22 May 2026

Your VC money is leaking through workarounds

Here is a question nobody asks in board meetings: how many hours this week did your team spend copying data from one system to another?

Not building. Not selling. Not improving the product. Just moving information from point A to point B because two systems do not talk to each other. The answer, in most VC-backed e-commerce companies doing 50 to 500 orders a day, is somewhere between "a lot" and "we don't want to know."

The Debt Nobody Tracks

Technical debt has a name. People write blog posts about it. Engineering teams allocate sprints to pay it down. But there is another kind of debt that grows silently in scaling companies. Operational debt. Processes that worked at 20 orders a day, held together by capable people who knew the workarounds.

A returns process that requires someone to check the order in Shopify, update the status in the ERP, and send a manual email to the customer. Three systems, three manual steps, one person who knows the sequence. Pricing logic that lives in a spreadsheet maintained by someone who is on parental leave next month. Customer communication that gets sent when someone remembers to send it.

These are not failures. They are adaptations. Smart people solving problems with the tools they have. And at low volume, they work. The problem is that they do not scale linearly. At 50 orders a day, one person handles returns in an hour. At 500 orders a day, you do not need ten people. You need a different approach entirely.

The Workaround Multiplier

Every workaround has four cost layers, and most companies only see the first one.

Time is the visible cost. Someone spends 30 minutes a day on a manual task. That is 2.5 hours a week. Annoying but manageable.

Errors are the second layer. Manual processes produce mistakes. A price that does not get updated. A return that falls through the cracks. A customer who waits three days for a response that should have been automatic. Each error creates its own mini-project of investigation, correction, and apology.

Waiting is the third layer, and the most expensive one nobody measures. When a process depends on a specific person doing a specific thing, everything downstream waits for that person. The warehouse waits for the return authorization. The finance team waits for the reconciliation. The customer waits for the refund. This is not idle time that shows up on a timesheet. It is latency baked into your operations.

Person-dependency is the fourth layer. When the process lives in someone's head, that person cannot be sick, cannot go on vacation without a handover document, and cannot be promoted without training a replacement. Your organization becomes fragile in ways that do not show up until someone leaves.

We use a cost calculation model that maps all four layers for each manual process. The results consistently surprise people. A task that looks like "30 minutes a day" often costs the equivalent of a half-time position when you account for errors, waiting, and the risk of the person leaving.

Start With the Biggest Leak

The instinct when you see operational debt is to plan a big project. Replatform. Implement a new ERP. Buy a middleware layer. Six months of planning, six months of implementation, results in a year if you are lucky.

That instinct is wrong. Not because the big project is unnecessary, but because the bleeding does not stop while you plan.

Start with one flow. The single process that leaks the most. Maybe it is returns. Maybe it is product data updates between your PIM and Shopify. Maybe it is customer communication after purchase. Pick the one where the cost model shows the biggest gap between what you spend and what you should spend.

Build a solution for that one flow. Connect it to your existing systems. An entry-level integration might touch 5 data sources with 1 controlled write-back. It does not replace anything. It bridges the gaps where your people currently serve as human middleware.

One company replaced their manual post-purchase email process with automated customer communication across 6 trigger types. Order confirmation, shipping update, delivery notification, review request, return instructions, refund confirmation. Previously, two people shared responsibility for sending these manually. "Manually" meaning they checked a dashboard, copied order details, and sent templated emails from a shared inbox. The automation was live in weeks. Those two people did not lose their jobs. They started doing work that actually required human judgment.

Show the Board a Number

VCs understand metrics. They don't need an architecture diagram or a technology roadmap. They need a before and after.

Before: returns processing costs 47 hours per week across 3 people, with a 6% error rate and an average resolution time of 4.2 days. After: returns processing costs 8 hours per week, error rate below 1%, average resolution time 1.1 days. That is a story a board understands. It is also a story that funds the next improvement.

This is how operational improvements compound. You fix one flow. You measure the result. You use that result to justify fixing the next flow. Each improvement makes your operations a little less dependent on heroic individual effort and a little more dependent on systems that work while people sleep.

Your VC money was raised to grow the business. Not to pay talented people to copy data between systems. Every workaround you eliminate gives those people back to the work that actually matters. And that, more than any single technology decision, is what separates companies that scale from companies that just get bigger.