You’ve got masses of information, a collection of disparate information systems and no easy way to consolidate your data and turn it into business intelligence. You want to bring it all together without having your users think you are into some sort of propellerhead project that the words “business intelligence” imply. What you want is clarity to emerge from your wads of data. That was the challenge facing Forest Resources’ FIS/integration manager Bruce Talbot until recently. Now he’s well on the way to achieving his objectives. And, hey, what better than to call his project Clarity? No propellerhead stuff in that name. And why not do all the training in-house on the actual data staff are familiar with? Familiarity would aid the learning process. Clarity would aid the business.
“We had a situation where the term data rich/information poor could have been applied to us,” says Talbot. “We wanted to move away from that.” There was certainly a problem with the amount of data on the various computer systems, although its lack of integration meant it did not always match consistently. Much of the information was the bread and butter stuff of forestry — data concerning the management of forests throughout the Carter Holt Harvey empire, of which Forest Resources is an important business unit.
“What we wanted to do was make that data accessible to our end users,” says Talbot. “We had some data in our forestry system and some — basically our financial data — in SAP. Other bits were contained in Excel spreadsheets and some specialist products were recorded in Access databases. We wanted to pull it all together.”
And pull it all together they did, starting in June 2000. The man named as project manager for the job was Diarmaid O’Reilly, as Irish as they come but with a firm grounding in the type of structure required for the new systems to function satisfactorily. “After three or four months of putting the strategy together we knew it was going to work and then we selected the products (Cognos’s business intelligence tools) that we would use,” says Talbot. “Four months after we got sign-off on the project, we delivered our first subject area. And we have been delivering subject areas every three months since then.”
Carter Holt Harvey’s structure is based around multiple business units, with five making up the forestry division. These segments include the nursery component, the estate management in Forest Resources, the harvest and distribution component, the Lodestar shipping business and a smaller unit that looks after small forests — those owned by farmers and others who need help when the time comes to harvest. Talbot’s IT unit has traditionally supplied a range of services to these businesses, and the new analytical tools now mean users and management can finally get a comprehensive picture of where their investments are at. Before the new tools came into service, former chief executive Jay Goodenbour was reported as saying the system was extremely complex and difficult to use. Data wasn’t the problem. The challenge was to mine for meaningful information.
The lack of solid integration in the past meant that the business units weren’t getting an entirely accurate picture of how they were performing. The problem lay not in the final figures that were coming through but in the detail. Inconsistencies would occur in some costings that were correct in total but uneven in detail. “Those sorts of inconsistencies became clear to us as we moved the project along,” says Talbot. “People were putting in data but they weren’t getting the information out at the other end.”
O’Reilly rounds out Talbot’s summation: “The data needed to be fixed at source. At the time, those who were entering it would rather just ‘fix’ it to make it look right rather than correct the detail.”
The result of all this complexity was that business information was in danger of becoming distorted. The final cost of business in some markets was slightly off target, with resulting confusion over costs versus sales. “In some cases we weren’t making as much as we thought we were making,” says Talbot. And vice-versa, presumably.
In order to overcome these deficiencies, Talbot and his team had devised a number of tactical responses — by delving deeper into data via SQL inquiries, for example.
As Talbot describes it, the remedy involved both a top-down and bottom-up response. “We had the ability to pull data out of our current forestry system, but not efficiently. That met some of our operational needs but it didn’t meet management’s needs for data. So we approached our new system first from a top-down perspective, to take the chief executives of each business unit and their other managers to get the drive, the impetus to encourage those at a lower level. If those people at the top started questioning the data, then the people down below would ask where it came from. They would want access to the analytical tools to ensure data accuracy before the people above them started asking questions. That approach helped to drive the new system through the business. If you start at the bottom it can be a lot harder.”
As planning got under way, O’Reilly and his staff spent the first couple of months going around the business units to ask people what they wanted from the system, and what they couldn’t get at that time. Basically, he says, they were crying out for information. Everyone was going off in directions in their quest for improved data analysis. O’Reilly’s team focused on the areas that were perceived as offering the best value.
“We developed a matrix where we could look at a subject area versus cost,” says Talbot. “We grouped the information together by sales and marketing, financial or other areas, and graphed the results on our matrix to indicate what areas we should attack first. We didn’t want anything too complex. We had a team of four and we contracted others as required. During the initial phase we probably used an additional three people.”
Talbot predicted that two things would happen as the project got under way. First, people would discover data inconsistencies they had not known about. That assumption proved correct. The second prediction was that staff would identify areas of information they wanted but weren’t currently collecting. That assumption also proved true. Today, even without the business intelligence tools, Forest Resources has greater data integrity than ever before.
“Staff who were accustomed to using Microsoft Excel to make it look right at least now have an accurate data source,” says Talbot. “Everything that flows from there benefits as well.”
O’Reilly says access to the new system is simple. Users enter their own home pages via a challenge and response system. Once they are through, the information is all there for them, with all the appropriate reports and analysis derived from a single source of data. Previously, if people had done identical analysis from their own data sets, each user would have had a different set of results. It was easier to manipulate the details to make the results look right rather than to fix up the data at its source. That was all very well as far as it went — after all, the final figures were correct — but it hardly helped the sales and marketing teams, the decision support people, the engineering staff and the forest leaders as they looked at their cost structures and struggled with their sums to stay within budget.
The application of the Cognos tools had two obvious impacts right from the start. In the past staff would come to Talbot and describe how they had been struggling to make sense of data for weeks. With the new tools — the PowerPlay, Impromptu and Visualiser products — they would be shown how the information they wanted was now at their fingertips. “And then,” says Talbot, “you could say to them: ‘Do you want to see this month’s summation as well?’ It was enough to make them cry. That’s not to say they weren’t used to doing a lot of analysis in the past — they were, but it was time-consuming. Now they can access the information as they need it.”
Join the CIO New Zealand group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.