A smarter approach to business intelligence

The trick to prioritising investments and activity in the area of BI is to go back to basics and consider what is important to your business

It is very easy to be blinded by the science and sizzle of business intelligence (BI) and analytics. The technology advancements in this space over recent years ­– in-memory databases, real-time analytics engines, “big data” management and analysis – are enough to get even the deepest space techies fired up. And there is no shortage of highly visual tools such as cockpits, dashboards and other interactive front ends to get business people excited too. BI is even moving into the cloud, with vendors such as IBM and SAP providing data analytics capability as an online service, and you can’t get more up to date than that.

The difficulty with all this, though, is a tendency to get so wrapped up in the technology and “solutions” that we lose sight of the original problems we were trying to solve. Some IT vendors don’t help here when they attempt to distort our view of the universe to one in which everything depends on solving the specific problem with which they can help. This is a common trait among smaller, niche players, but even product specialists in larger vendors with more comprehensive portfolios can exhibit a rather blinkered view of the world.

The trick to prioritising investments and activity in the area of BI is to go back to basics and consider what is important to your business. At one level, this is about identifying what represents a “successful” or “good” outcome from a business activity point of view, then figuring out the metrics you would use to measure that. We are talking here, of course, about defining key performance indicators (KPIs), which may range from financial measures such as revenue, cost and margin, through to more operational metrics such as process cycle times, quality of output, number of exceptions and so on.

From here, there are two ways of developing your activity. The first is to focus on the job of actually measuring – that is, populating your KPIs with meaningful numbers. This is about looking at which information sources need to be accessed, and how that data needs to be processed to derive the KPIs themselves. You can then at least assess where the business is in performance terms, and with knowledge of how KPIs relate to each other, move on to predict and model future performance. For example, a poor metric on quality will probably lead to a poor metric on revenue down the line.

There is a great deal of technology and know-how that can help in this area, but it is all geared towards telling you what has happened or what might happen – it does not help you to make things happen.

Furthermore, it is down to you to specify what is important. Industry models and maps provided by some vendors as templates can help, but you need to make sure that the measurements, weightings and dependencies are meaningful in your business environment.

The other area of development addresses the question of how to make things happen in a more efficient and effective manner. We need to look at BI from a different perspective here. The problem we are trying to deal with is how information can be used during business execution to either drive a better result or allow the same result to be achieved more efficiently. In a sales situation, this might be introducing propensity-related intelligence into the sales process, which is a grand way of saying that if you have a customer that meets a certain profile and they buy product X, for example, then offer them product Y because there is a good chance they will buy that too.

Whether it is enabling this kind of cross-sell, up-sell activity, providing guidance on how much discount it makes sense to give, or whether or not to ship a replacement product in a customer service scenario, decisions such as this are being made on the front line of businesses on a continuous basis. Individually, they might not amount to much, but add up the benefit of driving improvements at this operational decision-making level and the cumulative impact can be huge.

Identifying where a difference can be made need not require some expensive, top-down business modelling exercise – simply get out there and ask people on the front line in key areas of the business what information would help them become more effective or efficient at their jobs.

So, while we don’t want to undermine the value of some of the great technology that exists and is emerging in the BI and analytics arena, we would encourage business and IT people to make sure they know what is important to them before funding that next major BI investment. It may be that a simple integration exercise would drive a much greater and more rapid return.

Dale Vile is managing director at Freeform Dynamics