Toronto-based business intelligence consulting company SQL Power Group Inc. wanted a BI interface easy enough for end-users, but one that didn’t leave experienced SQL users frustrated with its lack of power.
The company has released Wabit, a drag ‘n’ drop visual query builder that automates joins and report generation.
“You can do almost everything you can do in SQL with the visual query,” said Jonathan Fuerth, who goes by the inside-joke title of director of global variables (he’s the manager of software development for SQL Power Group).
Users logging into the free tool specify how Wabit is to connect to the database. They then drag the tables they need into a design tool called the “Playpen.” As multiple tables are dragged aboard, the interface performs the joins automatically based on the available metadata, Fuerth said. If the metadata is incomplete or if a skilled SQL user wants to customize the query, he or she can toggle to a screen where the code can be edited manually.
That feature is helpful when the query runs up against the visual tool’s limitations – it can’t perform unions of queries, for example – or to optimize queries, he said. For most purposes, though, even experienced SQL users will find it more efficient to use the Playpen.
“We still find it quite quick to use the query builder as opposed to writing it by hand,” Fuerth said.
Fuerth said the program updates filtering, sorting and join changes in near-real time for previewing with the default row limit of 1,000. Larger data sets might cause a couple of seconds of latency, he said.
Wabit works with Microsoft Corp.’s SQL Server, Sun Microsystems Inc.’s open source MySQL, Oracle Corp., IBM Corp. DB2 and the open source PostgreSQL.
SQL Power isn’t charging for Wabit for a couple of reasons, primarily that the company is first and foremost a consultancy, and it’s a tool staff uses on customer premises. “We’d be doing this anyway even if we weren’t releasing publicly,” Fuerth said.
Also, releasing it “in the wild” gets the company suggestions and feedback from environments they hadn’t considered before, he said.
A centralized enterprise server for Wabit will be available soon, which will allow “really tightly integrated collaboration.”
“That part won’t be free,” Fuerth said.
But does it make sense to roll out business intelligence tools to every Elmer Fudd in the organization?
“You have to define your business intelligence and the users actually consuming the data,” said Dylan Persaud, manager of enterprise applications research with IDC Canada Ltd. Not every end-user would be able to understand and interpret the data the way a business analyst, line manager or executive would, he said.
Someone in procurement, for example, ordering through a customer portal understands purchase orders, order dates, quantities, part numbers, customized orders and lead times. “Somebody like an end-user might not understand the intricacies,” Persaud said.
And there’s the need-to-know factor: Is this end-user going to make a worthwhile decision based on the analysis? “A picker in the warehouse doesn’t need to see how many units were produced yesterday,” Persaud said.
And more users can mean more siloes of information that might not be consistent with each other, said Bill Hostman, research vice-president with Gartner Inc.
Still, though, “we’re seeing a big interest in those sorts of tools,” Hostman said. They’re becoming especially popular in organizations that are “light on IT but heavy with ‘thinkers’” – business users who want to wring better decisions out of the data and don’t have to worry about inconsistencies in the BI gathered.
What about the danger of a user who doesn’t understand the context of the information using BI tools? “In a way, that’s what the learning process is all about,” Hostman said, but the “the tuition can be expensive” if inexperienced users are making million-dollar decisions based on a flawed understanding of BI.
Business intelligence should be a core competency in a business, not a tool, said Hostman, and BI leaders should be spending at least 25 per cent of their time coaching users. As for the simpler tools, “at some point, you’re probably going to have to learn SQL.” But that’s not the difficult part; it’s in the application of business rules, how the users slice and dice the information, what algorithms to apply, where to find that information in the first place, Hostman said.
“The use of the information is where the problem is,” Hostman said. “The distance between a SQL query and a decision is a long one.”
Persaud said IDC is predicting growth in business intelligence despite – or perhaps because of – the ongoing recession, as companies see its increased importance in wringing out efficiencies and risk mitigation, and the need to know the parameters of their businesses.
“As it stands right now, a lot of companies don’t understand the parameters of their business,” Persaud said.
