About ImportReady

1. Why this exists

We tried to write the import rules from the platform’s own documentation. Twice. Both rules were wrong.

Not subtly wrong — wrong in the direction that costs a merchant a catalogue. So we stopped reading documentation and broke a real store on purpose instead, recording every word it said and every product it quietly dropped. That archive is where all 14 of the import error pages come from, and it is where this product’s entire position comes from.

The finding that decided everything else: 10 of those 14 failure modes produced no error message at all. The import reports success, the catalogue is wrong, and there is nothing to paste into a search box. Those are the expensive ones, and nothing on the market was looking for them before the upload.

2. Who builds it

Haryz Adil, trading as ImportReady, operating under the law of Morocco. One person, not a company with a team page — and saying so plainly is better than a photograph of an office we do not have.

You can write to me at contact@getimportready.com. A person reads that inbox, because there is only a person.

3. How we decide what is true

This product’s only real claim is that it tells you the truth about what it did. That is not a slogan we can enforce with good intentions, so it is enforced with rules instead:

• 15 recorded observations from a real broken store, and a test in our own build that refuses to publish a page quoting an error string that is not in that archive.
• A corpus of 21 real supplier files — 19,885 rows, 11,405 products, 5 to 614 columns wide — that our engine is re-run against on every change. It found four defects that 140 passing hand-written tests did not.
• Baselines that record what the code actually does, not what we wish it did. One fixture is pinned as failing, because that is the honest current answer and hiding it would mean nobody notices when it is fixed.
• A published negative result. We swept 546 open-data sources looking for supplier catalogues, inspected 250 files, and found zero usable ones. The well was wrong, not the filter. It is written down because a method that only reports its successes is not a method.

4. What we will not do

We will not tell you an import is guaranteed to work. Nobody honestly can. Whether your import succeeds depends on your store, your image host and the platform’s current spec — three things we do not control. What we can be held to is finding and naming every blocker in the file, which is a promise about our own rule engine, and one you can test against a file you already know.

We will not sell to you if you should not buy. Under about a hundred products, a chat window will genuinely do this job, and the pricing page says so. Selling a subscription to somebody who does not need it buys six weeks and costs the only thing this business runs on.

We will not invent a customer. There are no testimonials on this site, no logo wall and no case studies, because there are no customers yet. When there are, they will appear with names attached and not before.

5. Where it is going

The plan is not to become a platform. It is to be the layer that runs before your data reaches one: read a supplier file, normalise it, tell you what will break, mark whatever had to be inferred, and hand back something you can upload. Everything else is a distraction from that.

Run your worst supplier file through it. It is free, it needs no account, and it will tell you more about whether this is worth your time than this page can.