How analysing and interpreting data achieves substantial economic gains

Introduction

We live in a world where an extraordinary amount of data is generated every day. Much of this data is simply collected, stored and sits unused. However, corporations that analyse and interpret their data can achieve substantial economic gains.

In this article, KordaMentha Forensic's data analytics specialists discuss how data analytics allowed a transport organisation to find its lost pallets and, in the process, save large sums of money.

Losing pallets

Pallets are one of the necessary evils in the transport industry. In theory, the rental of pallets should have minimal or no impact on a transport operator's bottom line - pallets are transferred onto the operator's account for the duration of the haul and then transferred off on delivery. Yet lost pallets seem to be a growing cost to transport operators in both rental and replacement cost.

In recent years, pallet audits by pallet providers CHEP and Loscam have found that thousands of dollars have been incurred by transport operators in lost pallets. Customers are charged upwards of $24 by CHEP and Loscam for each pallet which is 'lost' or cannot be accounted for. This has prompted many transport operators to review their procedures and controls to ensure better pallet management across their networks.

Ideally, the transfer of pallets should take place when pallets are either picked up or dropped off, however, this is often not the case, and differences arise between the pallet balance in the books of transport operators and the number of pallets physically on hand.

The following reasons play a part in creating this difference:

  • exchange of pallet dockets does not take place at pick-up and/or delivery
  • unbranded pallets are exchanged on deliveries
  • excess pallets are not returned to pallet depot but are sold on the black market
  • pallets are 'de-hired' on the wrong account
  • staff incorrectly enter the pallet documentation into the pallet management system
  • mismatch between consignment and pallet documentation.

With multiple warehouses, depots, transport terminals and distribution centres, the timely recording of movements can be a complex exercise prone to errors.

So, what steps can be taken to improve the capture of pallet movements? And is there a way to identify the location of the lost pallets?

Finding pallets...with data analytics

Where pallets are lost, the key to their recovery can often be found in system data. Good information management will ensure the transport operator is in a position to provide evidentiary data to assist in the recovery of these lost pallets.

KordaMentha Forensic were engaged to assist a large transport operator in locating their lost pallets using data analytics. After months of poor management and neglect, the difference between the book balance and physical pallets on hand was escalating and this was costing the company dearly. The company was not only paying rental for the pallets it did not have, it had to record the contingent liability for the replacement cost of the lost pallets.

We assisted the company by utilising data analytics techniques to reconcile data from the consignment, financial and pallet management systems. Due to the incomplete and inconsistent nature of the data across the systems, the matching exercise was performed in multiple stages using a combination of values and 'fuzzy' text matching.2

As an example, one stage of the matching exercise was to compare the number of pallets loaded onto the consignment at the pick-up point ('transferred on', Point A below) with the number of pallets 'transferred off' at the final delivery point (Point B below).

Any movement in pallet balance between data sets A and B suggested that the final delivery point was a likely location of lost pallets - as not all pallets were properly 'transferred off' at that point. Such consignments were flagged for further investigation and reviewed back to proof of delivery documentation. This exercise was undertaken for thousands of pallets.

The results were then summarised to identify the combinations of recipients and locations for lost pallets. From this analysis, we identified that the top ten recipients of consignments accounted for 70% of the lost pallets. This allowed for minimal effort by the company's management to negotiate and recover most of the pallets.

Saving costs with data analytics

The loss of pallets can have a substantial financial effect on the organisation, as is shown in the example below.

During the annual stocktake of pallets, 'Haulage Ltd' was unable to locate 50,000 pallets. As Haulage Ltd was charged a daily rental charge for un-returned pallets, this suggested Haulage Ltd had been paying $1,750 a day (based on 3.5 cents/day) in rental cost for pallets it did not have on hand.

At this point, Haulage Ltd had the option of either paying the replacement cost of $25 per pallet (amounting to $1.25 million), or attempting to locate the 'lost' pallets and return them.

But as every additional day added $1,750 in rental costs, timing was crucial. Manual reconciliation of 50,000 pallet movements would have been complex and time consuming but using data analytics, Haulage Ltd was able to quickly identify the potential location of the lost pallets.

As well as saving a substantial amount of money, Haulage Ltd identified the underlying control weakness to prevent the repeat of such losses.

The unpalatable truth: the user must pay

It appears that this issue is not just restricted to transport operators. Other companies who retain pallets delivered to them may also need to pay rental costs for lost pallets, as a case against hardware retailer Bunnings in 2010 suggests.

In May 2010, the NSW Supreme Court ordered Bunnings to pay CHEP $10.98 million plus interest for the daily hire of almost 65,000 pallets between January 2002 and mid-20072 . Bunnings was found liable because they had used CHEP pallets that were left behind with deliveries by transport operators for commercial uses such as:

  • to display goods in their stores
  • to rack goods for longer-term storage
  • to shift goods from store to store.

The case puts forward a significant proposition – the user of the pallets may be liable for the rental cost even though they are not a CHEP customer. This suggests that companies which receive their deliveries on large volumes of pallets may need to implement their own pallet management systems to avoid incurring similar costs.

Conclusion

Whenever there is data, analytics can be used as a powerful tool to drive substantial economic gains for any corporation. The benefits include:

  • forming high level decisions backed by data driven facts
  • identifying previously hidden insights through a reconciliation across multiple apparently un-related data sources
  • achieving competitive advantages.

In the lost pallets example, we were able to reconcile large volumes of our client's data across multiple data sources. This allowed our client to prioritise the recovery of pallets, negotiate a settlement with the pallet company and benefit from a reduction in substantial ongoing costs.

Footnotes

1'Fuzzy' text matching refers to the technique of finding words that sound or look alike. Matching the names 'Smith' to 'Smythe' is an example. Using this technique becomes a necessity when dealing with free form text fields, that is, data that is filled in manually.
2Chep v Bunnings [2010] NSWSC 301.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.