NEW YORK---IBM on Tuesday launched Watson Analytics, a service that's designed for line of business executives, uses natural language tools as well as predictive algorithms and offers a freemium pricing model. IBM is hoping to give Watson cognitive computing more exposure and usher in an era of bring your own analytics to the enterprise.
Big Blue has said it will spend $1 billion to commercialize Watson, but there are challenges with scaring since IBM's cognitive system has to ingest data, learn from it and has to be customized in many implementations across verticals such as financial services and health care.
Watson Analytics is an avenue to bring analytics to more executives on a smaller scale. "This is an analytic journey and we're dealing with people with varying skill sets," said Marc Altshuler, vice president product management for business analytics.
Under the freemium model, Watson Analytics is available on the desktop, Android and Apple's iOS via a Web site. IBM would charge for more sophisticated analysis, but the tiers and pricing model are still being worked out. IBM's challenge will be to highlight insights and entice users to drill down into the data and ultimately pay. Altshuler said that IBM is looking to get Watson Analytics into the hands of as many people as possible.
For instance, Watson Analytics could ingest 20 spreadsheets and data from Salesforce, Teradata and Oracle systems to start making connections and answering questions in a natural language format. A marketing pro could ask about what campaigns are delivering the most return, a HR leader could find data about a retention and a sales person can find about deals that'll close soon. In some respects, IBM is trying to push microanalytics via what it calls a "single business analytics experience."
And that experience will largely ride on how easy executives can plug data into Watson Analytics. A demonstration at IBM's Watson Analytics event revolved around data imports as well as visualization.
Neil Whitney, director of product experience and design, highlighted a bevy of icons and connectors to import data from everything from Salesforce, Google Docs, Oracle, Box and a bevy of other tools. Not surprisingly, connectors to IBM software such as Cognos and SPSS were in the fold. Under Watson Analytics self service model users can also upload files. Whitney noted that IBM spent a lot of time to make sure Watson Analytics could absorb data easily since that's a key hurdle to adoption.
Inhi Cho Suh, IBM vice president of big data, integration and governance, said behind the scenes Watson Analytics is utilizing the same natural language, algorithms and discovery tools as larger cognitive computing systems from Big Blue. Suh also said Watson Analytics aims to make data ingestion easier by automating things like quality, cleansing and creating sets of information for execs. "Behind the scenes we're adding predictive analytics and data science in a way that's consumable," she said. Suh added that collaboration is another key component in Watson analytics so business users can share data.
Watson Analytics, which will run on IBM's SoftLayer cloud infrastructure, is now in beta and will be generally available in November. There aren't apps for Watson Analytics yet as IBM is trying to cover its various platform bases with a Website that's desktop and mobile friendly.
IBM said Watson Analytics can carve out some mindshare with executives because it doesn't assume that users already have data ready for analysis. Watson's strength over other applications is that it cuts down on data preparation and loading, which can take 60 percent of the time involved with an analysis project, and deliver relevant information without questioning. There's even a "fix it" button to address data quality issues and repair them as well as various sharing tools.
Once data is dumped into Watson Analytics, the system looks key items such as sales, location, product and audience and then serves up a visualization best suited to the information.
Suh said Watson Analytics is designed to be personal and designed for execs of all stripes. "Watson Analytics is for the personal user who is bringing their own data," said Suh. Watson Analytics will be available as a beta in 30 days with general availability and a freemium model in November.
Here are some desktop screenshots for Watson Analytics.