companies gain access to a greater variety of business data and want to be able to explore different views of that data without rebuilding their data warehouses every time. “Fifteen years ago, the data you had inside your company was the information you used to run the company,” says Hewitt. “Today the amount of data companies have to deal with is doubling every 24 months. This is not a nightmare waiting to happen; it’s here.”
Kalido was created as a division of Shell in 1997 and spun off as a private venture in 2003. It now has 150 employees in its Burlington and London offices and works with 250 customers around the world, according to Hewitt. Atlas Venture and Matrix Partners of Waltham, MA, Benchmark Capital of Menlo Park, CA, and Balderton Capital (a London-based spinoff of Benchmark) have poured $29 million into the company over four financing rounds.
In the most recent version of Kalido’s flagship “Information Engine” product, released in January, the company has extended the graphical-modeling concept to the point that even non-experts can create and play around with the business models. In the Labatt case and other past implementations, business users and Kalido consultants would sit around a whiteboard or a PowerPoint screen creating diagrams of a company’s business structure showing which divisions reported to which, and where data should flow. The consultants would then use Kalido’s graphical configuration tool to create a customized data warehouse.
But in the new release, the visual modeling tool has been simplified so that the business customers themselves can do the driving. As users draw or redraw relationships between business entities, rules built into the modeling tool check whether the relationships make sense, and the software then implements those relationships in the underlying databases. “There is no intermediate stage of definition required,” to quote a recent blog post by Andy Hayler, who led the Kalido team that originally developed the Information Engine and is now an analyst at UK-based Bloor Research. “The model is the warehouse, essentially.”
Two weeks ago, Kalido released a free version of the Business Information Modeler and set up an online business-modeling community; created using the Google Groups service, the community is intended as a place where business managers can share models and trade stories about how the models are working in their own businesses. Without such models, expensive business intelligence software from companies like Cognos (also now part of IBM) might produce “pretty pictures and charts,” in Hewitt’s words, but it won’t reflect changing business realities. Recounts Hewitt: “We had a large pharmaceutical company recently where the CEO said, ‘We’re thinking of reorganizing the company by type of business—emerging versus established. How quickly could you give me a look at the P&L’s?’ We were able to give them a look within 24 hours. All of this data exists, but having it at your fingertips means business managers are making better decisions.”
As long as there are mergers, acquisitions, system upgrades, audits, and other disruptions to business processes, there will be a need for modeling tools to help smooth them out, Hewitt argues. “There’s been a lot of people saying that we need to make all transaction systems the same,” he says. “That’s just foolish; it will never happen. Even if you had one system that ran everything from soup to nuts, as soon as you bought another company you’d be working off their system and you’d have to deal with external data in a different format. Our system allows you to harmonize all of that data into a consistent language that business people can understand. They don’t want to know about table joins and cardinality and all of those things behind the screen—any more than you want to know what happens to a cow in a meat processing plant.”
Or to the barley, hops, and yeast in a brewery, for that matter. We just want the beer—and the data on who’s drinking it.
