What Retail Taught Me About Operations

When I tell people I worked at Sephora, they usually ask about the discount. (It's good.) But the thing that actually stuck with me isn't the products. It's the systems.

A Sephora store looks simple from the customer side. You walk in, browse pretty shelves, maybe get a sample, check out. What you don't see is the layer of operational complexity underneath that experience. Inventory management across thousands of SKUs. Merchandising planograms that get updated every few weeks. Staff scheduling that has to balance beauty advisor expertise, peak traffic hours, and labor budgets. Seasonal launches that require the entire floor to be reorganized overnight.

I didn't understand any of that when I started. I thought retail was straightforward: stock shelves, help customers, run the register. It took maybe two weeks for me to realize how wrong I was.

The best-run shifts were the ones where nothing dramatic happened. That's the whole point.

Operations is about systems, not heroics

My manager had this thing she said all the time: “If you're putting out fires, someone didn't plan right.” At first I thought that was harsh. Stuff happens, right? But the more I worked there, the more I understood what she meant.

Good operations means building processes that handle the predictable stuff automatically so you have bandwidth for the unpredictable stuff. When our restock cadence was dialed in, we never ran out of bestsellers. When the opening checklist was followed, the floor was ready before doors opened. When the tester protocol was maintained, customers didn't have to hunt for product samples.

The shifts that felt chaotic weren't chaotic because of bad luck. They were chaotic because someone skipped a step, or a process wasn't clearly defined, or there was no escalation path for a problem that was totally foreseeable. Operations isn't about being a hero who saves the day. It's about building a system where the day doesn't need saving.

Cross-functional coordination in a 3,000 square foot store

Here's something that surprised me: even in a single retail store, there are multiple teams that need to be aligned. Beauty advisors on the floor need to know what's launching this week. The stock team in the back needs to know what's selling fast so they can pull from overstock. Management is balancing store metrics against district targets. The district merchandising team sends planogram updates that someone has to interpret and execute on the floor.

When those groups communicate well, the store runs smoothly. When they don't, you get situations like the time we had a huge promotional display for a product that we'd actually been out of stock on for three days. That's not a supply chain failure. That's a communication failure. The stock team knew. The floor team didn't.

I started thinking about this as cross-functional coordination, even though at the time I probably would've just said “we need to talk to each other more.” Same idea, different vocabulary.

Data-driven decisions, before I knew that was a thing

Every Monday, our manager would pull the previous week's sales data and use it to make decisions. Which products to feature on the endcap. Where to place testers to maximize trial. How to handle the out-of-stock list — do we substitute with a similar product, leave the space empty, or rearrange the shelf?

She didn't call it “data-driven decision making.” She called it “looking at the numbers.” But that's exactly what it was. And the decisions she made based on those numbers were better than the ones she made on instinct. Not always, but consistently.

I remember one week she moved a mid-range skincare brand from a lower shelf to eye level based on its trending sales velocity. That brand outsold the premium product that had been at eye level for months. It wasn't a genius move. It was just paying attention to what the data was saying.

The invisible goal

The thing I keep coming back to is this: the best operations are invisible to the end user. When a customer walks into Sephora and has a great experience, they don't think about the planogram. They don't notice that the testers are freshly sanitized or that the bestselling serum is always in stock on the endcap. They just have a good time. That's the whole point.

I think about this when I look at tech operations roles, especially something like creator operations on a platform like TikTok. The goal is the same: build systems that help creators succeed without them needing to think about the system. A creator shouldn't have to figure out optimal posting times or navigate a confusing support process. That stuff should be handled by the ops team, surfaced through good tooling, and delivered in a way that feels effortless.

The parallel isn't perfect, obviously. A Sephora store serves hundreds of customers a day. TikTok serves hundreds of millions of creators and viewers. But the core principle is the same: operations done right is a multiplier. It takes friction out of the experience so that the people doing the actual work — whether that's a beauty advisor helping a customer find the right foundation or a creator filming content in their bedroom — can focus on what they do best.

What I'd bring forward

I don't put Sephora on my resume because I think retail experience is glamorous. I put it there because it taught me things that I don't think I would've learned in a classroom. How to spot a process that's breaking before it actually breaks. How to communicate across teams that have different priorities. How to use data to make decisions instead of guessing. And maybe most importantly: the discipline to build systems that work quietly in the background instead of waiting for problems and reacting to them.

That's what I want to do in a tech operations role. Not the retail part. The systems part. The cross-functional part. The making-things-invisible part.

Also, the discount was really good.

I worked at Sephora for about six months during my sophomore year. It wasn't a career-defining role, but the operational thinking I picked up there keeps showing up in unexpected places — in how I approach project management, how I think about user experience, and how I evaluate whether a process is actually working or just feels like it is.