Being intelligent about your business?

Manufacturers, their suppliers and distributors are realising they need to collect, analyse and distribute increasing amounts of data to measure how well they’re doing. For many, this data-management task – business intelligence (BI) – is beyond them. Arthur Parker, president of BI software developer Sagent Europe, believes this is why the market for BI automation software is booming. But though Sagent beat rivals Brio, IBM, Cognos, Infomatica, and Business Objects in a recent bid for BI for a German bank, his biggest competitor, he says, “is the company that thinks it can do the job itself using a code sheet.” Two years ago Robert Bosch Great Britain (RBGB) had the same insight. Its Denham, West London centre distributes automotive products and power tools, and provides sales and support for the supply of automotive electronics and systems to the car industry and of packaging systems to manufacturers. RBGB project manager Beverley Daniels says the company had been talking about adopting ‘policy deployment’ for some time. In 1999 it brought in consultant KPMG for advice and under its recommendation of ‘balanced scorecard’ began to devise the strategic objectives and critical success factors (CSFs) it would need to measure the business. “The problem we have here,” says Daniels, “is that Denham covers more than one division. So we have to have a corporate strategy that covers different divisions’ objectives.” Following change of project managers, Daniels says the adoption process was happening piecemeal in different parts of the operation. By the time RBGB did begin to co-ordinate it in 2000, she says, “We found it had so much data that data administration was becoming very difficult to handle,” and threatened to take over the whole scorecard project. “That’s when we looked for software to support the data collation side of it.” RBGB decided on CorVu’s CorManage balanced scorecard software. Adopting the software and linking it into RBGB’s intranet was relatively easy. But, that over, more problems emerged: “It wasn’t all very meaningful,” she says. RBGB thought it knew the CSFs and key performance indicators (KPIs) it wanted. Setting these measures sounds easy but it isn’t. Group managers can set goals for the entire business to cascade downwards. But managers in the business units have different issues to address. Says Daniels: “In certain areas market share is critical to success. In other areas it’s not market share, it’s profitability. It depends what your competitors are doing. You really need to do it both ways, top down then bottom up and check that they actually meet. It took us a fair time to get to that point,” she adds. Definition is key One problem was inconsistency with Bosch’s standard reporting method: “So we had to change a lot of our measures around, and the discipline behind the measures: ‘what should the definition be?’, ‘what format should it be in?’ ‘what standard plan should we use?’, ‘best?’, ‘worst?’ and so on.” Daniels estimates RBGB lost about three months when it had to go back to its senior management and ask for a list defining exactly what each of the measures should look like: “Then we had to do some rebuilding of the scorecards.” Most BI applications collect data from customer relationship management (CRM) or enterprise management (ERP) systems. Mark Mahara, managing director and VP EMEA for BI developer Silvon Software, says, “ERP can get the data but can’t do anything with it.” Businesses need to measure their quality, responsiveness and customer service, and ERP won’t provide this information without further processing. Many systems only yield the information users need if the IT department can keep up with a queue of report requests. Each report is processed individually, even though most of the requests are for reports nearly identical to others in the queue or to reports done earlier. UK Safety Group (USG), which makes Tuf and other industrial footwear, faced such problems in 1998. The shoe industry has been hammered by imports since the enactment of 1992’s pan-European manufacturing standard EN345. Imports were then 3%. Now they’re 55% and rising, and the UK workforce has been cut by 80%. Earlier this year the Tomkins automotive group put USG’s main rival, Totectors, up for sale. USG bought it for £20 million. Three years ago USG had already decided to install data warehousing across all its divisions. USG uses Protex, Datel’s clothing-industry ERP system. USG’s IT director, Tony Wood, says: “We had a lot of useful data in Protex but found it difficult to access in a form that was readily digestible.” His short-staffed department either wrote a lot of AS/400 queries themselves or paid Datel to write special report programs. In May 1999 USG installed Silvon’s DataTracker, since upgraded. Wood says DataTracker presents the information quickly, allows users to define what they want without having to explain it to IT staff, and takes queries off line, leaving USG’s AS/400 to do other things. Now, Wood explains, sales managers give their requests for reports and analysis to the marketing manager. USG now produces, in no extra time, more reports containing better business information for a variety of users, including the chief executive, John Newman. He says: “I now know what is being sold where, why and by whom. I can relate profitability to product, style, customer and territory. I can identify and exploit opportunities for cross selling between workwear and footwear.” “Most companies have no central repository,” Mahara points out, but adds that, wherever the data is collected, it is likely to need cleaning: “As soon as people get data together they find it’s not particularly good.” Costings or other information may be wrong, and they’ve made incorrect assumptions about the margins from different products: “Typically, people say ‘We didn’t realise that three products were doing the same things at different cost margins’.” Sagent’s technical director, Mike Lee, says the same: “The biggest issue is the data quality. That’s the really difficult piece. You think this part number is unique, and we can see that it isn’t. When you’re combining data from two systems – ‘month end revenue’ can mean the real month end, or the last Friday in the month.” Daniels encountered similar issues: “We had some reporting net sales in a percentage and some reporting in an absolute, so it wasn’t comparable. And it was only when we started to bring it into the software that it became obvious that of course you need to make it comparable.” Sagent is targeting the manufacturing market aware, says Lee, that manufacturers tend to need to collect and analyse data from several applications, and that they “have very low budgets for IT.” Joined up manufacturing According to Parker, the entry-level price for a Sagent system is £30,000, providing most of the functionality a small company might need: “In a simple organisation, with not a lot of data coming in from a single source, you can do everything we sell for that £30k. That is, collect data from existing systems, do simple analysis and present the data in web form. In a big organisation, it’s much more complex. You have to be able to collect data sources and feed them into a single data mart.” Sagent claims to interface with “every source of data in the marketplace”, but though it can interface to SAP ERP systems, Parker says only that it is “looking at the others”. Daniels says that, apart from a dispute about the measurement of answered phone calls, all RBGB data was “pretty clean… although some people would like to think it wasn’t. Now all of our departments have scorecards, and all our data in there is being updated monthly.” All staff can see the monthly data, and it’s being used in department meetings and linked to department heads’ individual objectives. RBGB doesn’t intend to measure individual employees’ performance by scorecard entirely, but is “looking at all possibilities”. Nor is the company yet using the system to improve sales forecasts. “What’s important for us is to have a management tool that allows us to drill down into base data, particularly sales data, so that we can provide our sales force, our product managers, with very detailed, up to date information.” Managers can now see yesterday’s sales figures. The information is presented in a way that’s easy to understand and those who see it can do something about what they see: “It really allows people to manage their business better.” A main working-level benefit will be that managers no longer have to ask the IT department to produce the reports they need. In the future, they will be able to go to the screen and call the reports up at the click of a button. But an even bigger benefit, says Daniels, will be that everyone in each department will know what their strategic objectives are and how close they are to achieving them. The results are worthwhile in cross-department communications alone. Departments have learned, usually for the first time, what the priorities are elsewhere. Purchasing, for example, is finding it difficult to handle the number of suppliers the sales division and other departments specify for sundry supplies like pens. The problems, says Daniels, “have come flooding out”. So many, in fact, that “we’re now a bit overwhelmed”. But RBGB is tackling them one at a time: “We don’t have to do all of it at once.” The cultural impact has been “huge”, she says. Often, objections to the scorecard turned out to be objections to what the technique revealed. One department head complained that the targets the scorecard set couldn’t be met without an extra member of staff, which wouldn’t be allowed. Daniels pointed out that the technique had worked because it had focussed on the problem. Others saw it as a tool to manipulate sales and other figures. But once they see the benefits for them they welcome the change. Her advice to would-be BI adopters is, “Do not underestimate the cultural change and the effect on the perception of people. You need complete management buy-in. And you really must agree and believe in your vision, your strategic objectives and CSFs, not just on an individual basis but across company.” Daniels says the most difficult message to get across is that the company is measuring, not blaming: “There are multitudes of examples where people say ‘this affects my business but it’s not my fault.’ But of course it’s nobody’s fault. It’s just something we need to look at.”