Stop buying stock on intuition.
SKUM reads your sales, turnover, stockouts and each supplier’s real lead time, and proposes what to buy, how much and when — with the reasoning written alongside.
The problem, as it is actually lived
Over-buying and under-buying are not opposite problems. They are the same problem.
You have cash asleep in stock that does not move and, at the same time, you are losing sales on the products that do. Both, every month, in the same warehouse.
This is not a management problem, it is a volume-of-decisions problem. Past a few hundred SKUs, no one can hold each product’s turnover, each supplier’s lead time and each category’s seasonality in their head. So buying happens based on last month’s sales, on what the rep pushes, or on what is on offer.
The cost is double and almost never measured together: capital locked in what does not sell, plus the margin lost when what does sell is missing.
How it works
It connects to the data you already have. No ERP migration, no manual data entry.
Reads the history
Sales per SKU, with seasonality and trend. Not a flat average of the last few months.
Calculates turnover and coverage
How many days of sales you hold of each product, and which are above or below what is reasonable for their category.
Detects stockouts
Including the invisible ones: the days stock sat at zero so there was no sale to record. That is the number almost nobody measures.
Uses real lead time
Per supplier, measured on the orders you have actually placed — not the one their price list promises.
Proposes the order
What to buy, how much and when, with the reason for each line written and editable before you confirm.
What makes it different
It is not another report. A report tells you what happened and leaves the decision on your desk; SKUM arrives with the decision proposed and the reasoning visible.
And the business criteria are written down and adjustable. If in your industry a stockout in line A costs far more than idle capital in line B, that gets configured and shows up in the proposal — instead of being buried in a model nobody can audit.
The practical consequence: the purchasing team stops assembling the order and starts reviewing it. That is a different and much shorter job.
Who it is for, and who it is not for
It makes sense if
- You handle hundreds or thousands of SKUs and nobody can hold them in their head anymore.
- You buy from suppliers with long or variable lead times, especially if you import.
- You have at least a year of sales history in some system.
- Stockouts cost you concrete sales and you can estimate how much.
We do not recommend it if
- You have a few dozen SKUs: a well-built spreadsheet is enough and cheaper.
- Your demand is project-based, not recurring. Without a historical pattern there is nothing to learn from.
- You have no digitised history. That has to be solved first, and it is a different project.
- Your ERP already has a replenishment module you use and that works for you.
Real case
Integrations
It connects to what you already use. If your system is not listed, we look at it on the call.
- ERPs with an accessible database or API
- Excel or CSV exports, if that is what exists
- Your own database, when the project includes one
- E-commerce platforms, for the online channel
Frequently asked questions
Do I need to change ERP?
No. SKUM reads the data wherever it lives and returns the purchase proposal. If your ERP has an API or accessible database it connects directly; if not, it works from periodic exports.
How much history is needed?
A year minimum to capture seasonality. You can start with less, but the first proposals will be more conservative and need closer review.
Does it decide on its own, or does a person decide?
It proposes; a person confirms. Every order goes through human review before being issued, and each line shows why that quantity is suggested. The point is to shorten purchasing work, not remove it from the loop.
What about new products with no history?
They are handled separately, by analogy with similar products in the same category, and flagged as low-confidence estimates so the buyer pays attention to them.
Does it help if I import and lead times are months long?
That is exactly where it pays off most. The longer the lead time, the more expensive a mistake is and the more value there is in putting it into the calculation instead of estimating by eye.
Apply this to your company
We can look at whether this makes sense for your operation — and if it doesn't, we'll tell you.