Shared services provide an environment in which automation can be efficiently implemented. Part of the rationale for establishing shared services is often to provide a technology platform which is scalable to the growth aspiration of the business, thereby realising investment efficiencies; the implementation of automated tools is one example of this. In a typical shared service environment the utilisation of development skills can be maximised, and a single best practice model implemented using the widest possible scale. Implementing automation in an unconsolidated environment is very likely to be fragmented, inconsistent and piecemeal. The results will be sub-optimal. Implementation costs are higher and benefits lower than they could be, and information sharing is made more difficult.

There are many paths to achieve automation. This article focuses on the use of a tools layer to supplement the core functionality of an ERP system. When a shared service operation is created, there is normally standardisation on a single instance of a recognised ERP; this is frequently a prerequisite to driving out efficiency. This ERP provides the core transaction processing and holds corporate data. Once established, there is an understandable reluctance to re-engineering the functionality of this core system. Depending on the ERP there may be adequate functionality. But it is likely that automation will involve some use of external tools, to support data capture and validation, reporting, analysis and self-service.

The Benefits

Automation business cases are often highly attractive. Frequently the cost of implementing the tools concerned is modest, and they can be implemented in manageable phases. Individual payback is often quick (e.g. 18 months or better). There often is a step-change in service improvement/ reliability, information accessibility, headcount savings & improved cash flow. Some of the key benefits cover:

	Reduced working capital/ improved cash flow. A Finance shared
	service function will provide a key role in ensuring cash is collected on time and in controlling and reducing DSO (Days Sales Outstanding). Automation typically helps to reduce errors, to focus resources in collection activity, to provide analysis capability and to enable the quick management of queries.
	Reduced income leakage. Errors are likely to count against an
	organisation; people will always complain if they are over-charged or charged incorrectly. Error reduction, assisted by automation, can therefore have a significant effect on ensuring all income is billed and realised.
	Substantial reduction in manual work. The effect can be dramatic;
	for instance, taking an Accounts Payable function from 5,000 - 10,000 invoices per year per FTE to 20,000 – 30,000.
	Consistent compliance and assurance of compliance, through
	embedding controls, authorisation and policies/ rules (e.g. expense rules, multi-jurisdiction tax requirements) in the automated process.
	Improved operational information through:
	Faster, more up-to-date reporting. On-line access throughout the organisation to reports and   source documents

Example Applications

It is not possible to go through all the automation opportunities in one article. The discussion that follows has therefore been selective, and the rest of this article therefore concentrates on two applications of automation:

	The reduction of manual intervention in the purchase to pay process.
	The reduction of errors in the order to cash process and focusing
	collection effort.
	Both processes involve the control of cash and accurate recording of
	liabilities of and debts due, and are critical to the well-being of the organisation. In both processes, there is frequently a lot of administration to get right. Issues normally faced are:
	How to deal with processing large transaction volume much of which
	is of a routine, compliant nature, and to avoid backlogs building up?
	How to ensure accessibility to data on transactions?
	How to ensure that the commitments the organisation has entered
	into are honoured (both as a supplier and a customer)?
	How to optimise cash flow?

Purchase to Pay
A good well controlled purchase to pay process can be considered in a linear fashion with Finance playing the leading role in the latter parts of the process.

Of course it sometimes does not work this well. This too is important as we will come back to below. The Finance role is to receive invoices and, after logging and sorting, match them with purchase orders and receipts and ensure mismatches are resolved. In an unautomated P2P function, about 70% of the effort is required to process invoices to the point at which it is released for payment.

Automating this process to reduce significantly manual intervention requires:

	Automatic data capture through either open exchanges or intelligent
	scanning
	Matching against order and receipt data, and Workflow to route
	queries to requestor/ buyer/ authoriser for resolution
	On-line archives to enable users to access invoice images
	The results of these are better accuracy, and enable more
	accessible and timely reporting.

The first task is to capture information from purchase invoices. EDI has been around for some time, enabling electronic communication on invoice data from supplier to customer. But this is only effective for supplier/ customer relationships with very high volume/ value of transactions, because the links/ protocols need to be set up individually with for each relationship. It can suffice for some areas, particularly direct purchases. But, for many purchase areas, classical EDI is not viable. Most data capture has therefore remained to be dealt with by traditional paper-based processing, as shown in the diagram below.

However, “open exchanges” are now available that make it feasible to eliminate paper processing for supplier with only moderate data volumes. In some ways this is like EDI in that communication of invoice data between supplier and customer takes electronically. A service provider will translate the output from the supplier into those intelligible to the systems of the receiving organisation. There can be wins in this for both suppliers and customers, in that both are not saddled with the problems of paper processing. The obvious inefficiency in using what is in effect the print out from the supplier computer to key into the customers is eliminated. However, there are costs, efforts and time involved in getting suppliers to sign up, and some paper processing will always remain.

Another method is to use “intelligent” scanning. This differs from dumb or ordinary scanning in that it incorporating OCR features and often some form of “learning” – the software can learn to identify invoices and different types of invoice. Paper is not eliminated, and therefore the issue is how to process a large volume of invoices from different sources efficiency and accurately. Documents are typically semi-structured, in that they have commonality (e.g. supplier names and VAT details) but they have differences (the VAT information is not in the same place). “Intelligent” scanning has some advantages. There is minimal change from the supplier perspective; they still submit invoices exactly as they did before. However, “intelligent” scanning takes time and effort to set up, and organisations sometimes have difficulty in achieving a satisfactory extraction rate of information from source documents.

So far we have discussed data capture. But, once captured, invoice data needs to be processed. This involves matching with order and receipt information and then managing the exceptions that result typically through a workflow system. Matching is most likely to be an operation that is reliant on the corporate ERP systems which should capture ordering and receipting information. Workflow applications can either be supplied with the ERP or be acquired independently.

Finally to make information available, a copy needs to be held in an archive. This archived copy is available to authorised staff through on-line enquiry. A typical way information can be made more accessible is for budget holder to be able to drill down as far as source documents if necessary.

It is important to remember there are pitfalls in automating purchase invoice processing, and many implementations fail at least initially to achieve expectation. One is to ensure that there is a good process flow. The purchase to pay process is essentially linear and requires that order commitments and receipts are accurately recorded on time. If they are not, then many of the benefits can not be realized. Automation tends to put greater emphasis on clean process, and it is not just a matter of acquiring a tool and plugging it in. It also relies on a good understanding and definition of the processes in use; in practice processes are often more reliant on the judgment of clerical staff (e.g. how to deal with particular types of invoice), than is usually realized.

Order to Cash
Strengthening end to end management of the order to cash process usually represents one of the easier ways to improve cash flow – reducing DSO and managing bad debt risks.

Perhaps a little simplistically the typically order to cash process can be considered to have four stages:

It is important to measure and reduce error rates and delays throughout the process. In a broken process, credit control frequently ends up spending a disproportionate amount of time sorting out order and fulfilment errors earlier in the process, the root causes of which have not been addressed. To be effective, each stage in the process needs to ensure the next gets the right service. For instance:

Order taking needs to be accurate – validation, for instance through automated order confirmations, and the encouragement of self-service can help.
Fulfil – proof of delivery needs to capture promptly and accurately and available for subsequently enquiry; for example, through “track and trace” facilities and the capture of delivery data at the point of delivery using hand-held devices,
Bill preparation from order/ fulfilment data needs to be fast and accurate; using electronic bill presentation can cut out despatch and mail delays.
Collection is aided by easy access to relevant information, such as proof of delivery, earlier in the process.

An important product of automation can be more accurate and faster sales information (e.g. through cleaner capture of delivery data) to enable better responsiveness of market changes. Some of the above do require major system change and therefore to an extent fall out of the realm of discrete, relatively easy-to-implement tools.

Much time in many Accounts Receivable functions is spent collecting debts. If many outstanding items represent disputes not yet surfaced, this costly and delays cash flow. Obviously more accurate billing and better capture of proof of delivery can go some way to eliminating disputes/ queries and hence reduce the cause of collection workload.

In addition a number of tools are available that enable better focus in credit collection and control. Typically these will be interfaced to the Accounts Receivable application and provide:

	Prompting items follow up based on corporate policy, customer
	parameters and previous customer interactions.
	Recording of customer interactions.
	Facilitating internal follow up (for instance by email) to be
	undertaken and monitored
	Ability to see customer profile and history, and look for recurring
	reasons for dispute, risk reviews, etc.

One of the advantages of these is to reduce the non-contact time that was previously spent on prioritising, preparation and follow-up. But not only is contact time increased, prioritisation functionality ensures the right items are prompted for follow up. There is also a much better flow of information. Reasons for recurring disputes and risks are better identified and addressed. The effects on DSO can be significant particularly where there is a large diverse customer base. Finally there is potential to reduce bad debts (through being more on top of clearly deteriorating situations) and to improve customer satisfaction (through being more on top of queries).

Conclusions
Automation tools are an effective way to improve the efficiency of a shared service operation. But over and above that they can improve its effectiveness as well as just makes it more efficient. Examples of this are:

	The provision of faster, more up to date information to budget
	holders and on sales performance, delivered direct to the desktop. Uses in the business can drill-down to source documents, if necessary, rather than requiring Finance to do this.
	The provision on a better control environment through embedding
	controls, authorization procedures/ levels, etc.
	The focusing of collection activity to enable more effective debt
	collection performance and reduced DSO.
	The analysis of debt follow-up activity to identify root causes of error
	and likely bad debt risks.