Maybe Software isn’t the Solution — A Product Manager’s Perspective
Our lives are ruled by software applications- and for the most part, we can’t get away from it. This article focuses on how this fact of modern life translates to the B2B software space, dominating the processes at the backbone of our global economy
As a Product Manager, I spend a lot of time listening to clients talk about how the software products I manage are not solving the problems they thought they’d solve. I ask them to re-explain the problem to me as if it were the very first time, and explain it back to them until we are 100% on the same page on what needs to be solved for.
For those unfamiliar with Product Management: before you think “your product sucks”, think again- it’s the PM’s job to dissect problems, and abstract the bigger picture of what is going on across the market. Because I manage a suite of mature products, I spend most of my time with clients who are unhappy and need a new feature to complete their workflow, which has likely evolved since they first started using our products 15 years ago.
Over the past year, I spent hundreds of hours interviewing clients and sitting in on demos given by our rock-star team of Sales Engineers, noting down questions and requests. I wanted to understand where the process of setting expectations was going wrong, until the answer hit me: software only works if you have predefined standard processes that everyone agrees to follow. Else, it could even do more damage than good.
It became apparent that many of our users were not following the predefined processes, and this wasn’t clear until we asked multiple stakeholders the same questions, receiving different answers.
When dealing with very large organizations (Fortune 100 size), it’s sometimes hard to get airtime with users directly, especially if they are located in different time zones. In the interest of keeping a large project on track, Product Managers are often stuck having conversations with with Project Managers and Systems Integrations Engineers on the client side, rather than with the actual users of our software. By the time information reaches us, it’s been filtered through multiple people, resulting in a corporate version of the telephone game.
The Problem with Traditional Industries
I work in the Maritime Commerce industry, which is pretty messy. Despite the fact that 90% of all goods in the world at some point are handled through Maritime, there is no standardization of processes across companies or regions. Every company defines their own workflows and rules, and often uses a combination of Excel, email, and in-house built software to keep track of things. It works so far because there are a limited amount of companies in the world, and new players in the space are rare.
The Maritime industry is at least 10 years behind in implementing systems to collect and measure data. When interviewing people who manage port operations for transporting billions of dollars worth of products every year, it blew my mind that most large commercial ports don’t even keep track of the time required to load and unload cargo- and then wonder why there are constant delays. This means we lack historical data, and have no proof or track record of which processes are being used, and what problems they carry.
Having been a sailor and spent time at sea for most of my life, I can tell you that around the ocean, stuff breaks all the time. For maritime companies, that means constantly having to repair equipment worth millions of dollars a year. For ports, that means having to take entire sections of the port out of order to make equally costly repairs. As of 2022, we have very little data on all these interruptions — people just deal with them, as they have for centuries.
The most Common Issue in Software Implementations
I’ve worked for SaaS companies for the past ten years, mostly in banking and energy. During implementations with large corporations, consultants spend months collecting data on the current process, designing flow charts, and finally getting everyone in the room to agree on what the standard processes are. Only when the software implementation , the questions start pouring in: “how do I handle this exception?”
Both parties soon realize that there are way more exceptions than foreseen during all the research and interviews, normally because there is someone in operations who just handles the problems as they come along, on top of their regular job, normally with a hacky, sometimes not-so-compliant solution. Most of these exceptions happen when incompatible processes between departments collide, or someone built an unstable integration between two pieces of software.
So now you need to redesign the process, accounting for every exception that comes up. You’ll discover that many solutions you come up with are not compatible with how the project was architected. You realize that if the exceptions are the norm, it doesn’t make sense to use software to manage a process, because the reality is that there is no process. We asked for a process because we needed one to fit our software, and spent months putting something on paper that doesn’t really exist in reality.
This is because the human mind is still better at solving complex problems with too many variables than most commercial software is even close to being. The workers involved in those messy processes were subconsciously aware of this.
Is there a Solution?
No problem is impossible to solve, but solving problems in traditional industries, such as shipping or manufacturing, requires going much further back to the source than most PMs can do. It requires disrupting decade-old practices, and lucky for us, this might be exactly the right time in history to do it.
Most companies in traditional industries are behind in digitizing their processes. This is mostly due to the fact that most older workers are reluctant to use software to accomplish something they have been doing manually for decades. There is only so much management can do to push an entire company of workers towards digitalization.
However, a paradigm shift is happening. Many Boomers are getting ready to retire, and the current post-Covid worker shortage is forcing companies to restructure teams and adopt new tech to manage their processes, rather than hire the same number of people to replace those who leave. So companies are hiring younger, tech-savvy workers to do jobs previously done by multiple people with the aid of the right software.
At the same time, many supply-chain workflows have collapsed during the pandemic, and now that it’s over (fingers crossed), they are ready to be re-built. A great example of this is the extreme congestion in container-shipping ports for the entire year 2021, which is still not solved. With months worth of goods accumulated in shipping yards, the delays are becoming so absurd that they are forcing shipping companies to revisit their processes.
These systems collapsed in the first place because for decades, we kept building software to solve each individual problem, and then were forced to integrate all the pieces with each other, accumulating tech debt along the way. Everything is interconnected, and we made it absurdly complicated.
Many workers out in the field were miserable fixing exceptions, and then having to record the fixes in software not built to account for so many exceptions. This is why so many boomers thought they were better off before using software to manage their extremely messy processes- and they weren’t wrong about the problem, even though they came to the wrong conclusion.
The answer is not to avoid using software. The answer is to change the way we do things. Most processes we currently have in supply-chain have evolved from what was put in place post-World War 2, at a time where most big decisions were taken behind closed doors in boardrooms. Management designed processes and pushed them on the workers, who had very little leverage to change things, so they handled exceptions by thinking on their feet. Eighty years later, this no longer works, as the amount of exceptions got out of control (one could argue it never really did work).
In 2022, the right way to build a process to function at a large scale is to collect feedback from stakeholders at all levels on what the process should be. Once you have buy-in, then you can build brand new software that reflects the work being done by the people on the ground, and helping them actually track exception handling.
And yet, some corporations still buy software from third-party providers (or build it in-house), hoping it will fix all their problems, and then blame who built the software when things still go wrong. No amount of software can fix a broken process. But people can: trust your people, then build software.
The pandemic brought us many bad things, but I do believe the collapse of many supply-chain processes is a positive byproduct of it. This is our once-in -a-lifetime chance to start from scratch, building systems that will last for generations to come. Will we get it right this time?
Thank you for reading. If you’d like to continue the conversation, you can find me on Twitter and LinkedIn.
