Old databases and spreadsheets don't usually die. Unlike old soldiers, they don't even fade away. Plestel used BI to bring its legacy data into the modern era and provide its key decision-makers with competitive intelligenceAfter 18 years operating within Telstra's small systems division, telephone systems integration company Plestel found itself in an unusual situation. It was spun off as a separate business in a joint venture with Plessey and became a brand new company encumbered with legacy information systems and business processes.
Plestel's primary activities are selling and installing the Commander telephone system. Its target market is small and medium enterprises, as well as branch offices of large corporates. With annual revenues more than $150 million, and 200,000 active customers, it is a sizeable business in its own right. However, it had accounted for only 1 per cent of Telstra's total revenues. Plestel's challenge as a separate entity was in becoming more responsive to the needs of its customers, the changing market place, and what its competitors were doing.
"We had to get a handle on the business and provide our key people with intelligence about what the market trends are; what our customer base is doing; which products are performing well; and which products are being churned by our competitors, so we could respond appropriately," explains Steve Evans, Plestel's general manager of information technology. "We are the market leader, but without that information we were going to suffer and lose market share. We inherited a plethora of sheltered workshops from Telstra, by which I mean people were extracting data from myriad different databases and Excel spreadsheets all around the country and were trying to massage that data into some form of information [without adding much value]. This was very labour-intensive and clearly had to stop."Consequently, in February this year, Plestel started looking at a number of business intelligence tools and has since gone live with PowerPlay and Impromptu from Cognos running under Windows NT. The company is also rolling out a Web-based version of PowerPlay for more casual and less resource-intensive users.
Evans admits the evaluation and selection process happened quickly, given the urgent business imperative. However, he was personally familiar with Cognos products; he considered the choice low risk; and it provided a growth path into other tools, such as Cognos' Scenario and Forethought, which Plestel also has plans to implement. In addition, Evans claims that the tools are simple to use, friendly and intuitive, and that user training takes less than a day.
Ultimately, Plestel will have more than 100 users of PowerPlay and Impromptu, consisting of senior executives and the marketing department. PowerPlay enables them to build OLAP models within its SQL Server-based corporate data warehouse and analyse trends and activities within the business. The warehouse contains information extracted from Plestel's core application, Service Plus, a legacy system it also inherited from Telstra. Impromptu allows the company to drill down and report on more detailed data underlying the trends and patterns uncovered in PowerPlay.
"Impromptu and PowerPlay work well together," Evans says. "Impromptu lets us look at individual transactions and customer records; so, for example, if we're looking at customer cancellations, PowerPlay and Impromptu will allow us to see who has cancelled and what the attributes of those people are. Later on, Scenario will provide us with intelligent characteristics we may not otherwise have detected. Further down the track, Forethought will enable us to devise a mechanism that will allow us to proactively stop those people from cancelling and so reduce our churn."Due to the need for speed, Evans says this was an IT-driven project. Plestel had pilot models up and running within six weeks of purchasing PowerPlay and Impromptu. However, in hindsight he believes it was a mistake to release the initial models so soon, as the source data from Service Plus was of questionable quality. It had migrated through a series of legacy systems within Telstra over the years, without diligent administration, and lacked validation at the data entry phase.
Consequently, the integrity of the end models was also thrown into question, as they were producing inaccurate information, and Evans and his team lost some credibility within the user community. These data quality issues, which Evans describes as the biggest hurdle to success, and now thinks he should have addressed earlier in the project, necessitated a rigorous cleansing exercise to normalise the data, make it more accessible and remove inconsistencies.
With an IT group of only 20, Evans' other challenge was to manage resourcing conflicts between this project and numerous other operational projects and demands on his staff. In addition to ongoing support, Cognos assisted in the initial design of the OLAP models, in order to get them up and running quickly, although Plestel has since taken responsibility for this in-house.
While data cleansing and ensuring its integrity will be an ongoing process, Evans believes Plestel now has meaningful customer and contracts bases and that the company is getting real value out of the Cognos tools since going live in July this year. He estimates that the initial investment was around $200,000 to $300,000, including consulting. He says it was very simple to justify the cost, on the basis that the company was being starved of the information it needed to survive. Although he can't say when payback will be, he anticipates the primary return will come from increased revenue through greater customer retention.
"By analysing trends we'll know which customers are likely to leave --such as customers in a particular industry, who have a particular type of equipment; or those who have been with us for a certain amount of time and are paying a particular monthly rental. Then we can be proactive in trying to keep them," he says.
"We're now starting to focus on the service aspects of our business, such as looking at the number of faults we get, what the attributes of those faults are versus product, and how long it takes to fix problems. We're also capturing and storing information about our competitors, so we can see what they're doing.
This enables our marketing people to devise alternative campaigns to combat that competitor activity.
"It's only a matter of time before we start seeing [these benefits] flowing through to the bottom line. I think this will change the way we operate as a business, as we'll have information available as it's happening rather than having to wait for somebody to manipulate and distribute that information," Evans says. "Our corporate objective is to generate information as a by-product of the business. We want to have a common source of information and eliminate inter-system dependencies. And we want decision-makers using that information, rather than have people just analysing data," Evans concludes.
Military intelligence is a contradiction in terms, one of the Marx Brothers once said. The cynics might argue that in some cases the same sentiment can apply to business intelligence.
Be that as it may, like most buzz terms, business intelligence can mean different things to different people. To John Murcutt, communications systems manager at Australian resources company North Ltd, it means the opportunity to evaluate and analyse information quickly and then being able to advise senior management as to its impact.
According to Murcutt, North has several underlying transactional information systems, including JD Edwards and Mincom. However, the company found that none of them were very adept at consolidating large amounts of information. This was particularly a problem for North when it went on the acquisition trail.
"We were using a multitude of spreadsheets that were very difficult to maintain and led to a lack of consistency," Murcutt says. "While they contained most of the necessary data, all the knowledge tended to reside with one person. Once that person left the organisation or moved on to something else, the spreadsheets often became unreliable, and there were instances where we didn't have confidence in the integrity of the data."Consequently, in late 1996 North embarked on a group-wide implementation of Hyperion Enterprise, a financial consolidation tool. About 100 finance people now use it across all of North's business sites, including the Robe River iron ore mine, the North Parkes gold mine and its Tasmanian forestry operations, to create consistency and standardisation in its management and statutory reporting, planning and analysis. Whenever North makes an acquisition, it is now company policy to immediately roll out Hyperion Enterprise to the new unit.
Towards the end of 1997, North also began implementing Hyperion Pillar, an analytical budgeting application. However, unlike Enterprise, North is only using Hyperion Pillar at selected sites that feel they can obtain value from it, and there are currently fewer than 50 users. "Hyperion Pillar enables us to go down to a low level in the budget. For example, in our mining operations underground managers can potentially look at the impact of tyre maintenance and what it is costing the organisation," Murcutt explains.
North looked at a number of different solutions, but Murcutt says it soon became clear in the evaluation process that Hyperion was a fair way ahead of its competitors in the functionality its products had to offer. He was also impressed with their flexibility and maturity, and the research and development that went into them gave him confidence that they would continue to be upgraded.
North initially assigned two people from its finance team to the project full-time, along with a number of people part-time from other areas of the business to identify the company's requirements. Implementation began in October 1996 and the financial consolidation tool was up and running live by February 1997. The company ran its own training courses for end users, structured over two days. According to Murcutt, the implementation was successful, with few problems arising.
Although Hyperion Enterprise was the most expensive solution North looked at, Murcutt claims that the benefits gained significantly outweigh the cost. "It allows us to collect much more information and in a more timely manner than previously and do a lot more 'what if?' analysis," he says. "The key benefit is the consistency it has given us across the business units and the ability to change structures if circumstances have changed. It allows us to test certain scenarios and mix of structures, so if we were thinking of either divesting or acquiring a business unit, we could add that to a particular structure and analyse its impact.
"It has also speeded up consolidations significantly. When results come in from the business units, we now turn around a first cut of the consolidation within a day or so, and we've been able to cut our year-end reporting timetable back by two weeks. I think all the finance people would view the implementation an outstanding success. We couldn't contemplate life without it now. The only downside, if any, is that now that we are able to deliver this information, the more management have, the more they want."According to Murcutt, the upfront investment was around $500,000. However, while nobody in the company is now questioning the cost, Murcutt did find it difficult to justify it at the outset of the project. The benefits, he says, were hard to quantify and prove in advance; ultimately, it required somewhat of a leap of faith on the part of the company, which has since paid off.
"The biggest challenge was to sell to the business units [the concept] represented a better way. Its primary purpose was as a consolidation tool for the organisation corporately, and the business units themselves didn't think they were going to get any value out of it. Two years on they see they are getting value from it by having a higher quality reporting system in their own unit.
"In hindsight I probably would have bought it 12 months earlier and tried to sell its intangible benefits more. If we could see what the intangible benefits were as we do now, it would have been simpler to cost justify," he says.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.