We have all had these or similar conversations: Question: “How was production today/yesterday?” Answer: “Good!”
Follow up question: “How much did we/you produce?” Answer: “A lot!”
This conversation is not producing anything, it is just pumping air. You know where I am going with this. If you want to have more than feel-good small talk you need to quantify properly.
A better answer would be: Production was good; we produced 145 units, which is about 5% above our average output. This involves more than just randomly naming a few numbers. Having numbers which mean something needs a bit of organization. It needs basic understanding of what you are measuring. It needs an agreement amongst the organization how you measure and how to express them. With some basics, we can communicate a clearer picture of the situation.
I know I have written about this many times. I am passionate about the subject and I believe in order to have an attempt to improve anything or achieve anything you need to have facts.
Measurements are facts.
When visiting a plant, I often asked the question: How do you know, you are doing a good job?
Answers such as “if my boss does not complain…” are common. When you dig deeper for an answer you find that there is usually an objective or a goal. Many companies have not specified or documented their objectives. Intuitively they know what they want to achieve. The only good thing about not defining your objectives is that, in the end, you can adjust your goal and say you have achieved it. This is not really what you need. Of course, the answer I am pushing for is to have measurements, or even better, have a system of Key Performance Indicators.
If you do not continuously record, the human mind plays a trick on you. We typically only remember the worst, best, and last event. We are usually not able to identify the average over a defined period unless we keep records.
We live in a world of measurement. What would sports be without measurements? If you play golf or baseball, there is a scorecard or board, and you know at any given time how the players stand in an overall ranking and if they are winning. The higher the stakes, the more individual KPIs are developed. The batting average is just one of them. Imagine a hockey game with no scoreboard. After a tough game the players and the fans leave the stadium with no indication of who won. Imagine if, at the end of the month, or on a quarterly basis, the score keeper writes a report to the owners telling them how their team did. Ridiculous! This is what a lot of companies do with their business.
By not keeping a score, companies are missing out on a motivational tool. By being in a position with no beginning, no end and no objective, the employees could have the feeling of being on a hamster wheel. Having a scorecard allows them to participate and see the bigger picture. The president of a company once told me that he started to post the daily volume per order picker. The volume per person shot up by double digits within the first month. “We thought you were not interested…” was one of the employee’s responses when he was asked for the reason.
I have seen many companies collect data. Data collection is often automated and emailed to the boss almost in real time. Every day the data is accumulated for the management team and discussed at the next possible production/operations meeting.
Sounds good? Not quite!
Too often I see two key elements missing:
First, the workers/operators must own the data. The data is a reflection of their work and they should see it as it is collected. They need to see it, understand it, (be able to relate to the units of measure) and trust the number’s accuracy.
Second, more important than the (daily) data is the average and the trend.
For example, if you collect daily production information this will fluctuate up and down caused by so many different factors (statistical noise). As you accumulate it weekly and monthly the true average performance becomes visible. If it improves (or deteriorates) over several consecutive months you have a trend.
A performance improvement of one or two per cent per month is not visible without a measurement. For example, producing 101 or 102 units per day instead of the usual 100 is less than the statistical noise. However, a one to two per cent improvement per month could add up to a 12-24% improvement per year. Only when you measure/keep score will you see small but statistically significant changes. The trend is more important than the actual numbers.
An additional factor is that as soon as we set a goal and set a method of measurement, we give the team an opportunity to drive towards the goal. As an old poet said: “Only he who knows his destination finds the way!”
Setting goals and setting up how you monitor the performance will influence the direction of the company. There are many performance criteria you can measure. There are so many that you can go overboard and have too many. You need to find your own ‘sweet spot.’ My suggestion is to start small and expand as required. You should select only what is important for the customer and for the company. If, for example, you always ship your orders on time and your customers are not particularly interested in on-time performance, then on-time data is not important. If, on the other hand, being on-time is rather weak and your customers pressure you for more on-time shipments, the on-time performance is the right thing to measure.
What you measure gets attention ... what gets attention will improve.
Look at your organization and establish a few things that are not quite right. What keeps you awake at night?
These can be company-wide issues, if you are the owner or key manager, or a departmental issue in the area you supervise. You do not have to wait for top management to establish a score. A supervisor/team lead can start to keep score in her/his department. Of course, an overall system would be better, but if a local score card is possible, do it.
Quantify the current situation and the ideal target (standard) and you have established the measurement.
You can measure different things with almost unrestricted units of measure:
• Volume per time (cabinets, pieces, linear feet.... per day, shift, hour)
• Productivity (output vs. input) (volume per worker, volume per labour hour)
• Quality (reject rate, cost of quality %)
• Safety (number of incidents, near misses, cost)
• On-time performance (i.e. % of orders shipped complete on the original promise date?)How much measuring is enough?
Keep it to a minimum, is the short answer.
Any measurement costs something to collect, to tabulate, to analyze and to archive. If you do nothing with the data, it is a waste of resources. If you already have sufficient control and a deep understanding of your business/operation, you might not need many or even any additional measures. Visualize results
If you collect data daily, you need to process and display it, but you also need to calculate and display averages over a larger time frame. Weekly, bi-weekly or monthly measures are usually the best practical reporting cycles. More important than the actual measurement is the trend over a period. With whom do you share this data? Good measures and performance indicators should be shared with a bigger group. Everyone in a group should understand what the goals are and how their group is performing. For this reason, you might want to create different levels of reporting:Shop floor:
This can be volume, quality, on-time measurements and safetySupervisor/department managers:
Productivity, overtime per week, material yields Management:
Financial performance, labour cost productivity
This reporting should become a part of the daily, weekly and monthly work routines. The measures and the terminology must go hand in hand.Continuous improvement
The company strategy and the goals for continuous improvement projects need to be reflected in the measurements, and more importantly, they need to show up as measurable improvements. If you think you are improving something and it does not show up in the results - you have a problem. Either the improvement action is not working, or the measuring is wrong. Both these results would be a critical indicator.
Starting a project of implementing key performance indicators (KPI) is not a short-term project. It should be considered a long-term project. Too many attempts at starting a program get cancelled and/or run out of steam at the first economic challenge. Therefore, start small with few measures, and build on the successes of those. It is easy to go bigger later.
By implementing these KPI systems, productivity gains of 10-15 per cent can be made within the first year of implementation. This gain is not due to the measurement or because the progress is being charted. The difference is that focus is being put on what is declared as critical. This focus allows the organization to align itself to the company objectives. This alignment and focus takes the slack out of the system. Once the slack is out of the system, it becomes very clear where there is a bottleneck. The measurements provide the communication tool for management and employees alike. And by having a common language and objective, it is much easier to address the next critical steps.
It takes management commitment to get a KPI system started and sustained. This system may well become the best management tool you ever implemented.