Free feed analysis
What grade does your feed get?
Paste the feed URL your store already publishes. In about two minutes you get a letter grade for every product and a ranked list of what to fix first.
No login · no card · nothing written back to your feed
Avg score
82/100
Products graded
1,284
What happens
Two minutes, start to grade.
Nothing about the scan needs a decision from you. You paste a URL and read the result. The timings below are what a mid sized catalog takes.
If the fetch fails, you find out immediately and you find out why, rather than watching a spinner. The most common cause is a firewall on your side blocking our fetcher, and the message says so.
-
0s
You paste the feed URL
Step 1 of 4The same URL your ad channels already fetch. Nothing to upload.
URLpush ingestno file upload
-
8s
The feed is fetched and parsed
Step 2 of 4Whatever format your platform emits, read once and normalised into one shape.
XMLCSVJSONhttps onlyprivate-IP guard
-
45s
Every product is scored
Step 3 of 4Two engines read each item and their findings are composed into one letter grade.
30+ checks9 dimensionsA to F
-
~2m
The report is ready
Step 4 of 4The distribution across the catalog, then the fixes ordered by the points each recovers.
grade per productranked fixesnothing written back
The report
What lands on the screen.
Not a score out of ten with no explanation. The report is built so that the first thing you read is also the first thing worth doing.
How the catalog splits, A to F
The shape of the problem before any of the detail. A catalog with two hundred F products and one with two thousand C products need completely different work.
Which repair pays most
Findings grouped, each group showing how many products it affects and how many points it lifts. The order of work comes out of the data rather than out of a meeting.
Whether a model can understand it
The AI-Readiness figure, kept separate from the feed score, because a feed can pass every channel rule and still say nothing an assistant can act on.
The free scan reads up to two thousand products and opens two of the findings in full. The rest are counted and named, and unlock on a plan.
Limits, stated up front
What the scan will not do.
Worth saying plainly, because most feed tools ask for write access before they have shown you anything.
It does not touch your feed
The scan reads. It has no write path back to your store, by design rather than by policy, so there is nothing to undo afterwards.
It does not need an account
The grade comes first. You decide afterwards whether it was worth your two minutes.
It does not connect to your ad accounts
Nothing is requested from Google or Meta to produce the grade. The feed on its own is enough.
It does not invent a single value
A missing field is reported missing. Nothing is filled in to make the number look better than the catalog is.
After the scan
The grade is the diagnosis, not the treatment.
Reading the report costs nothing and changes nothing. Acting on it is where the product starts, and you can do that yourself or hand it to us.
Open the fix groups and work down the list
Thirteen groups, AI suggestions grounded in your own product data, IF then rules for the patterns, and export to the channels that need it. A fourteen day Pro trial covers the whole flow, and no charge is taken to start it.
What the product does →Let the consultancy run it
We came at this from the agency side, so the feed is rarely the only thing that needs attention. If the campaigns around it need work too, that is the same conversation rather than a separate one.
How we work →Questions
Before you paste the URL.
Where do I find my feed URL?
Wherever you set up Google Shopping. It is the same URL Merchant Center fetches. If you have never set that up, the app carries short guides for sixteen store platforms, from Shopify and WooCommerce to ikas, Ticimax and Ideasoft, each pointing at where that platform keeps the address.
Do you support Shopify and WooCommerce?
Verintra works with any feed URL, whatever produced it, so both are fine. To be precise about what that means: there is no app to install on either platform. We read the feed they publish, which is why the platform does not really matter.
Can I upload an XML or CSV file instead?
Not at the moment. Ingestion is by URL, or by pushing data to us. A file you upload once goes stale the same day, and a stale feed is the problem rather than the fix.
What if my server blocks you?
Some hosts block requests from datacenter addresses or from any user agent they do not recognise. Our fetcher identifies itself as VerintraFeedFetcher, so allowing that is usually enough. The error message tells you which case you hit rather than failing silently.
How is this different from the checks Merchant Center already runs?
Merchant Center tells you what it will refuse. That is a floor, not a standard. A product can be accepted, run, and still lose to a competitor whose title states the size and whose description says who the thing is for. The second score on this page measures exactly that gap.
What happens to my data?
The feed is fetched over https and the product data is used to produce your report. It is not sold, and it is not used to train a model. The scan needs no credentials from you, because a public feed URL is all it reads.
Is two thousand products a hard limit?
On the free scan, yes, and it is stated here rather than discovered halfway through. Larger catalogs are read in full on a plan. The two thousand are enough to see the shape of the problem, because feed problems are patterns and patterns show up early.
Two minutes, and you will know.
Paste the feed URL your store already publishes. Nothing is installed, nothing is connected, and nothing is written back.