Sergey Nivens - Fotolia
If you feel bewildered after looking through the collection of IBM analytics tools, you're not alone.
"Every year, their packaging changes, their branding changes, but the power is in how it all fits together," said Brian Hopkins, an analyst at Forrester Research. "Our clients can't necessarily get their heads around the value proposition."
Heading into the IBM Think 2018 conference, which takes place next month in Las Vegas, the complexity of offerings and lack of cohesion across analytics software product lines is the No. 1 issue analysts expect the company to address.
For example, Hopkins said he was recently briefed by IBM representatives on updates to Watson Explorer, an application that uses the Watson cognitive engine to analyze structured and unstructured data. He was surprised to learn that the particular product being discussed is an on-premises desktop application, though there is a separate private cloud option.
Hopkins said the lack of a public cloud version undermines IBM's messaging and product development aimed at making the company a modern, cloud-first software provider -- something that he wants to hear more about at the Think event.
"They were using the Watson name to brand a lot of cloud-based analytics tools," Hopkins said. "I want to see how that's progressing. I want to know how it all fits together."
Part of the confusion around all the IBM analytics tools comes from the proliferation of Watson-branded software. At times, it seems Watson is a part of everything IBM sells. But what Watson actually is, and what it means for a piece of software to be branded with the Watson name, remains a bit of a mystery.
For Ventana Research analyst David Menninger, this became especially problematic when evaluating Watson Analytics, a data analysis and visualization tool that uses the Watson engine to guide users through an analysis. Menninger said it took him some time to understand how this was different from the Watson platform itself. Given the time and money IBM has spent pushing Watson as the face of its company, this kind of confusion hurts its positioning with potential users, in his eyes.
"The company made a big bet on Watson," Menninger said. "The confusion was real, and it was justified. That's one of the things they need to address."
IBM has already taken some steps to do so. In January, the company reorganized the various offerings that fall under the IBM Analytics umbrella into three platforms: Hybrid Data Management, which includes the Db2 database and related technologies; Unified Governance and Integration, consisting of IBM's data integration, governance and quality tools; and Data Science and Business Analytics, which combines SPSS Modeler, Cognos Analytics and Watson Explorer, among other applications.
Customers can pay to use the different products in one of the bundles via a new FlexPoints pricing model. It gives them points that can be applied to a single technology or a combination of software; the products covered by points can also be changed over time as an organization's data management and analytics needs change.
Rob Thomas, general manager of IBM Analytics, said the vendor made the bundling and pricing moves in order to simplify the purchasing process for users. "I would call it a revolt against complexity in enterprise IT," he said. "The overarching thought is that less complexity is good."
IBM still positioned for AI success
Assuming that IBM can get its story fully straight, the company may be well-positioned to capitalize on the ballooning interest in artificial intelligence. Experts feel its analytics software functionality is strong. But what really sets the IBM analytics tools apart from the competition are their end-to-end capabilities.
"Algorithms get all the press in AI, but the prerequisite is data," said Forrester analyst Mike Gualtieri. "What most enterprises don't realize is that data is the underpinning of AI."
AI algorithms need lots of labeled training data to learn to recognize images or text. This is where IBM's broader portfolio of products comes in. Gualtieri said most other AI companies today offer software that handles just the machine learning and prediction aspects of AI. But IBM has a mature data platform in Bluemix -- recently renamed IBM Cloud -- and other data management tools that allow users to tie their AI applications straight to data-generating systems.
But while the underlying software is strong, Gualtieri said, it's still incumbent upon IBM to explain how its tools fit together.
"Data is the key," he said. "IBM has to show the importance of this. They're in a stronger position than most other companies to take advantage of AI trends. The key thing I'm going to be looking for is how [is IBM] going to bring all these things together."