The Operational Cost That Never Shows Up in Any Report

When I first started talking with Supply Chain teams in retail, I expected the conversations to revolve around KPIs: fill rate, service level, inventory turnover, lead times.

But there was one phrase I kept hearing over and over again:

"The numbers are there. But I don't know how much longer we can keep working like this."

That's when I realized the problem isn't always in the indicators. Most of the time, it's in everything that happens before you ever get to them.

The KPI Gets Hit. At What Cost?

Fill rate can be at 94%. Service level can be met week after week. Supplier-to-DC lead times can be, in theory, under control. Everything can look "fine" on the dashboard you present to the director.

But ask the distribution analyst how they got there, and the story is almost always the same:

They open the ERP to check current stock by store. They cross-reference that against the WMS to validate what's in transit. They pull up the purchasing team's spreadsheet to see which purchase orders (POs) are still open with suppliers. They manually verify each supplier's delivery lead time, because "the system doesn't always have it up to date." They calculate safety stock by category. They adjust replenishment orders for stores with unusual demand patterns. And at the end of the day, they check whether service level held.

By the time they've pulled all that information together, several hours of the workday are gone — and they still haven't made a single decision. They've only prepared the data so they can start analyzing.

That process, which takes between 4 and 6 hours, repeats every single day. If an analyst spends 4 to 6 hours a day preparing information, they end up dedicating more than 50% of their workday to consolidating data before they can even begin to analyze it.

The KPI gets hit. But 80% of the analyst's time went into consolidating information — not analyzing, not anticipating, not optimizing.

The Problem Isn't the KPI. It's How It Gets Produced.

Distribution in retail isn't just moving product from the DC to the store. It's coordinating, simultaneously, variables that live in different systems that rarely talk to each other.

The supplier lead time determines when product will arrive at the DC.

But that data is rarely up to date.

Most of the time it lives in emails, informal agreements, or spreadsheets that someone maintains by hand.

The same goes for lead times from the DC to the store.

In many operations, that's still a historical average that no one has revisited in months.

Meanwhile, SKU-level turnover by store stays buried in reports that arrive too late to prevent a stockout.

The analyst is the human glue holding all of it together. And it works — until it doesn't.

Scattered Information Isn't Just Inefficiency. It's Risk.

When information is scattered, errors don't surface right away. They surface when there's already a stockout across twelve stores. When service level dropped three points in the week you least expected it. When you discover a supplier has been delivering late for two weeks and nobody noticed because that data was sitting in an email no one checked.

The time an analyst spends manually consolidating information isn't just wasted time. It's time that wasn't spent detecting anomalies, anticipating stockouts, or optimizing distribution across store formats.

And there's another cost that never appears in the reports: the cost of error. When you calculate by hand for six hours, a misplaced decimal or a formula that didn't copy down correctly can result in overstock in some stores and stockouts in others. Both scenarios cost money.

What Changes When You Distribute Based on Real Store Demand

Most distribution operations I've seen replenish on a cycle: every Monday, every week, every time stock drops below a fixed threshold. It's predictable, it's operationally comfortable, and it's almost always disconnected from what's actually happening in the store.

Demand-driven distribution works differently. The system understands what is truly being demanded at the store level, accounting for stockouts, promotions, and seasonality.

Instead of replenishing by calendar or fixed rules, it considers the specific behavior of each store: demand patterns, promotions, seasonality, operational constraints, and replenishment lead times.

With that information, it generates replenishment recommendations aligned with inventory policies and the reality of each point of sale.

The result isn't just better replenishment.

It's an operation that responds to what is happening today — not to what happened last week.

The result isn't just a better fill rate. It's a fill rate that gets anticipated, not one that gets explained after it already dropped.

What I See in the Operations That Work Well

Operations that manage to move from a reactive model to a proactive one tend to have one thing in common.

They centralized critical planning information and automated repetitive replenishment decisions.

That allowed them to reduce the time spent consolidating data and shift their analysts' focus toward managing exceptions, interpreting demand, and making higher-impact decisions.

That is precisely what we see working alongside retailers across different segments in Latin America.

That's exactly where Celes comes in.

The Next Step

If one of your analysts was out for a week starting tomorrow, would the operation keep running the same way?

If the answer is no, operational knowledge is probably still too dependent on people and not enough on process.

That is one of the highest — and least visible — costs in Supply Chain.

If you want to identify where it's happening in your operation, let's talk.

At Celes, we help supply and distribution teams reduce that operational cost through replenishment decisions driven by real demand and intelligent automation.

---

Valentina Alarcón is a Supply Chain Transformation Consultant at Celes, specializing in distribution and replenishment operations for retailers across Latin America.