In the late 18th century, a Hungarian inventor toured Europe and the Americas with a chess automaton that played a wicked game of chess. It was initially regarded as a marvel of technology but turned out to be a scam. Suspicions were confirmed when it was shown that the world famous “Mechanical Turk” had a human chess player inside who was moving the pieces on the board.
Today’s buyers of enterprise software are trapped in a similar illusion about the systems that they buy. It’s true that information technology properly deployed can create predictability, precision, consistency, discipline, and scalability. But at the same time, business information technology bears a number of uncomfortable similarities to the Mechanical Turk – and organizations ignore them at their own peril.
Let’s take a look at three of TriFinance’s most recent cases (anonymized of course), to illustrate the point.
A midsize corporate client wanted to streamline its procure-to-pay process, the process to capture a purchase request, create a purchase order, receive the goods and the invoice, process the invoice, and pay on time.
Most companies nail this process, because it’s fairly easy. And it was straightforward with this client, that is until they installed a new ERP system and additional software to enable digital supply chain finance with suppliers inside and outside the company. We were called in when users of the new system revolted: they were unhappy because “others” were “not doing their job,” the number of errors in the system were exploding, and suppliers were complaining.
To avoid endless finger pointing between all the departments, we decided to “mine” the procure-to-pay process. We knew that if we would see how the process was set up in the ERP system and could look at the data, we would be able to spot the weak links in the process. Unfortunately, ERP systems don’t allow you to download a process. In the end, it took a new database with more than 40 tables to reconstruct the invoicing process, plus a tool to extract the data and another tool to merge the tables after extraction. It was only after these preliminary fights that we could start on our real work.
At a large industrial multinational, we faced a similar complaint. A seasoned executive with more than 35 years of experience in the industry compared the implementation of ERP to someone “putting a blanket over [his] production plants.”
“Before”, he said, “I used to feel the activity. People were in close contact with operations, they knew what was going on, and they understood how business operations fit together.”
Since the ERP implementation, he feels he’s lost his grip on his business: “If I want to tackle a business problem, I need to talk to five people and puzzle their answers back together. While in the end, what we do, is just one simple flow: we source products, manufacture and convert them in finished goods, and sell and distribute to customers.”
These two cases are by no means an exception. In a lot of companies, technology has not only failed to obliterate old processes and create new, better processes, but also made it virtually impossible for employees to even fully understand the processes.
ERP systems have separate modules for procurement, accounting, controlling, sales, production, maintenance, transportation, etc. Each department manages, at best, two modules, and barely understands the others. Due to this fragmentation, people lose understanding of the flow — which is crucial to understand where their own work adds value. They tend to substitute the larger goals of the department or process for the narrow goals of their own interaction with the system.
The damage to the organization can be devastating: the cross-functional perspective wanes, issues start to fall between the cracks, ownership fades, the bigger picture is blurred, all while motivation and ultimately performance also suffer.
Like the Mechanical Turk, business information systems are a black box. They look shiny on the outside, but the specifics are never fully revealed to the user. This leads to stress at every level of the organization: “Am I seeing the full picture or am I missing something obvious?”
Another telling case: in a global e-commerce company, a new technology platform was installed to pay freight partners.
Previously, freight partners would send their bills to the company on paper. The administrative assistant would validate the bill details. The bill would be paid, and it would be retained in old-fashioned binders. Then the company switched to a cloud-based platform where all freight carriers upload their bills. These bills are supposed to flow into the finance system automatically.
The problem is that the upload is anything but automatic. Uploads are riddled with errors, and employees have to correct them manually. After correction and validation, they need to export the data from the cloud solution into the finance system of the company. However, because the two systems are not integrated well, employees have to manually match the syntax of some fields from one application to the corresponding fields in the other application.
The entire process is a tragedy of wasted time and motivation. “It takes up to 4 hours per upload, three times per week, to correct files and upload them into the finance system,” complained an employee. It’s hardly a wonder that morale hit rock bottom at this company.
This is another problem with business information technology that we see again and again: the new tech is supposed to be autonomous and automated, but it isn’t.
Like the Mechanical Turk, the system relies on high quality input and continuous maintenance from people in order to facilitate a set of processes.
The three cases share a central theme: technology has not only led to feelings of professional alienation and frustration for the users, it’s also actively doing the opposite of what it was supposed to do, namely increasing productivity and decision making.
We think that this is because technology is often deployed wrongly. Organizations should spend more time rethinking processes and allowing their imagination to guide them, instead of looking for IT solutions that will replicate their broken processes in a digital way.
Sometimes decision makers choose a solution oblivious of the fact that it will require frustrating workarounds or manual interventions, while this really transforms the meaning of work.
But the darker, deeper truth is that ERP systems have simply not lived up to their promise, which is that all data-related matters of an organization could be brought together in one single location and tool.
Looking at the current information technology landscape, it’s clear that there is no “one tool.” Legacy infrastructure is typically overlaid with a wide range of applications and interfaces. These add-ons are often marketed as plug and play with the underlying technology, but often, they are not.
The software industry adds to the drama through constant releases of new functionality. In his 2003 book, “Facts and Fallacies of Software Engineering,” Robert Glass explained that every 25% increase in software functionality increases the complexity of the system four-fold. At the same time, the impact of new functionality on cost reduction, operations, and performance is marginal.
As most of us are getting used to well-designed consumer applications in our daily lives, we are getting less and less tolerant to business software that is poorly set up and requires our continuous assistance. Business applications are now compared to consumer applications. And the employees’ opinions are becoming more and more vocal and cynical.
So what are leaders to do? Of course they need to keep stressing the authority of the process and not the tool. But there’s more: leaders who decide to select and roll out new technologies in their companies should move beyond mere functionality discussions and assess the true organizational impact of new tools through the reality of their employees.
Enterprise technology can be challenged aggressively through three simple questions; (a) will the software application significantly facilitate collaboration between employees?; (b) will the software application significantly reduce the “processing” complexity of work?; and (c) will the software application significantly increase the employee understanding of the full operations?
When the technology vendor is affirmative on all three axes, follow up with the question: “How?”
In the end, the real challenge we face today is not digitalization; it’s people — users who are getting the distinct impression they are considered by management to be an inferior machine, which exists only to feed a superior machine: the technology platform.
Van Caeneghem and Bequevort are management consultants with TriFinance in Belgium.