Habit, distributed costs, no pressure to change: why inefficient procurement processes survive and why automation alone does not get rid of them.
It is rarely a lack of attention that keeps inefficient procurement processes alive. They survive because familiarity feels safe, because their costs are spread across several departments, and because each individual transaction seems too small to justify a change. And they survive the one measure that looks like the obvious solution: automation.
Many procurement processes are not obviously bad. They work, they satisfy internal requirements, and they map cleanly into the ERP system. A new supplier is vetted, set up, and approved; the order is placed and the invoice processed later. From the perspective of each individual participant, the workflow makes sense. It only becomes a problem in the aggregate, when the same process logic is applied to small and one-time purchases as well.
That is exactly where the effort spreads across multiple functions. The business unit specifies the need, procurement vets the supplier, finance requires accurate data, compliance runs the necessary checks, and at the end the invoice has to be processed. Nobody experiences the full effort of a single transaction. Each function sees only its own slice.
This is why inefficient processes stay under the radar for so long. Each individual transaction seems too small to prompt any larger change. Over months and years, however, hundreds of these cases accumulate. The problem then lies less in any single flawed process step than in the constant repetition of a workflow whose total cost is barely visible.
Most procurement processes were not deliberately designed to be complicated. They grew over the years: an additional check was introduced after an audit, another approval level after a compliance incident, and extra master data fields were added because finance or the ERP system required them. Every single step had a plausible reason at the time. We have described elsewhere how procurement creates this complexity for itself. This article covers the other side: why that complexity survives once it exists.
Over time, these accumulated steps become a routine that hardly anyone questions on a fundamental level. Employees know which documents are required, who has to approve, and where each piece of information needs to be entered. Precisely because the process is familiar, it feels manageable. Change, by contrast, initially brings uncertainty, because responsibilities have to be renegotiated and established routines adjusted.
The result is a typical imbalance: the known costs of the existing process are accepted, while the potential risks of a change are perceived more strongly. A process can therefore run stably for years and still be needlessly expensive. Stability is not proof of efficiency.
Inefficient processes are most persistent where individual transactions barely register economically. A small one-time purchase rarely attracts enough attention on its own to justify a process change. But when such transactions are repeated regularly, many small efforts add up to a significant cost position.
This effect is especially pronounced in indirect spend and tail spend. The individual order value is low, and the supplier in question is often of little strategic relevance. The internal process costs remain all the same: supplier setup, compliance checks, approvals, vendor master data maintenance, and invoice processing generate effort regardless of whether the subsequent order is for a three-digit or a five-digit amount.
That is why the administrative share of small purchases can be disproportionately high. It rarely becomes visible, though, because companies measure purchase prices far better than the internal cost of an order. The price of a product sits in the system; the time of several employees does not. What a single transaction actually costs, and how the math works out in a concrete case, is the subject of our analysis of process costs in procurement.
More control does not automatically mean more efficiency. Supplier vetting, approvals, and master data maintenance serve important control functions, and a strategic supplier with a high purchasing volume justifies the full workflow. For a supplier used once for a small amount, however, that same workflow is oversized. This does not mean controls should be dropped. The decisive question is whether they have to be carried out in the same way, inside the company, for every single purchase.
Many companies respond to process inefficiency first with digitization. Supplier onboarding is automated, approvals run through workflows, invoices are processed electronically. This can significantly speed up individual work steps and is sensible in many cases.
The underlying process logic, however, remains untouched. If a company still sets up a separate creditor for every new one-time supplier, the same administrative transaction is simply executed faster. An automated transaction is a faster transaction, not an avoided one. The number of transactions does not change.
The bigger lever therefore sits one level higher. Companies should not only ask how a process can be executed more efficiently, but also how often it needs to happen at all. As long as the number of transactions stays the same, automation optimizes the execution of the problem, not the problem itself. Process automation and process simplification are not the same thing.
This structural level is exactly where the single creditor model from Pedlar comes in. Companies no longer have to set up and manage a new creditor in their own systems for every individual one-time supplier. Pedlar acts as a central billing partner through which different one-time suppliers can be handled; delivery still comes directly from the supplier.
Functional responsibility stays with the company. The business unit defines its need, procurement can set rules and approvals, and the company's compliance requirements remain in place. What changes above all is the administrative structure: many changing external suppliers become, on the company's side, a single creditor relationship.
This does more than make an existing work step faster. The total number of internal transactions goes down, because the checks the company requires anyway no longer have to be repeated for every new supplier. In the long tail especially, the result is a considerably leaner process structure.
Another reason change comes slowly is distributed responsibility. Procurement may own the process, but it does not carry all of its costs. Business units lose time in coordination, finance processes invoices, IT runs the systems. Each function optimizes its own slice while the underlying process architecture stays in place.
The result is a series of local improvements without anyone questioning the workflow as a whole. Procurement shortens processing times, finance automates invoices, IT improves interfaces. But as long as nobody looks at how many people, systems, and process steps a simple purchase triggers in total, the actual problem remains.
An end-to-end view is therefore essential. Only when companies consider the full effort behind a purchase does it become visible whether a process still stands in reasonable proportion to order value, risk, and strategic importance.
Not every existing procurement process needs a fundamental overhaul. Many workflows serve their purpose and are exactly right for strategically relevant suppliers. The problem arises where the volume, frequency, and risk of a purchase no longer make any difference.
A modern procurement process should therefore be able to treat different types of purchases differently. Strategic suppliers need intensive management, recurring needs can be handled through established supplier structures, and one-time purchases need an economical path that does not create a complete new creditor relationship every time. The single creditor model does not replace classic procurement; it complements it where legacy processes generate a lot of administrative effort and comparatively little additional value.
Inefficient processes do not disappear because someone recognizes them. They disappear because someone reduces the number of transactions they generate. Would you like to discuss where this lever sits in your procurement? Schedule a call →
