This is the checklist we run before recommending any tool to a shop. It applies whether you end up buying a connector, building a custom piece, or mixing both.
Our short answer: you almost never need new software
We say it on our retail page and it holds here: automating a shop does not require buying separate automation software. Automation is built on top of what you already use — online shop, marketplace, till, invoicing, email — by connecting them to each other. No migration, no extra dashboard to learn.
That said, plenty of shops arrive with a licence already signed, or with a connector their supplier recommended, and the question they bring is “does this actually help me?”. The rest of this guide is for those cases: what to check before paying, and how to tell a tool that really fixes the mismatch from one that only displays it on a dashboard.
The question almost nobody asks first: who is in charge of stock?
A shop selling through three channels has three stock figures, and they are almost never the same. Before paying for anything you have to decide which one wins when they disagree: the till in the shop, the online store, or a central system. That is what we call a single source of truth, and it is a business decision, not a technical one.
If that decision has not been made, every piece of software will do the same thing: overwrite one number with another number, in whatever order the notifications arrive. And the result you will see is the familiar one — selling what you no longer have.
If you are going to pay for a licence anyway, the seven things to check
1. Does it write, or only read? Plenty of integrations “sync” in one direction only: they read stock from the online store and show it on a dashboard. That is a report, not a sync. Ask explicitly whether it writes back to every channel, and into which fields.
2. What triggers it, and how often? Reacting the moment a sale happens (because the channel notifies it) is not the same as checking every fifteen minutes or every hour to see whether something changed. In peak season, the gap between those two is exactly where double sales get in. Ask in writing for the real minimum interval and whether the plan caps API calls.
3. Does it understand what your catalogue actually looks like? This is where many tools fall over: variants (size and colour), bundles and kits that consume units from several products, the same item carrying a different reference in each channel, and more than one warehouse. If you sell fashion, the size matrix is not a detail — it is the whole catalogue.
4. What exactly happens when it fails? A channel goes down, an API stops answering, an order arrives twice. The good questions are: does it retry? does it queue what it could not send? does it tell a human when something is stuck, or does it disappear into a log nobody reads? A system that fails silently is worse than no system, because you stop checking by hand.
5. Who owns the data, and what does leaving cost? Full export of catalogue and history, API access, and what you take with you the day you change supplier. If the answer is vague, the exit cost is high.
6. What does it cost once you grow? Many prices scale by orders, by connected channel or by number of references. The honest comparison is not this month’s fee: it is the fee with the catalogue and the order volume you expect a year from now.
7. What stays manual? No tool does everything. A supplier who tells you in writing what remains manual — creating products, photos, returns, stocktakes — is the most reliable sign that they know the job.
Signs a licence will not work for you
- The demo runs on their sample catalogue instead of two or three of your real references, awkward ones included.
- Nobody asks about your returns. Stock coming back is a big part of the problem in fashion and footwear.
- There is no way to see, a month later, how many mismatches were corrected and why.
- They sell “AI” without saying what the machine decides and what a person decides. On the stock side most of the work is well-placed rules and alerts.
Stock sync, PIM and ERP are not the same thing
Three categories advertised with the same words:
- Stock sync: moves units and prices between channels. It solves overselling.
- PIM (catalogue manager): stores and publishes product information — names, attributes, photos, translations — and stops each channel from keeping its own version of the listing.
- ERP: runs purchasing, warehouse and accounting. The biggest, and the most expensive to replace.
Buying one and expecting what another does is a very common and costly mistake. If the pain is “I sell what I don’t have”, start with the first; if the pain is “every channel shows a different listing”, the work is in the second. We approach it as one process automated at a time, not as replacing a whole system.
A one-week pilot instead of a one-year licence
A small pilot answers in a week what a demo never answers:
- One channel and one product family. The family that drifts the most, not the easiest one.
- Two numbers written down before you start: how many stock corrections you make by hand in a week, and how many orders you cancel for lack of stock. Without those two you will not be able to say whether it worked.
- One week with the system actually writing, not in read-only mode.
- The same two numbers at the end, and the decision taken with them in front of you.
That is the order we always follow: first a diagnosis of one specific process, then the implementation, and the system running in plain sight before anyone signs anything long.
Frequently asked questions
My cash register is old — does that rule me out?
Almost never. It only has to let stock movements be read and written from outside, through an API or a scheduled export. If it cannot even do that, the middle path is to leave it where it is and build the central stock above it, with the counter deducting by hand only what cannot report itself.
Will an Excel sheet hold up as the reference stock?
As a starting point, yes, and it beats three disconnected figures. It stops holding up the moment two people edit it at once, or the order volume means it has to update without anyone opening it.
Is a PIM the same as a stock connector?
No. The connector moves units and prices. The PIM governs product information — names, attributes, photos, translations. A shop that oversells needs the first; a shop whose listings disagree from one channel to the next needs the second.
Can I start with one channel and add the rest later?
That is the recommended way. Each channel has its own rules for orders, returns and references, and doing them all at once multiplies the places to look when something goes wrong.
How do I tell whether the mismatch starts in the channel or in my warehouse?
By when it happens. If the figure goes wrong right after a sale or a return, the problem is how the notification travels between channels. If it is wrong all the time, even on days with no sales, what is broken is the physical count, and no licence fixes that.
Tell us the case on our contact page and we will look at it with your real channels. The technical step by step for the sync itself is in this guide.
You might also like