Pareto Rule, 80-20 distribution and what it can mean for you

Pareto Rule, the 80 / 20 distribution.

I love this one because it eliminates half the work. Let me give you some examples:…. But first let’s try a definition or two:

In 1893 Pareto was appointed as a lecturer in economics at the University of Lausanne in Switzerland. In 1906 he made the famous observation that twenty per cent of the population owned eighty per cent of the property in Italy, later generalised by Joseph M. Juran and others into the so-called Pareto principle (also termed the 80-20 rule) and generalised further to the concept of a Pareto distribution.

Indeed I think if he’d delved deeper or presented the information in a different way he would have probably ruled that 5% of the population owned 40% of the land or 10% of the population owned 60%.

For me, I think it’s similar to the standard bell curve, we’re looking for the results outside twice the standard deviation. In our case we’re looking at the top end, whereas at the bottom end none of the population ‘owns any land.’

One of the examples I spotted on Google relates to 80% of all the work on your computer relating to only 20% of the files!! You really can use it anywhere!

So back to my examples


a. Purchase invoices

In one company the Purchase Ledger team were hugely over worked and suppliers weren’t getting paid, so the level of phone calls demanding money rocketed, so the team were even more over-worked.

To see if we could reduce this problem I looked at the value of the purchase invoices being processed and found that 90% of the invoice accounted for just 10% of the value – which meant that they accounted for 90% of the over-working. They were all under £300 (pdv*) in value.

Incidentally the suppliers were pretty unhappy and the engineering managers were suffering because none of the suppliers wanted to supply them!

There were about ten engineering managers, so I organised them their own individual bank accounts and their own cheque books (all printed ‘amount not to exceed £300’.) They were delighted as were their suppliers and as was the purchase ledger team, the only ones upset were the central purchasing department who suddenly faced a work-load reduction of 90%. Oops!

b. Sales Invoices - Design activity

In a printing and packing company, we’d been taken over, had moved into huge new premises and had purchased two beautiful, huge and powerful four-colour printing machines. So the need was to get sales orders to use this capacity.

We needed to change the product mix and to free up the design team so that they could work to get these larger contracts.

A quick check of the sales invoices showed that 80% of the invoices accounted for only 20% of the sales value and were each less than £3,000 (pdv). We decided to risk losing 20% of our turnover by doubling the price of these jobs.

What we were surprised and delighted about was that half the customers accepted the new prices. This meant that that half paid twice as much and so we did not lose any of our sales turnover.

40% of our design capacity became available to win the new contracts.

c. product lines

This example is actually the other way round. At the popular end of the fashion industry, the company was having trouble with expanding too fast and it was obvious that the people and premises were going to be too few and too small within a year.

One of the problems was in the layout of the stock in the warehouse and the ‘picking and packing’ of customers orders. I did the Pareto analysis work and found that every product had virtually the same level of demand (What a compliment to our designer.) This even spread of demand meant I could arrange the product lines in a logical order for ‘picking’ without having to work out any complex algorithms. Phew!

We were able to collect the information from all the sales outlets and with a weekly computer bureau run, process all the sales orders so that on Friday the lorries were loaded for the Scottish and Cornwall runs (leaving on Mondays) and shorter distance ones were loaded later.

The ‘Picking and Packing’ was simplified and errors reduced, the warehouse staff were equipped with trolleys and because orders were picked in the most appropriate sequence we had more warehouse space available.

Even with doubling sales every year for seven years the original premises were able to handle the extra through-put. The directors and shareholders loved it.

The final benefit was that the sales teams also loved it because there was an immediate consistent pattern of when they could expect their next delivery and ….. therefore when they received their commission. Even the customers loved it because they could confidently rely on the forecast delivery times.

d. customers

Every company needs to rank its customers. It’s a simple process just list them off in MS EXCEL showing the sales value for the last year and sort the data by value into descending order.

Broadly chop the list into three segments. The top segment will broadly be your top 20% accounting for 80% of the sales, etc. The bottom 80% segment should be chopped into two on some almost arbitrary basis.

Now develop different strategies for each of the three segments, ‘A’, ‘B’, and ‘C’.

For example:

‘A’ customers,
fiercely defend these customers, love them, they are very expensive to replace, get to know everyone in their team as your key man might move, find out what they want your product or service to do in the future, ask them how you could improve your service or product, ask them what niggles they have (at all levels in their company). What other stratagems can you find? We could give you a report at the start of every month the ‘A’ customers who hadn’t received a visit during the previous month. Think laterally, look sideways, what about you organise a five–a-side football match with them – for charity (and then do a joint press-release)

‘B’ customers,
some of these need to be your next ‘A’ customers, what other products and services can you supply them, What stratagems can you devise to help your client?

‘C’ customers,
do you really want them, what else can you sell them, can you put up your prices, they need to become ‘B’s or be disposed of. But wait …could you find another route to market for them? If you could take out the selling costs and automate the sales process then you may have created a new division.

New customers have a very high cost of acquisition. For years Amazon were running at $75 per new customer and that was just for books with a very low margin. IBM used to budget $500,000 (pdv) for getting a major contract.

e. Works Orders

Works Orders are the internal orders given to sections or departments for the work to be done prior to them being sold.

This instance is about an engineering company who made hand tools. In engineering there used to be a ‘rule of thumb’ to allow one week for every operation in the manufacturing process and some of their products had seventeen processes.

The production team had previously noticed that works orders were taking longer to get through the factory so they had put more works orders into the system. When I arrived there were enough works orders on the factory floor to cover five years manufacture!

So that’s twenty times the raw materials, twenty times the number of tote trays and pallets, twenty times the difficulty of finding the right one to work on. The paths were blocked, the factory floor was filthy because it couldn’t be swept or cleaned. The inventory was huge.

Pareto Distribution analysis of the previous year’s sales and of the current year’s sales orders. (There is a fairly common trap of basing forecasts on actual sales (because it’s easy) rather than on sales orders that is on the actual demand. The key difference arises in that you can’t sell something because you don’t have it… and therefore it has not been included in the sales figures.)

First job was to scrutinise the current works orders and to schedule them by forecast demand. The rest were all put into an area, which became known as ‘the cemetery’, as they were not going to be needed for upto five years.

There was a Board battle of how to select the most important products, the sales director wanted to choose by looking after his favourite customers, the rest of the Board wanted the selection based on Gross Margin. No chance for ‘C’ products based on their Gross Profit so none got made for quite a time.

The Sales Director was actually delighted because the output soared as the factory was tidy, clean and the employees could see that the management team knew what they were doing.

With the increased output the sales figures also soared and more of his favourite customers were delighted with the level of deliveries.

The Board were delighted with the hugely increased Gross Margin.

f. The domestic appliance engineers

With Pareto you should have a reasonably-sized population, though as we are looking at all the members of the population we do have a statistically valid number. With the engineers we had over a thousand of them.

We identified key features of their work every week eg sales value generated, ratio of re-work and call-backs, cost of materials used, number of calls completed, miles driven and created an algorithm to give us one number to represent successful effort.

Then we did a Pareto analysis to find out who were the top performers so that we could work out what they did and therefore how to motivate and train others to their standards.

In doing this we also found the people at the other end of the scale! When they were asked to come in so that we could discuss the situation, they all resigned!

……. But that also meant that there was a new list of people at that end of the scale.