

    I can’t remember why now, but the graphing didn’t actually show me the information I was most interested in so I stopped using it.

    Another thing I don’t get is the frequency of wanting updated credit rating? Is this because of affiliation with the consumer credit companies? With my net worth, I can’t justify the monthly fees, so I get it once a year for free. I’d actually love to see get smart enough to provide it’s own rating of me — that is what it sounds like Nora Denzel is promising

    I also have to ask does Mint make printer recommendations to help Nora with her frustrations? — now that would require forensics and intense data analysis.

    Don’t get me wrong, Mint is an essential service for me (that I would pay for), but almost exclusively for the simple ability to search for transaction across my accounts, and the confidence that I have my records going back longer than six months.

    Thanks for the sparking my memories and thoughts Derrick! I love your cloud reporting and insights!