Go to content
AI assistant recommending three winter boots based on product data
  1. AI
  2. Agentic Commerce
  3. AI Shopping Assessment

AI Shopping Assessment

Your customers are already asking AI assistants what to buy. We show you whether your product data is good enough to be chosen, and what needs fixing first.
Book an assessment

The customer gets three suggestions. Not ten links.

In a traditional Google search, customers click around between webshops themselves. When they ask an AI assistant instead, they get a few specific product suggestions, and the assistant chooses them based on product data. Not on images, layout or brand. If your product is missing the data that answers the question, it won't make the cut. And a bigger ad budget can't fix that.

AI shopping is already underway

  • people connections icon

    Around half

    of consumers in Europe's five largest markets have used generative AI to learn about a product. (McKinsey, March 2026)

  • persentage icon

    42 % better

    higher conversion for traffic from AI sources than for other traffic on US retail sites in March 2026. (Adobe, April 2026)

  • a file with a speed watch icon

    31. January 2027

    is the date Google stops showing product images smaller than 500 x 500 pixels. (Google Merchant Center)

Can your products answer your customers' questions?

A customer writes: "Which children's winter boots under DKK 600 are waterproof and can be delivered by Friday?" The assistant can only recommend products where waterproofing, size, price, stock and delivery are available as data. If one field is missing, the product isn't a poor answer. It isn't an answer at all. The worst part is that you can't see it. Google Merchant Center warns you about missing required fields, but not about the questions your data can't answer.

What is an AI Shopping Assessment?

An AI Shopping Assessment is a fixed-price analysis of your product feed from Google Merchant Center. We measure how well your product data meets Google's requirements and how well it can answer the questions customers ask AI assistants. You get a score, a list of specific gaps and a prioritised plan for what to fix and in which order.

It's not an implementation project, and we don't need access to your systems. It's the first step towards being ready for AI shopping, so you can decide on your next move from an informed position.

No more guesswork. You'll know what needs fixing, how big a job it is and what to tackle first.

We check your data on three levels

  • a paper and setting circle icon

    Foundation

    Required fields, identifiers such as GTIN, titles and images. If a required field is missing, the product won't appear at all.

  • icon of different lines and symbols

    Format and rules

    Price format, fixed values, industry-specific rules and whether the feed matches the webshop. Data must be consistent before an AI assistant can compare you with others.

  • Brain icon

    AI semantics

    Can the feed answer questions about material, use, target audience and care? That determines whether your product matches the customer's specific question.

Each product gets a score from 0 to 100. Below 50 is red, 50 to 84 is amber, and 85 or above is green.

How it works

From link to plan in under two weeks

  • files icon

    1. You share your feed

    All we need is a link to the product feed you send to Google Merchant Center. If you use a feed tool, you can usually get the link from there.

  • Website icon

    2. We analyse your entire range

    Every product is checked against 88 data fields and a set of rules for format and consistency.

  • chart icon

    3. You get the report

    An overall score, all findings and a prioritised plan, plus an Excel list of the affected products.

  • envelope icon

    4. We go through it together

    In a meeting, we explain the findings and priorities, and we agree on next steps. From your first enquiry to the finished report typically takes one week, two at most.

What you get

  • chart icon

    One overall score

    Your AI shopping readiness in a single number, so you know where you stand and can measure progress afterwards.

  • a logic map and a pen icon

    Score for each area

    See whether the problem lies in core data, format or content.

  • a magnifying glass with a human icon

    Errors that block products

    An overview of what is currently keeping products out of Google Shopping.

  • a message search with an arrow icon

    Questions your feed can't answer

    See where AI assistants choose a competitor instead of you.

  • a target with an arrow icon

    Top 5 fixes and a plan

    Start with what gives the most for the least, and get a solid basis for deciding on next steps and budget.

  • pen inside the setting image icon

    Excel list and better data

    Affected products ready to fix, plus AI-generated suggestions for titles and attributes.

Example

This is what the report looks like

The report runs to around 20 pages: an executive summary for management and a technical section for those making the fixes. We also check your checkout and UCP readiness, so you know what's missing to be able to sell directly in AI assistants.

Fixed price. No surprises.

An AI Shopping Assessment costs DKK 35,000 excl. VAT. The price is fixed and covers your entire range in the feed you share.

Included: Analysis of every product in the feed, a report with score and findings, an Excel list, top 5 fixes, a phased plan and a review meeting.


Not included: The fixes themselves, integrations and implementation. We offer these as a next step based on the report's findings.

Book an assessment

First step of three

The assessment shows where the gaps are, so you can invest in the order that pays off. The plan is tailored to your systems and markets.

  • Binoculars icon

    1. Find the gaps

    We compare your product data with what AI assistants and Google require, and identify the quick wins.

  • a computer, tablet and phone icon

    2. Connect your data

    We agree which system owns each piece of data, such as PIM, ERP or commerce platform. Price, stock and delivery must come live from the source systems, because an AI agent checks them right before the purchase.

  • a raising chart icon

    3. Build and develop

    We connect the systems, enrich product data using AI with human review, and get your checkout ready for AI purchases.

The heavy lifting happens behind the feed

Many agencies optimise feeds and campaigns. But the new requirements shift the work from the feed itself to the systems behind it: PIM, ERP, commerce platform and checkout. That's where our core expertise lies.

We have performance marketing and development under one roof, so the analysis can be turned straight into action.

We make the diagnosis. And we can deliver the fix too.

Get answers to the key questions about AI Shopping Assessment

Has AI checkout launched in Denmark?
No, not yet. Checkout built on Google's new standard, the Universal Commerce Protocol (UCP), is currently available with selected retailers in the US. Google has announced Canada, Australia and later the UK as next. There is no date for Denmark. But AI assistants are already being used to find and compare products, and they read the same product data.
Why should we act now?
Because the data requirements apply now. From 31 January 2027, Google is raising the minimum requirement for product images, and the six new data fields for AI shopping can already be used today. A large range can't be fixed in a month, and those who are ready first get recommended first.
What do we need to provide?
Just a link to the product feed you send to Google Merchant Center. We don't need system access, and there's no project to set up.
Can't you just use our raw feed?
No. The feed in Merchant Center has often been processed in a feed tool or supplemented with extra data. That's the version Google and AI assistants read, so that's the one we analyse.
How do you calculate the score?
Each product starts at 100 points and loses points for each finding: 30 for critical errors such as a missing required field, 15 for important errors such as a title that's too short, and 5 for minor errors such as a missing brand. The overall score is the average across all products.
We already have a feed tool or an agency. Is it still relevant?
Yes. We measure the result, regardless of who produces the feed, and the report can be used directly by your current agency or team.
Is it only relevant for Google?
The analysis is based on Google Merchant Center, because Google sets the most detailed requirements. The same product data is also used by other AI assistants, so better data helps across the board.
Is it relevant for B2B?
Yes, if you have a product feed in Google Merchant Center. The logic is the same: whoever has the best product data gets chosen.
What happens afterwards?
You get a prioritised plan. Your own team can make many of the fixes using the Excel list. If you'd like help with data enrichment, integrations or checkout, we'll put together a quote based on the report's findings.

anchorLink

Book your AI Shopping Assessment

Fill in the form and we’ll contact you within one working day to agree how you’ll share your feed. You’ll typically have your report a week later.

If you have any questions, email Tobias Westergaard, Head of Performance, at [email protected].

You can also call us on +45 86 19 05 50
or email [email protected]