When I studied engineering, one of the most embarrassing mistakes an aspiring engineer could make was to derive a dimension from the wrong ‘base’.
Consider these two blueprints, related to the same object:
Both objects rest on a base and they need to be manufactured according to the dimensions shown on the blueprint. At a glance, both drawings appear the same. Yet, if you submitted the drawing on the left, you would have failed the subject.
Why? In the real world absolute values don’t exist. For example, the machine tool needed to make the objects operates with limited accuracy – it can cut too deep by one millimetre or too shallow by two millimetres. The blueprint should have reflected this, as shown below.
If you manufacture the object using the left drawing, the lower surface will be somewhere between 399 and 402 millimetres. The top surface will be somewhere between 698 and 704.
However, if you manufacture the object using the drawing on the right, the lower surface will still be between 399 and 402 millimetres but the top one will be more precise – somewhere between 699 and 702. The better outcome comes from using the common base, a single source of information. The top surface in the drawing on the left relies on secondary information.
But, things can get dramatically worse. While relying on derived information can be costly, the errors remain relatively contained. The biggest problems come from working off an unreliable base.
Many years ago I read about a factory worker who passed by a jewellery store each morning, where he stopped to adjust his watch to match the big clock in the front window.
One day the store’s owner asked the worker why he stopped to adjust his watch routinely. The worker explained that he was responsible for blowing the factory whistle at the end of the day and he wanted it to be as accurate as possible. The shop owner was flabbergasted. “For years I’ve been setting my main clock by your whistle…”
Many organisations suffer from the same predicament, effectively engaging in perpetuating corporate mythology. Information gets moved around loosely throughout the enterprise, from systems into spreadsheets and back again. The data becomes increasingly distorted.
Anyone who uses such numbers gets a false sense of diligence in their decision making, not realising that they have fallen victim to the computer-related disease described in 1963 as ‘garbage in, garbage out’.
Avoiding this scenario demands a single source of all your corporate data. To achieve this you need to capture all business transactions within a single database. All management and executive data must then be diligently derived from the ‘single source of truth’.
Sounds simple, but once your data becomes fragmented and repeated in business intelligence and other satellite systems (including spreadsheets), you face a truly challenging task – trying to stop further proliferation of your massaged data and developing a strategy to turn back the clock and eliminate the duplication.
Too many retailers still find themselves entangled in a complex technology web of unrelated systems, bolted together with unreliable and costly interfaces. These retailers simply cannot trust their data, for example, a recent study by Retail Systems Research found that two-thirds of retailers have inaccurate inventory records for their stores.
Without a single corporate database, no amount of money spent on technology can keep your data intact – like a house built on shifting sand, it will demand constant repair. To quote Bill Gates, “The first rule of any technology used in a business is that automation applied to an efficient operation will magnify the efficiency. The second is that automation applied to an inefficient operation will magnify the inefficiency.”