Do I really have to or can I already invest in new technologies if significant potential in the current systems has not yet been exploited? KPMG Value Audit: see more – understand better.

Y

ou all know the anecdote of a driver who hasn’t been shown how to shift into second gear before. Satisfied with himself and the world, the vehicle is used with virtuosity in first gear, stretching the equipment to its limits. In order to drive faster now and take advantage of all kinds of innovations, the aim is not to get a grip on the problem with gear changes, but to purchase a new car – with more engine capacity and more horsepower, so that the goal of “driving faster” is actually achieved (at least in some areas) without at the same time making any decisive changes to the way the vehicle has been driven for many years.

The same issue arises – albeit in the real world – around digitalization. Almost every day we discuss, for example, the use of artificial intelligence, high-speed analyses and the use of structured and unstructured information in so-called “data lakes” without asking with the same intensity whether the potential in existing processes and systems has been sufficiently addressed and used.

It is undisputed that all these “high-end topics” will change our working world and thus our lives in the long run. However, it is quickly overlooked that many of the practical innovations make high demands on the data (e.g. coherence, uniqueness), processes (e.g. stringency, no parallel processes, limited flexibility) and systems (e.g. homogeneity of the systems, comparability of the contents) without which the desired success cannot be achieved or only with a disproportionately high amount of additional effort. Systems that can achieve perfect results on a completely unstructured basis do not yet exist.

Digitalization: Status Quo in the German Economy

One of the central questions for future activities is now the appropriate location determination. What are the current relevant challenges for a particular company? Is it better to use artificial intelligence or to stick to more “traditional topics” such as master data, system usage and process variants?

A study on digitalization in accounting from 2017, conducted in cooperation between KPMG and LMU Munich, concluded that the “traditional topics” such as data quality management, system homogeneity and process automation are still the processes used at present.

This assessment also corresponds to the findings from the focus team activities of the 2017/18 audit season. On closer inspection, however, the picture is by no means homogeneous – i.e. there are a number of companies that have digitalized and optimized their processes intensively over recent years and are already in a position today to apply the latest technology in a targeted manner to an optimized data set – with some impressive results. None of this came quickly, unplanned or smoothly.

No time, no resources and no budget

On average, however, the picture is different. No time, no resources and no budget are the three most frequently cited arguments as to why the challenges of the past also represent the challenges of the present. The gap in the German economy with regard to processes and systems has thus widened more than ever before. In an environment of low interest rates, high demand and capacity bottlenecks, this development is usually not directly apparent – but this can change quickly in a gloomy economy. It is essential to use the time until then as intensively as possible.

Transparency is the key to success

In numerous operational processes, comparatively few key figures are required to get a good first impression of the status quo. It is not the number of KPIs (“Death through dashboards”), but the conscious selection and interpretation of the results in the respective environment of the company that makes the difference.

Example: Purchasing

A typically meaningful combination of key figures in the so-called purchase-to-pay process (short: purchasing) consists of the so-called order rate (purchase-order rate), the degree of automation in the so-called subledger as well as the lead time.

Order rate
The automation potential of SAP in the processing of purchase orders, for example, can only be used if the SAP module for materials management (so-called subledger) is also used. The order rate refers to the proportion of invoices processed via SAP MM in relation to the total document volume in the company. A high order rate is therefore one of the basic prerequisites for using the efficiency potential in an ERP system.

Degree of automation
An essential goal when using integrated ERP systems is to automate the process chain as far as possible – i.e. from ordering to invoice verification and booking to payment.
An essential key to this is the quality of the data in the system. Missing account assignment information in the purchase order, missing information on the invoice, incorrect price and quantity information etc. lead to the consequence that the basic automation potential available cannot be fully harnessed. This results in extensive manual interventions and releases.

Lead time
The processing time (e.g. invoice date vs. release date) is a good indicator of the degree of digitalization actually achieved; as such, high receipt volumes and below average transaction volumes (i.e. invoice amounts) with a processing time of 0 days (i.e. full automation) are essentially no problem. On the other hand, even the number of receipts and the average invoice volume are important issues for optimization considerations.

The “broader” the definition of processing time, the more information can be gained about the overall process. For example, there are quite a few companies that have excellent values for the key figures “order quota” and “degree of automation” – the unchanged high lead time for the entire process alone indicates that an essential component of the analysis is still missing (e.g. upstream, manual coordination processes).

“An essential goal in the use of integrated #ERP systems is to #automate the process chain as far as possible – i.e. from ordering to invoice verification and booking to payment. How you can implement this for your company… #AuditoftheFuture“

Twittern WhatsApp

How can KPIs be turned into concrete findings?

Current technology makes it possible to have a large amount of information at one’s disposal, including the key figures mentioned above. But what does a PO rate of 79%, a degree of automation of 82% and an average lead time of 12.4 days mean in an industrial company (series production), for example? Are these results good or bad? Where is the problem, if any? And which activities have already been proven in comparable cases to solve this problem?

It quickly becomes clear that the results of machine routines alone do not help. Only the combination of targeted results from digitalization with benchmark and best practice know-how, experience and expertise makes it possible to further develop digital results into concrete (company-specific) recommendations for action. It sounds almost paradoxical – but the more powerful the digital analysis tools become, the more important people and their experience from other companies become in order to actually be able to utilize the results.

In auditing, we rely on the focus teams for analysis

These teams are a special feature of KPMG and specialize in the typically ERP-based core operational processes of our clients (i.e. purchasing, production/inventories and sales). A distinctive feature of these teams is the combination of development and application of digitalization solutions in practice. In addition, these teams have an extraordinary benchmark and best practice know-how from more than one thousand audits per year, which we use for our clients not only to carry out the audit but also for targeted reporting from the audit (e.g. on the purchasing process).

First step in the analysis: In the example presented, the PO rate of 79% is well below the industry average of over 90%; i.e. compared to the peer group, more than twice as much volume is ordered and processed outside of materials management (manually). As far as the system-supported purchasing process was used, the achieved degree of automation of 82% is also well below the benchmark of approx. 95%; i.e. every fifth invoice that could have been basically processed by machine must be manually reworked and released. The KPIs, which are below average compared to the benchmark, are also reflected in the lead time of 12.4 days, which is also well above a best practice value of less than 5 days or the industry average (between 8 and 10 days).

Without examining the facts more closely, the example presented demonstrates that the first step will probably not be further developed using a data lake, artificial intelligence or a new, significantly more powerful database or ERP structure. Rather, the first step will relate more to traditional fields such as increasing the PO rate, improving data quality and optimizing processes as well as their processing. Of course, innovative tools can already be used at this stage to identify areas of action (e.g. process mining – visualization of processes) or to automate simple routine issues (e.g. RPA – robotic process automation).

How do I get this information?

For our clients who fulfill the prerequisites for the use of digital audit tools (KPMG Clara Analytics), information of this kind is already part of our standard reporting from the audit (KPMG Value Audit).

Do you have any questions or suggestions?