The future of the source-to-pay process

For the last 20 years, software companies have always done the same thing: digitized a proven process by trying to map as many process variants as possible as software.

The focus was always on mapping as many sub-processes as possible in the same tool in order to avoid interfaces and to have as little information as possible outside the tool in mail, Excel sheets or Powerpoints. The idea behind this was always process standardization, since only mapping in a fixed tool guarantees the same process implementation.

So, in the early 2000s, there were several startups that were all in the purchasing environment, but they all started with a different sub-process of source-to-pay. For example, some companies started with the analysis of spend on the basis of invoices, other companies focused on the sourcing process, and still other companies focused on the process of requisitioning and their own web store, although eCommerce processes became established later than the eProcurement processes.

Over time, all providers have successively expanded their portfolio of offerings in order to be able to offer an end-to-end source-to-contract process including comprehensive supplier management on the one hand, and on the other, the Procurement providers have focused on an end-to-end procure-to-pay process that also includes complex budgeting and workflow processes. Last but not least, larger source-to-pay providers were formed through acquisitions and takeovers, which were then ultimately able to map a standardized end-to-end process electronically, from spend analysis to invoicing and payment.

In the last few years, the larger all-in-one providers on the market have been busy closing the "gaps". Here, the focus is on "special processes" that are often not the responsibility of the CPO, such as "Expense Management" as offered by Concur or "Contingent Workforce Management" as offered by Beeline. The idea behind this is to offer existing customers the greatest possible upselling potential and thus generate additional business. This strategy is primarily driven by the major analysts such as Gartner or Forrester, who define which sub-processes a tool must offer in order to achieve a very good positioning in the Gartner MQ or the Forrester Wave.

But this strategy has its weaknesses, because the acquisition of more and more companies offering a small sub-process in the overall procurement context adds more and more complexity to the all-in-one procurement solutions. The integration of new solutions usually creates a number of problems and prevents vendors from bringing real innovation to the market. Instead, they are concerned with integration issues, with improving usability for the end user so that he does not have to operate different user interfaces, and of course with the normal maintenance of the applications, because existing customers naturally want to see their requirements covered in the tool at some point.

This is precisely where the opportunity lies for innovative startups that do not focus on the entire process but only on a sub-process, but do not just map this process in software, perhaps a little better or more flexible, but trigger real innovations.

Why are innovations more important today than ever before?

The reason is the much higher complexity in procurement than 20 years ago. Here are a few examples:

10 years ago, it was possible to define a category strategy and implement it more or less globally. Today, there are a large number of sub-strategies, as individual markets and regions behave completely differently and it is hardly possible to implement a global strategy.
In the last 10 years, many new laws and directives have been passed by governments that have an impact on purchasing and the purchasing process. Issues such as the Supply Chain Act are only the latest in a series of regulations that affect purchasing.
Supplier performance alone, and the impact of supplier performance on the buying organization, is a topic that has increasingly preoccupied purchasing over the last 10 years. At a time when supply chain networking is improving, suppliers need to exchange not only orders, invoices and quotations with their customers, but also a variety of additional data that allow conclusions to be drawn about their performance and stability. Identifying weaknesses, understanding the impact of these weaknesses on my own organization, assessing the strengths and weaknesses in the context of a bonus/malus system and their impact on the internal cost structure is often a small science in itself.
Recognizing impacts and reacting quickly to changing conditions have made risk management increasingly important, although risk management also has many different facets. In risk management, we talk about currency risks, country risks, financial risks in markets or suppliers, but also risks from environmental disasters or other external influences.
New trends such as sustainability, CO2 emissions, but also the topic of minorities and CSR are increasingly coming into focus. At the latest, when additional taxes have to be paid for the CO2 footprint, the CO2 emissions are just as important as the purchase price, as these 2 KPIs are then in a direct correlation with each other. However, a real calculation of the CO2 share is often not possible without the corresponding product cost analysis, which also includes an analysis of the cost and material components. Only this information provides detailed data needed for an analysis of CO2 consumption.
And add to this the complexity of volatility, because there is a big difference between whether the buyer is in a buyer's market or a seller's market. This can have a significant impact not only on the negotiation strategy, but also on the timing of the negotiation and the way it manages to generate competition between market participants.

If you now compare the classic sourcing process of 10 years ago with a process as it should be today, you will very quickly realize that no purchasing organization can any longer ensure that all decision-relevant factors are covered. Obtaining multiple quotes, comparing them, then negotiating and awarding the contract rarely covers what companies need today to stay competitive. Even the use of tools such as Celonis, which can be used for process mining in order to determine how a process is actually used today, has little positive impact because only existing and usually particularly inefficient processes are analyzed. The biggest advantage of Celonis here are the pre-build modules with which process standards are already delivered, even though the customer must often also bring in the innovation here.

But what do companies need to be better than others?

They need support in the form of a "digital co-worker". And the possibilities of a "digital co-worker" are also changing a lot at the moment.

Where in the last 3 years mainly ML (machine learning) was used to understand data series for pattern recognition, such as invoices in spend analysis or in the invoice receipt process, the intelligence of the digital co-worker will increasingly change. In the future, the focus will no longer be on recognizing patterns in large amounts of data, but we will see systems that can also derive decisions or at least suggestions for decisions on the basis of the data obtained.

Why is it so difficult to develop such systems?

This is easily explained: as almost all companies have a large number of developers who know different programming languages, database models, development frameworks, etc., but none of the developers knows about purchasing processes. Of course, most of the time this is not necessary, because every software company has its product managers, and they have to do the translation between customer requirements and developer specifications. But what comes from the customer as a requirement? Again, only more detailed information about how a process already looks today and what an end user would expect from a software.

But what would be needed in order to be able to go down truly new, innovative paths?

You would need to combine the know-how of the best buyers with data analysts and software developers and at the same time have a vision of how a purchasing process could look in the future so that companies actually have a significant added value or USP (Unique Selling Proposition). Since no software company has this know-how in total, most ideas today come from the purchasing companies themselves or from consulting companies that have purchasing know-how but also have data analysts and visions of how purchasing processes can be designed completely differently in the future.

So what should such software be able to do (different)?

On the one hand, it should help to incorporate more information into a possible decision in order to really consider all relevant aspects, and on the other hand, it should help to identify and implement the most important, next decisions. Several startups have already impressively demonstrated that this is already possible in some areas today, but it also shows that smaller providers have it much easier here than large ones because they focus on one topic and can link the know-how of buyers with the IT know-how much better and more efficiently.

The first providers such as Sievo or Orpheus (now part of McKinsey), for example, manage not only to present their customers with a data analysis in the form of a visually attractive dashboard, but also to trigger recommendations as to the most urgent action to be taken next on the basis of the insights. These recommendations can also cover different aspects: for example, they can recommend measures that are a priority for meeting purchasing goals, such as reducing material or process costs, or measures to improve a particular supplier's performance. This can also include the search for new suppliers, as can already be done today by Scoutbee many times faster and better than by most purchasing organizations.
However, they can also be aimed at increasing user adoption, namely when the recommendations directly influence tool usage. In the procure-to-pay environment, these can be recommendations that show the Procurement manager how he should adapt his product portfolio or simply optimize item descriptions so that users need less time for the search.
Tools can also suggest adjustments to workflows today, for example when they automatically recognize that if the approval rate is 99%, eliminating the workflow step is more efficient than maintaining the workflow at 1% non-compliance.

What do companies achieve with such tools?

To a large extent, they solve an HR problem, namely the fact that it is incredibly difficult to maintain the same high level of qualification at every location. This means they can distribute know-how better with such solutions and they are more agile, since decision trees in tools are easier and faster to change than the mindset of employees in the context of change management programs.

Additionally, processes are accelerated, and bad decisions are more likely to be avoided if you work through recommendation after recommendation and also build in control loops that measure the success of decisions and incorporate the results into future decisions.

A good example here are suppliers who have developed a product group management system. In the past, the product group strategy was created every 18 months and communicated to the entire team in Excel or PowerPoint format with the aim of ensuring that everyone adheres to the strategies defined in it. Since project buyers have to consider 10 or 15 commodity group strategies at the same time (since they buy different products from different commodity groups in a project), it is often tedious to check different documents for the correct strategy, which is why more and more companies are now trying to map the commodity group strategy in the sourcing tool, since you can then see the parameters in the sourcing tool and do not have to look in an external document. This would also make it difficult to understand afterwards, after a decision has been made, why and on the basis of which strategy the decision was made, if it is not stored and versioned in the tool.

However, the basic problem remains the same, namely that commodity group strategies are becoming increasingly complex because they can look different for individual regions and, on the other hand, the time lag is always there, which is why it is usually too late to react to new risks or changed market conditions. Completely new approaches are needed here, which allow a much more agile approach (and thus faster reaction) and still enable a high level of competence and decision quality. A typical provider here is nnamu, which takes a completely new approach to defining purchasing and negotiation strategy based on game-theoretical approaches.

So what does all this mean for the future of source-to-pay processes?

We are convinced that the success of source-to-pay providers will not be determined by more technical features for special processes that have little influence on the overall result of a company. Rather, it will be the power of innovation and thus the ability to combine the know-how of customers with the knowledge of data analysts, classic software developers or product managers and visionaries in order to generate new processes under changed conditions.

On the one hand, these new processes will revolutionize the operation of software because they will make it much easier to use, and on the other hand, they will generate so much additional value for the end customer due to clever recommendations, which in turn will be a significant USP for the software provider.

In the end, it is always the value that a software can generate for the user's company that counts and not individual technical features or the mapping of super complex (and at the same time often rare) processes in standardized tools.