Get Store Intelligence

We check every product you sell. You approve every change.

GSI sets the standard each of your categories should meet, checks every product against it, researches what is missing and proposes the fixes. Your team approves anything material before it goes live — and we verify that it did.

Scan my catalogue — freeOr talk to us first

The scan runs on your own products and returns real findings. Read-only, no obligation.

18V Cordless Impact Driver — Skin Only

SKU CID-18 · Power tools

$149.00

Battery platform
18V
Chuck
1/4in hex
In the box
Battery, charger and case
Max torque
—not stated
Weight
—not stated

The title says skin only — no battery. The inclusions field says a battery, charger and case are in the box. One of them is wrong, and either way the customer finds out after they have paid. Two more fields a buyer needs are blank.

The kind of fault a first scan turns up — usually more of them than the merchant expected.

Works with

  • Shopify
  • Neto / Maropost
  • WooCommerce
  • BigCommerce
  • Magento
300,000+
products in a single connected catalogue
Every change
reviewed by a person before it goes live
Verified
against the live store after publishing

Nobody is actually looking after your product listings.

You can probably name the last time you changed your theme, your ad accounts or your apps. Almost no merchant can name the last time someone worked through their product listings properly, category by category.

So the catalogue quietly becomes the oldest thing in the business. Suppliers change, ranges expand, staff leave, and the listings carry whatever was true three years ago.

Your store already gives you AI that can rewrite a description in four seconds. That was never the hard part.

The hard part is deciding what each listing is supposed to achieve, knowing which of your thousands of products are actually wrong, and having someone accountable for fixing them without breaking the ones that were fine.

That is a job, not a feature. It is the job we do.

A plan, with a beginning and an end.

Not an open-ended retainer where you never quite know what happened this month. Four steps, each one earning the next, and you can stop after any of them.

  1. Scan

    FreeA few days

    We check a sample of your products against the standards their categories should meet, and show you where the catalogue sits.

    • Where your products sit across the levels
    • The faults we found, with the products they are on

  2. Assessment

    $2,5002–3 weeks

    The full catalogue, scored. We define what each of your categories should require, then measure everything against it.

    • Category standards written for your store
    • Level distribution across the catalogue
    • A prioritised roadmap and a fixed quote

    Credited in full against a programme started within 30 days.

  3. Programme

    from $3,000/mo90 days

    We do the work. Findings investigated, recommendations written and reviewed, changes approved by you, published and verified.

    • An agreed set of products moved up a level
    • Your approval on anything material
    • Every change verified against the live store

  4. Maintain

    $1,200/moOngoing

    New products and supplier changes get held to the standard you paid to reach, so the catalogue stops sliding backwards.

    • New and changed products checked on arrival
    • Quarterly re-scan and report
Start with the free scan

All prices AUD, excluding GST. Programme scope and price are fixed at the assessment, not guessed beforehand.

Five levels. Your catalogue is somewhere on them right now.

Every product is scored against the standard its own category should meet — a garden hose should not be held to the same checklist as a laptop. That gives you something most stores have never had: a straight answer on how good your listings actually are, and a defined next state to move them to.

  1. 1

    Clean

    Accurate and consistent. Nothing on the product contradicts anything else on it.

  2. 2

    Complete

    Every field its category needs is filled, to the same standard right across the range.

  3. 3

    Useful

    Answers the questions that actually decide the purchase, and says plainly what is not known.

  4. 4

    Persuasive

    Sounds like your brand, and makes the case against the alternatives the buyer is weighing.

  5. 5

    Competitive

    Measured against what the buyer is comparing you to, not just against your own standard.

    In development

A catalogue almost never sits neatly on one level. A first scan usually finds it spread across two or three, with your best-selling range no better than the rest — which is the part that tends to sting.

Find out where yours sits

Free, on your own products, no obligation.

Nigel Frost

Founder. Builds the software, runs the engagements, and is who you will actually be speaking to.

It stays your store. We do the work on it.

GSI is not a subscription you log into and operate yourself. The software exists because doing this properly across tens of thousands of products is impossible by hand — but the software is the tool, not the service. Three parties do the work.

Ecommerce specialists
Work out what each listing needs to achieve. Buyer questions, category conventions, how your products should differ from the alternatives.
AI engineers
Build, test and refine the instructions that do the checking at catalogue scale, then validate the output is worth putting in front of you.
You
Own the store and the final say. Supply the product knowledge only your team has, and approve anything material before it reaches a customer.

AI does not mark its own homework. Nothing material reaches your live store without a person having looked at it and you having said yes.

Every change comes with its receipt.

This is the record behind the fault at the top of this page. Not a report written afterwards — the actual trail, kept as the work happened.

It matters because in eight months, when someone asks why a listing says what it says, there is an answer. That is the difference between a catalogue that was optimised and one that was rewritten.

Illustration of the real workflow. The product and the people are generic.

  1. Found

    Category audit on Power tools. The title said skin only; the inclusions field listed a battery, charger and case.

  2. Researched

    Checked against the supplier specification and the product photography, both of which showed the tool on its own.

  3. Proposed

    A corrected inclusions field, plus the two blank fields filled from the supplier sheet — written as a suggestion, with the evidence attached.

  4. Validated

    Reviewed by a specialist against the sibling listings in the same range, to be sure the kit version still read differently.

  5. Approved

    Store operations lead: “Correct — the kit is a separate SKU. Make sure the two never read the same.” Applied across the range.

  6. Verified

    Published, then the live store read back and compared to the approved value. Match confirmed. A later change made outside GSI reopens the record rather than overwriting it.

The questions everyone asks.

Do you need access to my store?
We agree that before anything runs. For some stores the live product pages are enough; for others a read-only connection gives a much more accurate picture. Either way it is read-only at the scan and assessment stage — nothing is changed until you have approved it.
What if we do not have the product information you need?
That is the usual situation, and it is half the work. We research what can be established from suppliers and specifications, then give your team a short, specific list of questions only they can answer. Products stay flagged rather than being quietly filled in with something plausible.
Will this sound like AI wrote it?
The instructions are written and tested by our engineers against your categories and your voice, and the output is reviewed by a person before you ever see it. If a recommendation reads badly, it does not reach your approval queue.
Can we just do it ourselves with Shopify's AI?
For one product, genuinely yes. The difficulty is deciding what matters across thousands of them, knowing which are actually wrong, and keeping it consistent as staff and suppliers change. We also write standards you can reuse in those tools where that makes sense.
How long before anything changes on my store?
Approved changes publish throughout the programme, not at the end. The first ones usually go live in the opening weeks, while the deeper questions are still being worked through.
What happens if it does not help?
The scan is free and the assessment is credited against a programme, so you can get all the way to a scoped, priced plan before committing to the work. We would rather tell you your catalogue is in good shape than sell you a programme you do not need.

For ecommerce agencies

Sell catalogue work you do not have to staff.

The reason agencies avoid product listings is not the writing — it is chasing product information out of the client. We give you a workspace under your own brand for assigning those questions, recording the answers, managing approvals and keeping the evidence of why work moved or stalled. You keep the relationship and the margin; we do the work behind it.

Talk about partnering

Find out where your catalogue actually sits.

Give us your store address and we will scan a sample of your own products against the standards their categories should meet, then send you what we found.

Free. Read-only. No obligation, and the findings are yours whether or not you ever speak to us again.

Would rather talk first? Book twenty minutes and bring your store up on screen.

Scan my catalogue

Three details. The store address is the one that matters.

We use your details to run and return the scan. No list, no sequence you did not ask for.