1 Why we need benchmarking
A survey of eight members of a trade association, all broadly similar businesses. The eight
businesses paid anything from £2,575 to £7,000 for IT maintenance. This begs the question - why
should it not be possible for the business that paid £7,000 to get a better deal?
Details of productivity in a GM car plant in the US, compared with a Toyota plant in Japan. The
Toyota plant took approximately half the time to build a car. It also did so in less space and with far
fewer defects in production.
The details in example 2 come from 'The Machine that Changed the World' the book of a MIT
survey of car manufacturers from around the world. This book helped encourage US and UK
manufacturers to become more efficient and close the performance gap with Toyota.
Both these are examples of why we need benchmarking.
Benchmarking can be defined as:
'a process by which a firm identifies critical activities within its organisation, compares its
performance in carrying out these activities with best practice, and implements changes in order to
close any performance gap'.
Key phrases are:
2 What to benchmark
Within any organisation, there are a whole set of activities that create value for the customer,
Inbound logistics - eg delivery of raw materials to the factory
Operations - eg the manufacture of goods
Outbound logistics - eg delivery of finished goods to the customer
Any of these can be benchmarked.
Equally, the organisation interfaces with a number of different stakeholders, each of which has
expectations of the organisation that can be benchmarked.
Customers - eg product quality, service standards
Investors - eg return on investment
Staff - eg working conditions, job satisfaction.
In practice, benchmarking focuses on critical activities. By critical, we mean activities that must be
carried out successfully if the business is to realise its strategic goals.
There are other important criteria as well in selecting areas for benchmarking:
Potential must exist for improvement - in the example given before, benchmarking obviously
had less value for Toyota than for GM.
Staff must be positive and eager to improve - because it is only through their efforts than
change will come about.
Relevant data is available or can be obtained - before MIT's study, there was anecdotal
evidence about the inefficiency of US car manufacturers, but the data that study generated
made it possible to make realistic comparisons between US and Japanese manufacturers.
Typically, the kind of areas that are selected for benchmarking are these.
Purchasing - as in example 1 above. Businesses can often get better deals from suppliers, and
benchmarking provides a way of showing what is achievable. Benchmarks can be set both for (a)
price and (b) product / service quality.
Customer service - in competitive markets, businesses succeed or fail depending on how
successful they are in satisfying the customer. Businesses need to understand what the
benchmark is for customer service and strive to achieve it. The elements of customer service that
can be benchmarked include:
Speed in turning around orders
Product / service quality.
Financial - Businesses can benchmark how efficient they are at running their finances. Specifically,
they might consider cash management - the ability to generate cash as quickly as possible.
Benchmarks can be set for the number of days stock is sitting in the warehouse before being
shipped out to customers (days in stock), and for the number of days from shipment to the
customer to payment (days in debtors). The fewer days between purchasing stock and getting
cash in from the customer, the better.
3 Types of benchmark
Benchmarking is about comparing performance. We can compare our performance with a number
Internal - within the business itself, it may be possible to make comparisons, for example between
External - it is more usual to go outside for the comparative data, ie to other businesses.
External benchmarks also encompass:
Best practice - ideally, the point of comparison should be 'best practice', because that is what we
are aspiring to. If it is possible to identify what is best practice in the sector or for the activity, this
should be the point of comparison.
Generic - Some activities are the same in every business, so it is possible to talk about generic
benchmarks. An example would be borrowing money - rates of interest are the same throughout
the UK, so we can all use the same generic benchmark, which is base rate. A business that is
getting a good deal from its bank might be paying 0.5% over base - a business that is not doing so
well might be paying 3% over base.
Customer - Some benchmarks may be laid down by the customer. For example, the customer
might insist on receiving delivery within 7 days of placing the order.
There are advantages and disadvantages to both internal and external benchmarks.
Internal benchmarks have the advantages of:
A common language - if one department talks about 'operating profit', the chances are that another
department will use the same definition.
Data easily accessible - within a business, there are no reasons why data should be withheld.
Quick payoff - because the data is available and relatively easy to use, we can identify any
performance gap quite quickly and take appropriate action. This ensures a rapid payoff.
Internal benchmarks have the following disadvantages:
Complacency - Within an organisation, the variation in performance may not be that great. This
may lead the organisation to feel that there are no major problems to be remedied. But if they were
to look outside, they might find that other organisations had made a quantum leap and were much
more efficient. Example 2 provides an analogy: if US car manufacturers had gone no further than
different plants within the same company, they would never have known how much more efficient it
was possible to be.
Inadequate payoff - Following on from the above, major benefits may not be achievable from simply
using internal benchmarks. They are not sufficiently far in advance of what the organisation is
No external focus - To get the most from benchmarking, organisations have to cast the net as
widely as possible in order to obtain comparative data.
External benchmarks have the advantages that they are more likely to reveal best practice, and
accordingly they can offer a bigger payoff.
However, the disadvantages are that it may not be possible to obtain data. Other organisations,
particularly competitors, will not be prepared to share information, for fear of losing a competitive
Even when data has been obtained, it may not be possible to interpret it. Going back to example 1,
which took IT maintenance costs for different companies, we do not know for certain that the costs
related to the same items of expenditure. For example, in one company 'IT maintenance' could
have been treated as including an on-line help facility, whereas in others this would not be included
within IT maintenance.
Finally, sharing data with other companies contains the risk that information that gives a
competitive advantage could be disclosed.
4 The benchmarking process
The benchmarking process includes the following steps:
Analysis and feedback
Select area(s) for benchmarking - using the criteria already referred to, ie critical activities, potential
exists for improvement, staff are positive and eager to improve, and relevant data is available or
can be obtained.
Obtain staff buy-in - their co-operation is essential. Fears about (for example) losing their jobs
need to be addressed.
Ensure understanding of the activity to be benchmarked - For instance, using our IT maintenance
example, what does 'IT maintenance' cover?
Assess activity - There may be some obvious points to be made about the activity even before we
benchmark it formally. Is the system currently working? If not, it may be worth re-engineering it
before we even think about trying to compare with best practice.
Define performance measures - Benchmarking contains at its heart the idea of performance
measurement. But what measures should we use? Financial measures are relatively
straightforward to define, but what about measures of quality? Many organisations use formal
quality standards as a way of defining what is meant by quality.
Identify potential benchmarking partners - If we are seeking external benchmarks, we need to
identify partners from outside the business. These will not necessarily be competitors, but their
activities must be similar enough to ours to make the benchmarking process viable.
Example 1 at the beginning of these notes shows how benchmarking partners can be identified - all
the companies in that example belonged to a trade association for their industry. All had an interest
in minimising costs, and - whilst they would not have been prepared to share information about
their core business activities - they were prepared to share information about purchasing,
Agree benchmarking methodology - How is the data to be gathered? How can we ensure it is
comparable? Who will be responsible for collating it? All of these questions need to be addressed
at the outset.
There are various ways of collecting benchmarking data.
In-house research - relevant for internal benchmarks.
Third party research - for example, employing a consultant to visit benchmarking partners and
collect data on our behalf.
Direct exchange of data - this is probably the best method. The data can be gathered either by
sending respondents a questionnaire or by carrying out a telephone survey. In both cases, care
must be taken in framing the questions in such a way as to elicit meaningful responses.
Site visits - Good for getting a 'feel' for another business, but needs to be linked to quantitative data
gathering. 'Industrial tourism' is a waste of time.
Analysis and feedback
Ensure that you are comparing like with like (Womack, Jones and Roos had to adjust their
figures in ‘The Machine that Changed the World’ for space taken to build a car to reflect the fact
that Japanese cars are smaller than US cars).
Ensure the data is in a consistent format (all GBP, not a mixture of currencies).
Ensure that comparisons are realistic. All the companies in example 1 had similar IT networks
and all were small businesses with no full-time IT staff, so the comparison between them was
Having gathered the data, careful thought must be given as to how it is presented. It is useful to
present it graphically - ie in pictorial form.
Finally an implementation plan is needed to set out how we are to move from where we are now to
Implementation consists of the following steps:
Start by identifying the performance gap. This shows what we have to do, so provides a model
for the future. So long as the gap is not too great, it also provides an incentive to improve.
Define the individual stages needed to achieve the plan.
Ensure adequate resources are available. This may not be something that existing staff can fit
around their normal jobs.
Communicate with staff. Essential to ensure continued co-operation and make closing the
performance gap a shared goal.
Finally, progress against the implementation plan needs to be reviewed. Also, from time to time,
the assumptions underlying the implementation plan need to be reassessed:
Is what we thought was best practice still best practice?
Are we using the right performance measures?
Are we focussing on the right areas?