Tableau Software has made a great name for itself among savvy data analysts (and occasionally journalist types), racking up more than 20,000 customers and revenues that are nearly doubling every year, but it’s still far from a household name.
That’s because no matter how much easier the company’s flagship desktop application is than historical options, and how much prettier and interactive its charts are, it’s still enterprise software. If the company is going to fulfill its vision of opening up visual data analysis to millions of small businesses and even consumers, it’s going to have to build something they actually want to use.
On Tuesday at its annual user conference, Tableau announced it’s working on a new tablet-based product, codenamed Project Elastic, that it thinks can finally bridge that gap. Designed specifically with mobile users in mind, the Elastic technology is supposed to be a paragon of intuitiveness, helping users answer the questions they have about their data using just two fingers and a modicum of brainpower. Simple, and at times even fairly complicated queries, are just a tap, swipe, scroll or pinch away.
The person leading Project Elastic is Dave Story, Tableau’s vice president of mobile and strategic growth, who has a history of building technologies with consumers and small businesses in mind. He has been with the company about a year, coming from LucasFilms, where he was CTO. Prior to that, Story was vice president of product development at Adobe, and CTO for Intuit’s small business division. At [company]Tableau[/company], as his job title suggests, he’s responsible for the company’s strategic growth, and mobile will play a big role in that.
“The idea is going beyond our core audience, which has been business, and reaching out to the consumer,” Story said, adding, “The motivation is what would you do now if you wanted to disrupt or wanted to start over with Tableau.”
He thinks the answer is that you’d figure out how to reinvent the product for the tablet market, giving users what he calls a “lean-back experience” (maybe on the couch or an airplane seat) rather than a hunched-over-the-keyboard experience. This way, the thinking goes, a small business owner with a trove of sales data from Square, or an individual consumer with a log of her activity from a fitness-tracking app, can try to make sense of that information in an intuitive way without fighting over the data on one hand and the software on the other.
Project Elastic looks good, but isn’t complete
I saw a demo of the Elastic prototype recently, and it’s easy to see why Tableau is excited about its prospects. Datasets open in the type of chart the software thinks is most appropriate (right now that’s basically limited to a bar chart for data in categories and a line chart for data by time), and it’s as simple as one might expect to scroll between categories or time frames. Want to focus on a specific data point? Just tap on it. Filter one out? Swipe it away, Tinder-style. Story zoomed out to a four-week window of sales activity and began scrolling through a rolling four-week window for an entire year.
He showed me an example using a dataset about sunspots, in which he zoomed in from centuries to days using the classic two-finger gestures we use to make the screen larger in most apps. The driving question behind the design, he said, is, “How do we make it so that when you reach to do something, we do the right thing?” Users should be able to assume everything is manipulable, he added.
Perhaps it’s because I was seeing a pre-formatted demonstration dataset, but there was no noticeable step of having to mess around with rows or columns, or change data types from numbers to dates, for example. Story said this is largely the work of Tableau’s product and R&D teams, led by Jock Mackinlay, which have been working on techniques to automate a lot of that work and simplify the parts that still need to be done manually.
“When the visualization part gets this fast,” Story said, “the data cleansing part overwhelmingly takes over.” Tableau plans to spend as much on R&D in the next two years as it has in the past ten, and solving that problem will be a major focus.
The two big things missing from Elastic right now appear to be built-in connectors to third-party data sources such as Square, Stripe, Jawbone or Nest, for example, and a broader array of options for visualization types. Regarding the former, Story said they’ll come, although that will likely require some work with the data providers in order to ensure Elastic is receiving data in a workable format.
When it comes to visualizations, Story said the mission of Elastic is “not so much what viz type should we have, but what kind of question are you asking.” If users have location data, for example, that should naturally come up in a map. However, he added that he doesn’t think tablet users will want or need the whole gamut of options for ways to visualize their data — everything from pie charts to scatterplots — nor will they want to analyze it across nine dimensions. Tableau Elastic will probably top out at between three and five dimensions.
“I don’t think we need to enable the user to visualize more than that,” Story said.
Now to find out who really wants to visualize their data
The company plans to release Elastic as a product at some point in 2015. Perhaps then we’ll see whether the market for personal analytics apps beyond Microsoft Excel is really a thing, or how big of a thing it can be. We’ll also see whether Tableau’s tablet-first strategy is the way to open it up. Other attempts to simplify and visualize data analysis for individuals, such as startup DataHero’s namesake cloud service, still focus on the browser. Some business-centric analytics startups such as ZoomData started off on tablets before realizing there’s still a huge desktop user base.
However, while the startups in the space are growing, they lack Tableau’s resources and name recognition. They also don’t have shareholders, who have already become accustomed to skyrocketing revenues, to please. Those factors put Tableau in a unique position as it tries to expand to a new class of user: It has the tools to build a great product that people should pay attention to, but it also might be a little risk-averse as it tries not to scare public market investors and to ensure its first foray outside its comfort zone isn’t a noble failure.
Story is convinced Elastic will strike the right note with its target users, who he thinks really do crave the Tableau experience in a format that’s easier to consume. He acknowledges that although Tableau Desktop is state-of-the-art for a desktop product, “it’s still too hard to use” for people who don’t deal with data for a living. So as Elastic heads toward its release date, the finishing touches will involve form, function and whatever it is that makes certain apps actually enjoyable.
“How do we just make it fun?” Story asked. “I’ve got on my backlog ‘Add whimsy.’”