Anatomy of an Operations Hub: stock, orders and suppliers in a single panel — with the numbers on what not having one costs.
by Nadia Maar ·
There is a precise moment when growth stops being good news. Orders double, revenue too — and suddenly the warehouse says one thing, the website says another, and a customer gets a confirmation email for a product that does not exist. Nobody made a mistake. The system simply ran out of road.
That system, nine times out of ten, is a spreadsheet. It worked beautifully up to a certain volume, and it keeps working just well enough not to look like the problem. This article is about where it actually breaks, why an off-the-shelf CRM often makes things worse rather than better, and what a panel built around ONE company's processes — instead of the average of ten thousand — looks like on the inside.
A spreadsheet does not collapse. It degrades, and every single step down looks acceptable until you line them all up. The signal is not “the file got corrupted”: it is that people have stopped trusting it and started checking by hand.
Take the time one person spends moving data between systems, multiply by their hourly cost and by twelve months. In most companies I talk to, that figure alone exceeds the cost of building the panel that would remove it. And it excludes orders cancelled through stock-outs, which is the part that hurts most because it only ever shows up in the margin.
The obvious answer to chaos is to buy a product: a CRM, an ERP, an all-in-one platform. Sometimes that is the right answer, and when it is I say so. But on specific operational processes the opposite of what you hoped for tends to happen.
A general-purpose product is the average of ten thousand companies. Your competitive advantage, almost by definition, lives in what you do differently from that average — and that exact part is what the product does not anticipate. From there come the three symptoms I always see together:
The rule I use to decide is simple: if the process is standard — e-invoicing, accounting, email — you buy, always. If the process is how the company makes money, you build. Buying your own difference is the surest way to lose it.
An operations dashboard is not “a website with charts”. It is a layer between the systems you already have — store, ERP, suppliers, courier — and the people who have to decide. Its job is to hold one version of the facts and show each person the part they need.
The figure at the top is not a photograph taken last night: it changes when the warehouse changes. The browser holds an open connection to the database and receives the update, instead of re-requesting the page every minute. Whoever is taking an order watches the last unit disappear as they look at it — the only moment when knowing is useful.
The margin is net: purchase cost, shipping and expected returns already in. It is computed in the database on every line, not at month end in a spreadsheet. It is also the number that most often changes a commercial decision — and the one nobody looks at when it arrives thirty days late.
The warehouse sees this row without the costs; management sees it with them. The rule does not live in the interface: it lives in the database, which simply will not return that column to an unauthorised user. A screen can be worked around by reading the network; a rule in the engine cannot.
Every bar is made of recorded events: who changed what, when, from what value. It is not there to police people — it is there so that when a number is wrong, the cause is traced in two minutes instead of two days of phone calls.
Under the hood the structure is always the same, and every piece is chosen for an operational reason, not an aesthetic one:
The other half: pipeline, margin per order, fulfilment times. The same data as the supplier portal, read from the selling side instead of the supplying side.
Not all custom software is worth its cost. These four are, and you recognise them because they remove work rather than add it.
This is the typical shape of the result on an e-commerce business between three and ten million in revenue, with a catalogue fed by several suppliers. Treat them as orders of magnitude to run against your own data, not as a certification: your numbers get measured together during the audit, which is exactly what the audit exists to produce.
| Before | After | |
|---|---|---|
| Fulfilling one order | ~15 minutes, three systems open | ~30 seconds, one screen |
| Stock alignment | 1–2 times a day, by hand | Continuous, automatic |
| Errors from misalignment | Routine, found by the customer | Near zero: the conflict is blocked first |
| Hours/week moving data | 10–15 | Under 1, exceptions only |
| Margin per order | Estimated at month end | Visible on every line, immediately |
| Who can see purchase costs | Anyone who opens the file | Only authorised roles, enforced by the database |
Payback does not come from saving on licences: it comes from the hours that stop being spent and the orders that stop being cancelled. Both are things the company already pays for every month — they just never appear on an invoice, which is why they stay invisible until somebody measures them.
If any of this looks like your week, the question is not what a custom panel costs: it is what not having one already costs you. That number exists, and it can be worked out in half an hour by looking at how orders, stock and suppliers run today.
The technical audit is free and it is not a presentation: we look at your processes, find the two or three points where most time is lost, and out comes a cost and payback estimate. If the conclusion is that you just need to configure what you already have properly, I will tell you — it has happened, and it cost me one project and gained me one relationship.
When the file is no longer a reflection of what is happening but its only copy, and two people open it at once. The signals: stock has to be checked in two places before answering a customer, the restock order comes from memory rather than data, and only one person really knows the file. From there the cost is not the spreadsheet, it is the time needed to trust what it says.
An off-the-shelf CRM solves the sales side and leaves out stock, purchasing and suppliers — which is where the work is. The frequent result is a team updating two systems instead of one, and double entry doubles the work instead of halving it. Custom is worth it when the processes costing you time are not the ones the ready-made product was designed to solve.
A single panel where stock, orders and suppliers are the same database, not three exports reconciled by hand. It does not add a system: it removes one. The value is not the look of the dashboard, it is that «how many do I have, where are they, when do they arrive» has one answer — and that anyone can read it, not only the person who built the file.
In the case described in the article operating cost fell by 35%, but the number to look at is a different one: how many hours a week go today into re-typing, reconciling and verifying data that already exists. Those hours are the measure of the problem, and the only honest basis for deciding whether a project like this pays for itself in your case.