A failing business intelligence strategy at ASB bank has been put right after the bank’s IT team went back to the business to find out what users really wanted.
ASB’s information solutions manager, Peter Newey, told an audience at a Microsoft business intelligence conference in Auckland last week that the bank had become very good at accumulating data but was failing to deliver this as useful information to the business.
Newey said ASB first embarked on data warehousing and business intelligence back in 1993. The initial IT partnership with business went well, as the bank sought to build a single view of the customer. The next stage of the project, linking data to views of customer profitability, was also successful, Newey said.
However, that’s where the project stalled. Not a lot happened from then on, except that more and more data was added to the warehouse — in fact, this was mandated, he said.
“[But] it’s not about putting data in,” Newey said. “It’s about getting information out.”
Missing was a partnership between IT and the business, Newey told the audience. IT put data in the warehouse and then just stepped back, which left the linkage between the warehouse and business strategy “tenuous”.
The original concept of the data warehouse — which as a model was developed in the early 1990s — was to serve business analysts, and then to feed this information upwards to the executive levels. An “army” of TSQL coders then grew within the business and information silos began proliferating, said Newey.
Business units needed information, too, but it was more effective to make copies and keep these locally than to use the warehouse. Newey said that about 75% of the warehouse’s data was eventually duplicated.
In 2005, ASB’s business intelligence strategy took a new direction, with the need to support new corporate governance requirements such as Basel II.
“Compliance and audit are your friends when it comes to getting value from BI and data warehousing,” said Newey.
However, the new strategy struggled as well, despite a “huge team” of developers and designers being assigned to it. After nine months not a lot of progress had been made. That was because the team was far too focused on technology, Newey said. Business users were consulted once a month but there was little user buy-in to what was going on.
“That went on, in hindsight, longer than it should have,” Newey said.
It became clear the project was out of touch when the project manager reported it would “take another 18 months” — another 18 months before he could predict how long it was going to take.
It was “time for a health check”.
“Without buy-in, what are you building?” Newey asked the audience. “Your next legacy system?
“Business units have to manage business knowledge. They are a partner. Make them a partner.”
The next move for ASB was to call in Wherescape, which makes the data-warehouse lifecycle management software Wherescape RED. It also partnered with Hewlett-Packard and began prototyping development on a two-week cycle, checking back with the business all the time. The project bagan focusing on the information needed to make business decisions, and that was a much smaller set of information than expected.
The level of business engagement was such that the bank’s credit and risk manager was engaged with the project for a half a day every day for three to four months. Within four to five months the bank could claim it was meeting its compliance requirements.
ASB has long been a Microsoft shop, and uses SQL 2005 and Microsoft’s analysis and reporting services.
Newey said that, while there are still silos of information at the bank, progress is being made towards managing data as a corporate asset and delivering group-wide services.
“Focus on the things that make the biggest difference,” Newey advised. “Business intelligence implies business ownership.”