How to Automate a Shoe Shop or Clothing Store

What can you automate in a shoe shop or clothing store?

In a fashion or footwear shop you automate five things: stock by size and colour across the shop floor and the website, the alert that tells a customer their size is back, returns and size exchanges, replenishment of the middle sizes, and sale-season repricing. What you don't automate is the advice given at the fitting room, which is exactly where the margin sits.

The difference from any other retail business is that here a product isn't a product: it's a matrix. One ankle boot in eight sizes and three colours is twenty-four separate references, each with its own counter. A shop carrying two hundred models in season is handling several thousand live references. Everything that another sector solves with a spreadsheet breaks here.

Daily task	Who does it today	What changes once automated
Deducting a shop-floor sale from online stock	By hand, at cash-up	The till does it in seconds, without waiting for close of day
Telling a customer their size is back	Nobody: the customer is lost	Automatic alert to whoever asked, in order of request
Handling an online return	Loose emails and a hand-written label	Returns portal, label issued, stock restored once checked
Spotting that the middle sizes have gone	When someone complains at the fitting room	Alert per model as soon as the size curve breaks
Changing prices for the sales	Reference by reference, in two places	One rule per season applied to every channel at once

Why a fashion shop is harder to automate than other retail

Because stock doesn't run out all at once: it runs out from the middle. The mid sizes of a model sell first, and once only the smallest and the largest are left, the system still reports units in stock. The model shows as available, takes up floor space and in practice can no longer be sold. That's what a broken size curve means.

A single counter per model sees none of this. Which is why the first technical decision in a shoe shop isn't which tool to use, but at what level stock is counted:

Model level: "Black Chelsea boot: 12 units". Fine for the shop window, useless for selling. It's the most common mistake at the start.
Reference level (model + colour + size): "Black Chelsea boot, size 40: 2 units". It's the only level at which anything can genuinely be automated.
Location level: the same reference, split between shop floor, stockroom and the units held aside in the fitting room or reserved for a customer.

Everything that follows — the alerts, the returns, the replenishment — depends on getting this level right from the outset. Changing it once the shop is running means relabelling the entire catalogue.

Step 1: tidy the catalogue before automating anything

Before connecting anything, the catalogue has to reach a state a system can trust. It's the dullest step and the one that decides whether the rest works. Specifically:

One code per reference, identical everywhere. If the boot is BCH-BLK-40 on the till and chelsea-black-40 on the website, no system will recognise them as the same item. The supplier's code is usually the best starting point, because it already comes printed on the boxes.
One single way of writing sizes. Don't let 40, "40 EU" and "6.5 UK" coexist in the same catalogue. If you sell abroad, the conversion is stored as product data, not baked into the name.
Colours from a closed list. "Black", "matte black" and "negro" end up as three different products for the system and one for the customer.
Photos and descriptions before the season, not during it. This is the real bottleneck in any fashion shop: the stock arrives, goes on the floor the same day, and takes three weeks to appear online because nobody has had time to photograph it.

Skip this step and what gets automated is the mess — only faster.

Step 2: connect the till to the online shop

This is the connection that recovers the most money, because it's the one that stops you selling online what has already gone over the counter. The till sends every sale to the central inventory, and that inventory updates the website and the marketplaces within seconds, reference by reference.

How it's built depends on what the shop already runs:

If your till has a native integration with your ecommerce (a till that already talks to Shopify or WooCommerce, say), you switch it on and check that it syncs at reference level rather than model level. Some integrations only send the total.
If the till is older or came from a sector-specific supplier, it will normally have a sales export or some way of querying its database. That's where a middleware like n8n comes in: it reads the till's movements and updates everything else without replacing the system the shop already works with every day.
If there's more than one shop, each location counts separately and you have to decide which stock goes online and which branch fulfils each order. That's covered in detail in the guide to syncing stock across channels.

The check you can't skip: sell one unit at the counter and watch the clock until it disappears from the website. If it takes more than a minute, you'll oversell during the sales.

Step 3: tell the customer when their size is back

Of everything you can automate in a fashion shop, this is the one that sells most and almost nobody has. When a reference runs out, the product page offers an alert by email or WhatsApp; when that specific size comes back into stock, the alert goes out on its own, in order of request and only to the people who asked for that size.

It works because the customer had already decided to buy: the only reason they didn't was that their number wasn't there. That's demand being lost today without leaving a trace. And it produces data no other source gives you: which sizes and models are being asked for and can't be supplied, which is exactly the information you need when placing the next supplier order.

Two details decide whether it works or annoys people:

The alert goes out when the unit is genuinely available, not when it appears on a delivery note. Telling someone about something they can't buy yet burns the alert.
If a single unit arrives and five people are waiting, you alert in order and give the first one a window to reserve. Alerting all five at once leaves four angry customers.

Step 4: returns and size exchanges without loose emails

In fashion a return isn't an incident: it's part of how people buy. Plenty of customers order two sizes knowing they'll send one back. A system that treats every return as an exception jams the moment the online shop starts working.

What gets automated here:

Starting the return, from a portal where the customer picks the item and the reason, without writing an email or waiting for a reply.
The shipping label, issued on the spot against the carrier.
The size exchange treated as an exchange, not as a refund plus a new order. This is what actually saves the sale: if the customer has to request a refund and buy again, plenty of them simply don't come back.
Stock restored only after the garment has been checked. Returned isn't the same as sellable. If the system restores it the moment the parcel arrives, you end up selling something online that wasn't in a fit state.

And the return reason, stored as data: when one particular model keeps coming back "because it runs small", that gets fixed on the product page and stops happening.

Step 5: replenishment and end of season

Once stock is counted by reference, replenishment stops being a walk around the shelves. The system flags a model when it loses its middle sizes — the point at which it still occupies the shop window but can no longer sell — and groups what needs ordering by supplier instead of firing off individual alerts.

End of season is the other half. Markdowns are applied through a rule — by season, by product family or by how long the item has been in stock — and that rule propagates to the shop floor and every online channel at the same time. Doing it reference by reference in two separate places is where mismatched prices between the window and the website come from, which is a customer complaint and, depending on where you sell, a consumer-law problem.

What order to do it in

You don't build it all at once. This is the lowest-risk order, and each step only begins once the previous one has run for a few days without surprises:

Order	What you do	Why it sits here
1	Catalogue tidied by reference	Without it, everything else automates errors
2	Till connected to online stock	It's what prevents overselling, the costliest failure
3	Back-in-stock size alerts	First thing that creates new sales, not just savings
4	Returns and exchanges	Becomes urgent as online volume grows
5	Replenishment and markdown rules	Needs months of your own data to tune

Mistakes that ruin a fashion shop's automation

Counting stock by model instead of by size. It's the original sin. Everything else is built on top of it and has to be redone from scratch.
Publishing shop-floor stock online without holding a buffer back. Between someone picking up a pair to try on and reaching the till there are minutes in which that pair is sold on the website and physically in the fitting room.
Automating alerts before the stock figure is reliable. A "your size is back" message based on a wrong number is worse than no message at all.
Treating a size exchange as a return. It turns a saved sale into a lost one.
Restoring returned items to sellable stock on arrival, before anyone has looked at the garment.
Building it all during peak season. Any inventory change gets tested in the quiet part of the year, not the week before the sales.

Run a clothing or footwear shop and want to start with what matters most?

At Zerolagia we build the full circuit for a fashion shop: stock by size and colour between the counter and the website, back-in-stock size alerts, returns and exchanges, and replenishment driven by the size curve. And you don't have to buy it blind — you can watch it running on real data before deciding anything.

See how we do it

Frequently asked questions

Is it worth automating a small, single-location shoe shop?

Yes, but not all of it. In a shop with no online sales, what pays back fastest is counting stock by size and colour and alerting customers when their number returns: those are the two things that recover sales being lost today. Automated returns and multichannel markdown rules only pay off once there is online volume. Automating a small shop all at once usually ends in a system nobody uses.

How do you manage stock by size and colour without losing your mind?

By counting every combination of model, colour and size as an independent reference with its own code, and using that same code on the till, in the online shop and in your management system. The model becomes a grouping used to display products on the website, not the unit you count with. At that level of detail you can spot broken size curves and trigger replenishment for a specific size, which is where the money is.

Do you need to change till systems to automate a shop?

Not necessarily. If the current till can export sales or allows its database to be queried, it can be connected through a middleware without being replaced. Changing till systems is the most expensive decision and the most disruptive to daily trading, so it stays as a last resort: only when the current system won't release the data in any form, or the supplier no longer maintains it.

What happens to returns before an item goes back on sale?

Returned stock is kept separate from sellable stock until somebody inspects the garment. The system needs at least three states: available, under review and not sellable. Only the first is published to sales channels. If stock is restored automatically when the parcel arrives, sooner or later you sell something online that wasn't in a fit state, and that sale costs far more than the one it saved.

How long does it take to automate a fashion shop?

It depends on the state of the catalogue far more than on the technology. If references are already coded properly, connecting the till to the online shop and setting up size alerts takes weeks. If the catalogue has different codes in each system, sizes written several ways and colours that were never standardised, tidying it takes longer than everything else combined. That's why the catalogue is always step one.

How to automate a shoe shop or clothing store, step by step