Understand business variation for improved business performance

Understand business variation for improved business performance

BRIEF CASE Understand Business Variation for Improved Business Performance David Pickton, Michael Starkey and Martin Bradford STATISTICAL PROCESSCO...

496KB Sizes 0 Downloads 57 Views

BRIEF

CASE

Understand Business Variation for Improved Business Performance David Pickton, Michael Starkey and Martin Bradford

STATISTICAL PROCESSCONTROLis a technique that can help managers and planners. To date it is most usually used in factories, but, since its value is in helping managers understand variation in performance, it has a m u c h wider application. Michael Starkey, in conjunction with industry colleagues, has recently formed an SPC User Network. He may be contacted for more information. All managers are concerned with improving performance, and analysis of current performance is a fundamental part of the business improvement process. Unfortunately, such analysis is rarely a straightforward affair and is a task that is frequently complicated by a lack of understanding of the processes involved. Judgements are made based on performance figures taken at specific (and often too infrequent) points in time; this quarter's performance compared with last, this period compared with last year's period. Typically, the measurements will vary (it w o u l d be more surprising if they did not), but the telling question is to what extent are such variations significant? Are the differences meaningful and, if so, what new course of action, if any, should be undertaken? To make accurate interpretations an understanding of variation and its causes are needed. Performance figures will always move up and down naturally over time; initiating management action based on a misunderstanding of variation in performance can misdirect effort, be inefficient and expensive. The search for a better understanding of the variations that occur in business activity introduces to us the world of Statistical Process Control (SPC). But do not switch off! This is not something you pass to the Production Manager to sort out. This is your world, or at least it should be according to an increasing number of managers who recognize the need for different approaches to handle the new challenges and Pergamon S0O24-63O1(96)OO034-9

pressures they now face. But SPC is not a new idea (although awareness of its potential in applications other than manufacturing has been exceptionally limited until recently). It was developed as a concept and technique for identifying causes of variation as long ago as the 1930s by Dr Walter Shewhart I and introduced by the renowned Dr Deming to Japanese top management from the 1950s onwards. The Japanese have since consistently applied SPC to revolutionize the quality of their manufacturing processes and products. The important points to note here, however, are that this is not another management concept being imported from Japan, its roots are strongly Western and its application lies not only in production control but also in other business areas for managers at all levels of the hierarchy. Organizations as diverse as 3M, Rover, Ford, the US Internal Revenue Service, the US Navy, the West Midlands Employment Service and the National Blood Service are all benefiting from the use of SPC techniques. It is not appropriate here to attempt a full explanation of SPC, our simple wish is to make the concept known and generate sufficient interest to encourage readers to find out more about the subject. SPC is a golden opportunity for managers but its full impact has yet to be realized in all areas of management and business improvement. It can be applied to any business process which can be measured, and that is a

Brief Case is a portfolio of commentary, opinion, research and experience. The editors welcome ideas from readers. ~ w Campbell, Marcus Alexander and Michael Goold at Ashridge

Long Range Planning, Vol. 29, No. 3, pp. 412 to 415, 1996 Copyright © 1996 Elsevier Science Ltd Printed in Great Britain. All rights reserved 0024-6301/96 $15.00+0.00

To determine the causes of variation, Shewhart constructed a statistical tool k n o w n as a 'control chart' and it is this charting system that is used in producing SPC analysis. Based on his French experiences, JeanMarie Gogue, 4 President of the Association Francaise Edward Deming, maintains that by adopting SPC and the use of control charts, CEOs and senior managers were able to cut by three-quarters the n u m b e r of corrective actions t hey previously had to undertake as part of their managerial duties, allowing them to focus on the one-quarter w h i c h was most effective. SPC analysis r e d u c e d 'tampering', released management time and focused management effort. Organizations that have begun to apply SPC in p r o d u c t i o n and nonp r o d u c t i o n processes alike have realized the opportunity that exists to make genuine i m p r o v e m e n t s to the performance of their business. Be assured, despite the term 'Statistical Process Control', the application of SPC does not require advanced mathematical skills, but an appreciation of the concept of variation is essential in order to interpret data w h i c h is presented in the form of control charts. For instance, supposing that we are studying the n u m b e r of customer complaints per week. A control chart w o u l d be drawn to show the n u m b e r of complaints on the vertical axis and the week n u m b e r on the horizontal axis. For a period of, for example, 20 weeks, the n u m b e r of complaints per week w o u l d be plotted on the chart. At this stage the chart w o u l d be merely a simple run chart (time series). The mean for the n u m b e r of complaints w oul d then be drawn as a horizontal dotted line. Upper and lower control limits D Mistake 1 - - T o react to an out c om e as if it came (drawn as solid horizontal lines) are then calculated from a special cause, w h e n actually it came from using formulae d e t e r m i n e d by Shewhart. These control limits are used to determine a 'band' of variation c o m m o n causes of variation. w h i c h would, by Shewhart's definition, be due to O Mistake 2 - - T o treat an out c om e as if it came from a c o m m o n causes--i.e, is part of the process and c o m m o n cause of variation w h e n actually it came according to the charts is entirely predictable. In this from a special cause. example the lower control limit is effectively zero as it is not possible to have less than zero complaints. Shewhart and others maintain that taking man- Our simple control chart might show that we had a agement action based on a confusion bet w e e n com- m ean of eight complaints per week and, by use of the m o n and special causes of variation (tampering) will Shewhart formulae, an u p p e r control limit of 21.5. result in making things worse rather than better. Man- The control chart is telling us, therefore, that we agement and quality consultant, Brian Joiner, 3 should expect in any week to have bet w een 0 and explains that c o m m o n causes of variation are always 21.5 complaints. Supposing in week 25 the n u m b e r present in a process, but their impact varies. Indi- of complaints rose to 23. As this n u m b e r of complaints vidually they have a small effect, collectively, exceeds the u p p e r control limit, it w o u l d indicate however, the effect may contribute to a lot of vari- the presence of a special cause of variation and that ation. Special causes are not always present and may management must find out what has happened. To appear from time to time. T hey originate from outside ignore this variant w o u l d be mistake No. 2 referred to the usual process and may make either a small or large earlier. After investigation it might be found that a contribution to the total variation. Usually they have new e m p l o y e e had been taken on and not given traina m u c h bigger impact on variation than any one com- ing in the appropriate procedures. Once the special m o n cause. Processes with only c o m m o n causes of causes have been identified and r e m o v e d from a provariation are stable a n d predictable, whereas those cess (or incorporated if they are beneficial), it is then management's job to work on reducing the c o m m o n with special causes are unpredictable.

surprisingly large number, despite our frequent reluctance to measure m a n y m a n a g e m e n t activities. Deming main tain ed that we have barely scratched the surface of its potential with 97 % of its application as yet u n t o u c h e d . Here he is specifically referring to m e a s u r emen ts related to "overall business strategy and planning and c o m p a n y wide systems such as personnel, training, pay, financial data, and purchase of materials and service etc". 2 The limit of potential applications is endless. Whereas the use of SPC has been more readily ad opt ed in the context of the shop floor (3% of processes), its w i der application, particularly in financial and strategic areas, are little k n o w n in both the academic and commercial communities. Acceptance of the pow e r of SPC by management at large is long overdue. The significance of the SPC concept lies in the recognition of two related issues. Firstly, that variation in performance, o u tp u t and quality is inevitable. Despite fervent attempts to achieve m i n i m u m variability and elimination of poor quality goods, late deliveries, unacceptable levels of service, or w h a t e v e r - - b u s i nesses will co n tin u e to experience less than perfect and variable performance with its consequential detrimental effect. Variation is a p h e n o m e n o n endemic in all business and m a n a g e m e n t activity. The second issue is a cons e que nc e of the work of Shewhart who " i n v e n t e d new ways to think about u n i fo rmity and n o n - u n i f o r m i t y " . 2 He saw two kinds of variation, those that came from c o m m o n causes and those due to special causes. He also noted two kinds of mistake:

Long Range Planning Vol. 29

June 1996

causes if the degree of variation is still d e e m e d to be unacceptable. A few words of caution are necessary, as emphasized by Professor David Kerridge. S The reader may e n c o u n t e r two quite different versions of SPC (and control charts). The first and most frequently found, especially in Europe, deals with monitoring and controlling processes and then making adjustments w h e n necessary. The second, as i n t e n d e d by Shewhart, performs all that the first does but is specifically concerned with process improvement. Dr Deming often used to say, " T h e first control chart y o u s h ou ld draw is on figures that cross the Managing Director's desk." Professor Kerridge 6 illustrates this point by referring to the case of a General Manager of a charity who applied SPC and control charts in making decisions covering a whole range of administrative and financial matters. For instance, he w o u l d draw control charts on all key items of both i ncom e and e x p e n d i t u r e from the m o n t h l y trading account and payroll. Because of the charity's legal liability, dangerous incidents were also charted. From the analysis the General Manager was able to identify that in m a n y instances mistakes were actually due to errors of data collection and not, as might have first appeared to be the case, a true variation in performance. The SPC checking resulted in a marked i m p r o v e m e n t in the quality of data p r o d u c e d and an increased confidence in the reliability of information the General Manager was using to make decisions. He knew when to intervene and when to leave well alone. This meant that he had more time • • • but when action was called for, he knew it was right to be decisive.6

Clear evidence of the seriousness with w h i c h the public sector are taking SPC can be seen in a recent publication from the National Health Service Executive. 7 Leading Improvements in Healthcare: a Resource Guide for Process Improvement was issued in November 1995 to all Regional Directors, District General Managers, FHSA General Managers and Trust Chief Executives. This v o l u m i n o u s guide contains substantial elements of SPC concepts as one of the means to achieve process improvement. There are other instances and examples of where SPC is being successfully used at top levels, but due to 'sensitivity', organizations are not prepared to be quoted here. However, in 1995 an SPC User Network was formed comprising companies who have experience in applying SPC techniques, and one of their prime objectives has been identified as "the sharing of knowledge of SPC to the benefit of members of the Network and to convince and convert others". At a recent meeting the members identified two main barriers to the effective adoption of the principles and practices of SPC. These were identified as: [3 an ignorance and a m i sunderst andi ng of SPC; [3 a lack of appreciation and support from top management. SPC cannot be seen as a 'cure all'. As Deming often stated, "there is no instant puddi ng". But the woeful lack of appreciation of SPC principles may well be holding back i m provem ent s to business performance. Application of SPC by senior management may yield significant benefits that are simply there for the taking.

References 1. W,A. Shewhart, The Economic Control of Quality of Manufactured Product, Van Norstrand, New York (1931). 2. W,E. Deming, The New Economics, for Industry, Government and Education, Massachusetts Institute of Technology Center for Advanced Engineering Study, Cambridge, MA (1993). 3. B.L. Joiner, Fourth Generation Management, McGraw-Hill, New York (1993). 4. J.M. Gogue, personal correspondence (1996). 5. D.F. Kerridge, Two Types of SPC, British Deming Association Statistics Research Group Working Paper (1995). 6. D.F. Kerridge, personal correspondence (1996). 7. NHS Executive, Leading Improvements in Healthcare: a Resource Guide for Process Improvement, NHS Executive (1995).

Brief Case--Understand Business Variation for Improved Business Performance

i ~!ii i ~i i i ~i ~i!i i!~ii i i i i¸ii!i!ii!iiiiiiiiiiiiiii!iiiii!ii!iiiiii!i!i!ii!ii i i i i i :i ~ili i ~!ii~i!!iiiiiiiiiiii!iiii!i!!i!iiii!i!iiii!!!i!i!i!!!i!ii!ii i~!i!!!ii i i!il

Michaet Stark~ is a

Long Range Planning Vol. 29

June 1996