Step by stepNovember 19, 20266 min read

432 products got a new code, and 415 silently lost their category

A pharmacy group recoded 432 products and 415 lost their category with no warning on screen. What happened, why it is so easy to happen to you too, and how to check it.

Pol

Fundador de AutoBoost

Product catalogBusiness data
432 products got a new code, and 415 silently lost their category

If products in your business occasionally get a new code (a recoding, a catalog merger, a listing that had to be fixed), something may be happening that never shows up on any screen: the product keeps selling normally, but it loses the category it had assigned. There is no error. One day it simply stops showing up in the report for its product family, as if it never had one.

What happened when a pharmacy group recoded 432 products

In the management software this pharmacy group uses, a product's category (the one that feeds reports and family breakdowns: prescription drugs, over the counter, skincare) is not attached to the product itself. It is attached to its numeric code. As long as that code never changes, you would never notice: the category stays there, quietly doing its job.

The problem shows up when a product gets a new code, something that happens fairly often for ordinary day to day reasons: a mistake when the product was first listed, a change of supplier, a merge of duplicate references. The new code is born clean, with no category attached. And the management software, at the same time, deletes the record for the old code: its category disappears with it, for good.

When the group looked into it closely, they had recoded 432 products. Of those, the system had only kept the category for 17. The other 415 had quietly lost their classification without anyone noticing, because nothing on any screen flagged that gap: the product kept selling, kept ringing up at the register, kept showing up normally. The only thing missing was its family, and that only becomes visible if someone builds a report by family and wonders why the total does not add up. After the fix, 419 of the 432 products got their real category back.

Why does a product lose its category when its code changes?

Because the data (the category) lived attached to an identifier (the code) that the system itself treats as disposable. The moment that identifier changes, the source system does not move the category over to the new code: it simply loses it, because as far as the system is concerned the old code no longer exists. This is not a rare glitch in one particular piece of software. It is the normal behavior of any system where important data hangs off a number that can change: a product reference, a customer code, a contract number.

You do not need to sell medicine for this to happen to you. A hardware store that renumbers its catalog after switching distributor, a wholesaler that merges two duplicate codes for the same item, a training academy that merges two records for the same student: in all three cases, if something (a category, a sales history, a customer type) was attached only to the old identifier, it can disappear the moment that identifier changes, with nothing flagging it as an error.

How does this show up day to day, before anyone builds a new report?

It almost never shows up directly. It shows up as a figure that, month after month, looks a bit low for one product family, or as a total that does not quite add up when someone manually sums the categories and compares them to the overall total. It is exactly the kind of mismatch that, without a report crossing both numbers, stays disguised as "normal variation" for months.

The decision rule: does your history survive a code change?

If an important piece of data in your business is attached to an identifier that can change (a product code, a reference number, a customer ID), ask what happens to that data the day the identifier changes. If the answer is "it gets lost" or "nobody knows", you do not have a history: you have a history that works as long as nobody recodes anything.

In the pharmacy group's case, the fix was not changing how each product's category gets calculated. It was saving a plain snapshot of the day's classification every night and merging it with the history of past categories, so that a retired code keeps the category it had while it was alive, even after the source system has forgotten about it. The data stops depending on the source system choosing to keep it.

How do you check this in your own business?

You do not need any new software to start looking. With what you already have:

  • Find a product, customer or contract in your catalog that changed code or reference in the last few months, and check whether it kept its category, its sales history or its customer type. If it lost it without anyone noticing at the time, that data does not survive an identifier change in your system.
  • Ask whoever maintains your management software directly: "if I recode this product today, what happens to its category?" It is a question almost nobody asks, because recoding feels like a minor chore, not a change that could erase anything.
  • If you have a report by family, by customer type or by contract category, compare its total to the overall total for the same metric. If they do not match and the gap has no clear explanation, that gap may be exactly the history that some code change quietly took with it.

This is the same work we do when we bring the data from several of a business's programs together in one place: it is not enough for the data to arrive correctly today, you also have to decide what happens to it when the source system changes something on its own. You can see what that looks like in the case of a data platform for a pharmacy group.

If you want to check whether some important piece of data in your business silently depends on a code that can change, in our services we explain how we approach this with the data you already have. And if you would rather talk it through directly, reach out and we will look at it with your own catalog.

Share article