Following the Value

My company's Expense Management system really ticked me off. So, in a period of wanting to connect with others at the company, I created a side project out of exploring why, and revealed $883k in annual savings.

Quick tangent: One of the earliest life lessons I remember was about the Bystander Effect. A psychological phenomenon where individuals are less likely to intervene in a serious situation if there are other people around— assuming that someone else has made, or will make, a decision. I vowed to never be a bystander at a young age.

Resuming: As I was saying— Our Expense Management system really ticked me off.

It was slow, it was buggy, few people abided by the purchasing policies, and it lived inside an accounting system that even accountants struggled to use, let alone a seasonal non-accountant who is just trying to be reimbursed for a hamburger. The purchasing policy violations really grind my gears, because, as a personal credit card aficionado, I knew we were leaving money on the table by submitting PO's and paying by check or ACH instead of using our company cards. Being on the FinTech side of the company, and having had prior experience using a non-ERP-based Expense Management system, my intuition was telling me "There is an opportunity here.."

My first step: Validate my feelings. Do others feel the same?

During breaks from meetings during the day, I decided to poke around all corners of the office with a pen & paper and casually interview people about their experiences with our expense system. I took note of everything—what frustrated them the most, what people seemed most confident about, etc. By keeping it casual and conversational, it was both a fun way to connect with people, and a candid way of getting real opinions. The more conversational I kept it, the more people relaxed, and the more they opened up and told me what they really thought.

What I learned was fascinating, because much of it was unexpected, sometimes in funny ways. For example, receipt management was one of the biggest pain points for both submitters and approvers. But specifically… HEIC files. High Efficiency Image Container format — the native iPhone image file format. The company is a Microsoft shop, and most people take photos of their receipts with their iPhones and then upload them into the system. The administrators reviewing expenses couldn't open the receipt attachments on their Windows computers. I recall laughing to myself when I noticed such a silly pattern, and wondered why I hadn't heard of it sooner.

Lo and behold, after interviewing 30+ employees, and surfacing hundreds of issues, I was not the only one dreading submitting an expense report. It dawned on me that for certain "power users" (especially executive assistants), the pain points were far more acute than I could have imagined. Even the AP processors were struggling with the administrative capabilities of the current system. The functional pain and inefficiency of using the system was there in abundance, for everyone.

My second step: Great, we all struggle with expenses.. Have there been any attempts to change systems before? Why or why not?

Now that I had validated a multitude of common functional pain points, I knew we had a problem (well, many problems). What I didn't know was whether these issues amounted to anything material, or what the reasons were for continuing to use the existing system. That's when I more seriously started to discuss the idea with stakeholders, who were simply too busy to be thinking in the way that I was. I decided to do the lifting and I got my hands on the contract for the company credit card provider, which was my first time analyzing an enterprise contract of any kind. What I found was that the rebate program details were a project layup that paved a pragmatic path towards switching systems. Game time.

My third step: Quantify.

The aforementioned rebate program with our credit card provider was dynamic— meaning that it fluctuated based on our account spend. What I found was that, as a large enterprise account, we were always at the top of our rebate bracket (1.87%). This immediately caught my attention as I happened to already know that the Expensify credit card (back in 2022, at least) was a flat 2% rebate regardless of spend. That may not seem like a significant difference, but when an enterprise account has tens of millions of dollars in expenses, it adds up quick. Ding ding ding— VALUE. There was now a real opportunity to switch systems, and pay for the project just with a rebate bump through changing credit card providers.

But it wasn't enough. With my sights now set on a potential project, I decided I needed to turn over as many stones as possible and get real with every facet of what I was about to propose. What were all the costs of switching, and how much money could be saved by efficiency gains and process changes?

This moment brought me back to the purchasing policy problem. I knew we were underutilizing our company credit cards and that the company had an affinity for PO's. I enlisted some help from our highly skilled data analyst and asked him to create a report that showed all our PO spend that, by threshold, should have been purchased via company card. The results floored me. I found that 56% of every PO submitted since 2021… Should have been purchased with a company card. 68.5% of those individuals had a credit card and still chose to submit a PO.

After doing some more interviewing with AP processors that approve our PO's, and calculating all the labor spent submitting and processing PO's, I had identified an opportunity to save $883,234 every year, simply by enforcing our existing credit card policy and shifting our spend.

My fourth step: Communicate.

Now here's the kicker. Switching expense systems would have netted another $100k a year — real money, but it became clear the majority of the value didn't live where I'd assumed it would. So I tucked away my attachment to the idea I'd started with and built the pitch around the part that actually mattered: a change in how we pay for things. The system switch got demoted to an optional side quest.

I took it up the chain. The analysis was sound, the number was big, and nobody I showed it to really argued with it.

And then nothing happened.

Not because anyone disagreed — because no one whose call it was wanted to make it. The CTO was straight with me: this wasn't his decision. The person it was didn't make it either. The project went on a shelf, and I went back to my day job.

Three years later, the company brought in an outside contractor to review operations. One of the first things they flagged — as one of the most obvious problems they found — was the exact issue I'd handed over for free. Same finding, same fix. They just had the standing to say it out loud.

What I learned:

I took this on out of curiosity and a desire to connect — an excuse to use some muscles that were asking for work. The $883k was the headline. The part that stuck with me was quieter.

Remember the Bystander Effect? Less likely to act, assuming someone else has made — or will make — the decision. I'd spent the entire project refusing to be a bystander. I didn't expect to watch the phenomenon play out one floor up, where responsibility was diffuse enough that the obvious call just never got made. It didn't need a villain. It only needed everyone to assume it belonged to someone else.

So I don't file the shelving under "loss." Good analysis doesn't expire — it waits for someone with the standing to act on it. I trusted my read, did the work, and took the shot. The shot was good; it just took an outsider three years to say so. Given the choice between taking the swing and assuming someone else will, I'll take the swing every time.

Thanks for reading!
Keelan