Reinforcing Loops
Even to an untrained eye, news reports of demand-and-supply bottlenecks, labor and equipment shortages, and pandemic-driven uncertainty were hard to miss in 2020. Although the intensity of those disruptions has since subsided, interest in expertise of organizations’ operational ecosystems has remained high. Witnessing the ripple effects across global value chains, the rising demand for supply chain expertise, and recognizing how my systems-thinking mindset naturally aligned with these dynamics, collectively motivated my interest in pursuing a career in supply chain.
With several years of experience under my belt as a practitioner and consultant, the industry faces a host of challenges spanning, data integration, workforce evolution, demand volatility, geopolitical risk, overall visibility, skill gaps, aging infrastructure, regulatory complexity, customer expectations, supplier risk, energy costs, climate change. The list goes on. With a brief glance at this list, it’s intuitive to say that potential solutions to meet such challenges ought to be proactive in their reactive responses. For example, geopolitical risk is very real, but this is completely out of the control of a company. Yet, the company can take measures to diversify its supplier base to hedge risk. In many cases, when organization do not manage exposure to a risky external variable, they may face reinforcing feedback loops.
Reinforcing feedback loops are found whenever behaviors or events inside the loop reinforce one another. These loops amplify the effect of the process. Compound interest is a common example. The more money one has in the bank, the more one earns on interest. That money is added to the balance and so the interest adds even more. The cycle repeats. In this way, reinforcing feedback loops are exponential, not linear. In this example, you have two variables inside the loop: the balance on the account and the interest earnings. An external variable influences this loop – the interest rate. This variable will influence the output of the feedback loop but doesn't change the core mechanism of the loop.
Let’s examine two, real world examples of this in supply chain.
- Owing to cost cutting pressure, an automotive manufacturer’s sourcing team is looking to slim down their costs. Their buying strategy selects lower-cost suppliers to meet short term financial targets. The suppliers deliver lower-quality materials. This leads to production delays, rework, or customer complaints. These operational issues cause higher overall costs, which can lead to further cost cutting measures. The cycle repeats and worsens.
- An insurance provider’s procurement department has limited visibility into their spending habits. Data is spread across different departments in disparate systems. Over time, this leads to more untracked purchases across departments, often at higher prices or from non-preferred vendors. As a result, spend data becomes even more incomplete and unreliable. With less accurate data, the procurement department cannot effectively negotiate contracts or enforce compliance. This weakens purchasing power and encourages more off-contract buying. The cycle repeats and worsens.
Loops in Spend Management
As evident in each of these examples, reinforcing feedback loops are critical to identify, correct, and continuously monitor over time for an organization’s financial health. The problem is that companies are historically bad at tracking internal the internal data required for rigorous analytics and key decision-making. Just a year ago, a market leading S2P software provider published a study that said that 46% of finance leaders lack complete visibility into spend data across the company. This leads many to ask: if leaders don’t know where or why money is being spent, how can they strategically cut costs and strengthen financial planning. This data aligns with our examples. For instance, in the our first example, better budgeting data could have avoided the decision to cut costs with materials. Better spend data in the second example could have reduced maverick buying.
Solutions that promise comprehensive visibility into company spend are diverse, yet they largely follow the same industry-wide best practice: automating and unifying processes—and by extension, spend data—through digitization. This focus has given rise to a market composed of full-suite platforms (e.g., Coupa), hybrid systems (e.g., Jaggaer), and standalone tools (e.g., Zip). Within this B2B technology landscape, software generally becomes more modern and modular as one moves from full-suite to standalone offerings, with the latter often representing newer market entrants (newer technologies). Still, the ecosystem is largely dominated by full-suite, market leaders whose platforms, beneath the surface, depend on a dense web of integrations connecting disparate enterprise systems. On the surface, the end-to-end value proposition is compelling; yet once we peel back the layers, we encounter a range of persistent limitations such as:
- High implementation and maintenance costs - high license costs, customization costs, subscription, consulting fees, in-house IT team
- User adoption barriers - unintuitive interfaces, shadow procurement practices (offline usage)
- Limited flexibility - predefined workflows, rigid system administration, industry-agnostic features
- Vendor Lock In - high switching costs (integrations, supplier enablement, historical data)
- Slow time to value - 12-24 month implementations, misalignment with business value, flawed project execution
Loops of Innovation
To me, the prevailing spend management solutions of the past two decades have become negative reinforcing loops themselves—albeit a step in the right direction. The insurance provider adopts a new system, but soon needs a dedicated team to manage the backend, remains locked into multi-year contracts to realize a net positive ROI, and finds that business decisions are increasingly constrained by the tool itself because of pre-defined workflows. In absolute terms, they are better off, but the nature of their challenges has simply changed. Many of these challenges persist, or even intensify, as innovation within the software stagnates.
In some ways, solutions to problems can become reinforcing feedback loops themselves. The market identifies a pain point, but in the process of implementing a solution, new and unintended frictions emerge. Like many software-dependent sectors, the source-to-pay market may follow the growth dynamics described by the 2025 Nobel laureates Joel Mokyr, Philippe Aghion, and Peter Howitt. With the advent of AI, I expect a new wave of creative destruction, as AI-native—not merely AI-enabled—platforms begin to take shape. Once these entrants gain a foothold, they may start absorbing the capabilities and expertise of older, standalone solutions, leading to consolidation across the ecosystem. Time will tell whether the benefits of these emerging tools will outweigh the high switching costs that have long defined this space.
