"Imagine having a Google Photos kind of search for your enterprise data," says Box CEO Aaron Levie.
Levie, whose company offers file-management in the cloud, was referring to the image-recognition feature that's able to identify family members' faces in your photos, as well as objects in the background, without being asked. Soon enough enterprise software "will be telling you what you should be working on, who you should be communicating with, what information you should be looking at," as Levie put it in an interview with Inc. last year.
Why stop there? Imagine a system that tells you what to search for before it occurs to you. Imagine a system that runs the search automatically, as well as hundreds of others, and tells you which results are the most important.
"Big data" has been buzzed about for years, but automated solutions for coping with it are hitting the market now. Sean Byrnes is the cofounder and CEO of Outlier.ai, a startup that automates business intelligence, which raised $2.2 million in February. "There's so much data that you can't look at it in spreadsheets," Byrnes said. "You can't have three dozen dashboards. The world doesn't have time for that."
Outlier.ai is part of a new surge of software products with the capacity to "learn," to put it in human terms. CEOs like Byrnes are eager to make their fortunes by enabling other executives, as well as employees, to understand their own businesses better. Possible advantages range from better analyzing growth metrics and customer churn rates to identifying where staffers are spending their attention and what resources could help them be more productive.
Outlier helps customers see patterns and the deviations from them, said Byrnes: "What's normal for this time of the year, this day of the week, for selling shoes in the UK? What's normal for this time of year, this day of the week, the trend over time, for the cost of this Facebook ad campaign? And the second step is then detecting what doesn't fit." Anomaly detection has been around for years, Byrnes noted, but automatically cross-correlating the anomalies and being able to highlight which ones are worth paying attention to is new. Crucially, Outlier.ai's technology notifies you only of the business changes that it "thinks" are significant.
We live in the age of SaaS apps that store event logs meticulously, completely, and forever. Work product is captured in the cloud, but so is the metadata about what was done to the content and how it was done. The sheer volume of business data being generated on a daily basis is both a challenge to decision-makers and an asset that companies can't risk ignoring. Businesses that stick to manual analysis methods will be left behind as competitors leverage technology to find surprising patterns and act on them.
DocuSign CMO Brad Brooks gave an example involving drug trials. A number of pharmaceutical companies use DocuSign to manage legal forms and records of whether patients followed directions. Brooks said, "We can give [the pharmaceutical companies] the data as to what's actually going on and the behavior between the customer interaction, because it's now all electronic and you can actually see that digital footprint of what's going on."
Brooks explained, "We capture all this metadata or event data that's happening around that, to allow for businesses to become much more effective in terms of how they're running the business." Like Box, DocuSign intends to use AI technologies in coming years to augment its users' workflows.
Software that produces automated insights is becoming robust now, as opposed to five years ago, for a cluster of reasons. Nick Elprin, CEO and cofounder of Domino Data Lab, explained that he couldn't have launched a work environment for quantitative analysis without the hardware improvements that cratered the cost of high-volume data-processing.
Being able to spin up a business without capital expenditures on servers, instead relying on cloud hosts like AWS or Microsoft Azure, means that startups can access every phenomenal advance in basic computing just as soon as established companies. Machine-learning innovations have expanded the fruitful methods of data analysis itself (along with prompting overuse of the term "AI").
Meanwhile, the secular shift to cloud-based software applications has reached a tipping point, long after being initiated by Salesforce in 1999. Most sophisticated companies use a variety of SaaS options instead of resorting to one bespoke solution, necessitating an integrated analysis layer.
Increasingly ubiquitous APIs make that analysis layer technically feasible, since data can be pulled in from every different software service. Aggregating and cross-correlating the data from marketing, sales, product, and so on, helps to track which changes in a given area cause ripple effects throughout the rest of the complex system that comprises a business.
The first layer of the enterprise AI ecosystem exists in the SaaS apps themselves. "It can be about the app, it can be about a lot of other things, but what customers really find valuable about us is the unique data that we're capturing, and how we can present it back to them to make their business run differently," DocuSign CMO Brad Brooks told Inc.
The next layer up integrates a couple of services, such as Zuora Insights uniting its subscription billing data with Salesforce stats to easily upsell and cross-sell users. The third layer, the highest layer of integration, lives in services like Outlier.ai and Periscope Data. The two companies have different approaches -- Outlier's algorithms are meant to do 99 percent of the work for you, whereas Periscope is intended for data teams -- but both systems want to be your source of business insights.
Specialist business-intelligence shops aren't the only ones with their fingers on this pulse. More and more huge companies are announcing AI initiatives. IBM was early on the scene with Watson. Salesforce has Einstein, Adobe has Sensei, and Domo has Mr. Roboto.
Even when they don't go for a cutesy name, incumbent software companies are keenly aware of the need to incorporate advanced analytics capabilities into their products. Younger players are beefing up their war chests; Looker announced a $81.5 million Series D round on Thursday, bringing the company's valuation close to $1 billion.