Principle 5: Improved Decisions (Item 2)

Effective use of facts, data and knowledge leads to improved decisions.

Don't gather unnecessary junk

One of the most difficult issues is what data to collect. Gathering data costs money. Storing it so you can find it again costs more money. Setting up a system to find the exact piece of data you want when you want it costs even more money. So do not gather data you will not use.

One of the greatest difficulties is to stop gathering data that you do not want. Most companies are so swamped with useless data that they cannot find the useful data on which they might be able to make a decision. The temptation is to gather everything because we do not know what we will want, so we had better gather everything. We frequently find companies that describe their policy as "we measure everything that moves". Very wasteful! Very impractical.

You have to decide in advance what you will need and do not collect stuff just because you might need it.

Unfortunately, there is a lot of risk elimination associated with data collection. People gather it just in case. Because you know that as soon as you decide not to gather some data, you will need it. Many of us have been in the situation where the boss says, "we need this piece of data, why don't you already have it".

The widespread available of electronic recording in databases has also led to unnecessary collection. Data is so easy to collect that people collect it. Shovelling huge amounts of data into electronic databases. Of course, the hope is that some day it will be useful. Many companies find that the expensive database that they built expecting it to be full of vintage wine turns out to be full of rusty old cans – if they can extract anything at all.

Measure everything? No way!

Years ago Geoff Bell caught birds as part of Australia's catch, band and release program. When you catch a bird and have it in front of you, you know you will probably never see it again. The temptation is to gather lots of information about the bird just in case someone can use it. Consequently, people measure everything they can think of: wing length, wing span, tail length, weight, head length, bill length, claw length, plumage molt. All this takes time — not only to measure but also to record and store the data. It is also stressful for the bird.

Studies indicate that all of it is useless. No one uses it. There is too much error in the data for any researcher to use it. There is more variation between individual bird catchers than there is between the birds. No piece of this measured data is of any use. What a waste!

You should have a good reason to measure something before you go to the trouble of measuring it.

Collection criteria

Companies that do this very well have collection criteria to help decide what to collect. The people who will be using the data should design the criteria and describe how they will use it. That is, the customer of the process should design it.

Collection criteria should be based on good frameworks. These frameworks are usually at six levels. More detail on some of these is given later.

• The data and information needed to conduct daily work: such as customer account data, transaction data, data collected on forms, tolerance of a part, mixture details of a batch.
• The data needed to manage those processes: such as numbers of transactions, throughput, flows on inventory, rework, timeliness, accuracy, tolerance of batches of parts; capability estimates and calculations; targets, benchmarks and comparisons with other companies. (Principle 4 `To Improve the Outcome, Improve the System')
• Stakeholder needs data: owner needs, customer needs (Principle 3), employee needs (Principle 7), community needs (Principle 9)
• Performance data for the company: such as performance against the Vision, Mission, Goals, objectives, strategies and KPIs. (Principle 2 `Focus on Achieving Results')
• Knowledge: such as experience and learning from success and failure. (Principle 8 `Innovation')
• Strategic data and information: such as market behavior analyses, competitor information, competitor behavior analyses, technology trend analyses, political trend analyses; market and currency forecasts; scenarios. (Principle 2 `Focus on Achieving Results')
• Your frameworks and the data collected using them should be reviewed periodically to ensure
• the data is used
• it is still what the users want
• if some other data should be gathered instead or as well

Don't let IT experts decide what you should collect

Unfortunately, design of collection criteria is very often left in the hands of the Information Technology specialists. This never works. Many companies fall into the trap of handing the design of the data-gathering system over to Information Technology people – because "they are the experts". This frequently happens because many people are overawed by IT experts. However, for all their expertise, your IT people are not the customers of the data. They are suppliers of an expertise. They may be experts at the technology, but they are not experts at deciding what to collect.

Your IT people are suppliers and, like all suppliers, they must supply what their customers want. The IT involvement during design should be restricted to advice on what is possible. Even then considerable effort needs to be exerted to ensure that the IT specialists deliver intuitive (you should not need a manual), user friendly systems that provide what you would have designed if you had those particular skills.

We have all seen computer systems that were so unfriendly that no one could use them (with the possible exception of the designer). Most IT departments hold their companies to ransom because they have special skills. Unfortunately, they seldom deliver what their internal customers want.

Information versus data

There is an ongoing debate about the difference between information and data. It appears to be semantics. The definitions are that `data' is raw, unprocessed material. `Information' has been processed, analyzed, interpreted or summarized. By `analyzed', we mean that some `meaning' is interpreted from the data, preferably a prediction or a cause and effect relationship, that was not apparent before the analysis. Another way of thinking about this is — information has been processed to the point it will assist your decision making.

However, this distinction between data and information is too restrictive. What you often find is that the results of one person's analysis, their `information', becomes the next person's `data'. As you go further up the decision making tree, what for one level is the final piece of information obtained after much analysis and interpretation, is just the input (together with the results of other analyses) for the next level to begin their analysis.

Although the semantics have raged for many years, the distinction between data and information has little value except to the Information Technology professionals and to show the different levels of uses the customers of the process have for data and information. So, forget it.

It is more important to know how you use the data or information for making decisions.