Site readiness
Site readiness checks whether AI search engines and shopping assistants can read your store. NameIO looks at your robots.txt file (the file that tells crawlers which pages they may read), your sitemap (the list of your pages for search engines), your product pages, the product data in the page's code (called JSON-LD), your policy pages and a few small files for AI agents. It runs the same checks on the competitors you track, with their public pages only. A crawler is a program that reads web pages; NameIO's crawler is described under Our crawler.
Open it from Site readiness in the sidebar, under AI Visibility, or from the Site checks failing tile on the dashboard's AI visibility card.
What the page shows
The top of the page says when NameIO last checked your store and how many checks pass: "Checked Nov 2 · 28 of 34 checks pass". The count is the checks that pass, out of the checks that pass or fail. Checks that don't apply, checks NameIO couldn't do and checks for information only are not in the count. NameIO shows counts only. It never turns the checks into a score or a percentage.
Below is a table with one row for each check. The rows are in groups: Access, Product pages, Structured data, Product data, Trust, Speed and Agent files, then For information (not counted). Point at a check's name to see what it checks. The first column after the checks is your store, then one column for each competitor. Each column's heading gives its own count, such as "28 of 34 pass". A competitor's heading also says "Public data only".
Each cell has one of four results:
- ✓ Passes. The check passes on that site.
- ✗ Fails. The check fails on that site. A check that looks at several pages shows how many pass: "✗ 7 of 10" means 7 of the 10 pages NameIO read pass and 3 fail.
- — Doesn't apply. The check doesn't fit the site. For example, a product page that shows no reviews has no rating to check. In a competitor's column, "—" also marks a check that NameIO runs on your store only. For a competitor, the five feed field checks come from its public product list.
- Not checked. NameIO couldn't do the check, for example because the site didn't answer or its robots.txt closes the page to NameIO. Point at the cell to see the reason. Not checked is never counted as a fail.
The Product data rows check the product data your store gives out. Where they get it from, and how to add your product feed, is under Your product data. Public product list needs nothing from you: NameIO reads it from your store's public pages.
The checks under For information (not counted) show what NameIO found, in words and with no ✓ or ✗, such as "Blocked: 2 of 5" for AI training crawlers or "Found" for an llms.txt file. They are your choice, so they never count as passing or failing, and the panel gives no "What to do" for them. One exception is Bot protection answers for pages, described next.
Bot protection. Some stores use a bot-protection service, such as Imperva Incapsula, Cloudflare, Akamai, DataDome, PerimeterX or Kasada. When the service doesn't know a visitor, it can send a small challenge page instead of the real page. A challenge page asks the browser to run a script or to show that a person is there before it sends the page. NameIO's crawler doesn't run scripts, so it gets the challenge page. NameIO doesn't guess what your real page says: it reports the challenge page instead. The checks of that page show Not checked, with a reason such as "Imperva Incapsula answered instead of the page", and a line under the top of the page says how many of your pages the service answered. AI crawlers that your bot protection doesn't allow get the same challenge page, so they can't read those pages either. To fix it, allow NameIO-Audit and the eight AI search crawlers (OAI-SearchBot, Claude-SearchBot, PerplexityBot, Googlebot, Bingbot, Applebot, Amzn-SearchBot and Meta-WebIndexer) by their user agent in your bot-protection service's settings. The panel of Bot protection answers for pages gives the steps for your service. If NameIO can't tell which service it is, ask your IT team or your bot-protection vendor.
Turn on Only failing to see only the checks that fail on your store. The list next to it shows one group of checks, or all of them. On a phone, the table shows your store and one competitor. Pick the competitor from the list above the table.
Select a cell to open what NameIO found. The panel gives the check and the site, the result in words (such as "Fails: 3 of 10 product pages"), what the check means, and what NameIO found: links to the pages it read, the values it saw and the robots.txt line numbers. Links open in a new tab. For a check that fails on your store, it also says what to do, with a button to get the fix NameIO builds for it (see Fix what fails). For a product data row, it says where the items came from, how many items miss each field or have it wrong, and the ids of the items that fail, 20 at a time, with how many more there are.
Passing checks over time compares your store with your competitors: one line for your store and one for each of up to four competitors, with a point for each day NameIO checked that site, over the last 6, 12 or 24 months. Choose No comparison to see one site, and pick it from the list. The change compares the site's last check with its last check in the months before, when both covered the same checks. Select Show as table to see the same numbers as a table, with one row for each day and one column for each site. When NameIO changes what it checks, the chart marks that day with a dashed "Checks changed" line, because counts before and after cover different checks. The chart shows after NameIO has checked your site on two different days. More on the chart and its table is under Charts and tables.
The dashboard's AI visibility card has a Site checks failing tile for this page, such as "6 of 34": the checks that fail on your store's latest check, out of the checks that pass or fail. Under it, a small line shows the last 7 days. It links here. Before the first check it says "Not checked yet".
Your product data
AI shopping answers use the product data your store gives out: names, prices, stock, brands and barcode numbers. NameIO reads it in one of three ways:
- Your public product list. Some platforms (Shopify, WooCommerce and Magento) publish a list of products that anyone can read. NameIO reads one page of it in each check of your store, so you don't need to do anything. A public list doesn't show every field a feed has, such as the seller name, so NameIO doesn't count the fields it can't show.
- Your product feed, on any platform. A feed is the file of your products that you send to Google or ChatGPT shopping. In the Product data box at the top of the Site readiness page, paste the feed's address and press Save, or press Upload a file to send a CSV, TSV or XML file. NameIO reads an address in each check of your store, and starts a check when you save it, if you can check again. It checks an uploaded file at once. If NameIO finds no products it can read in the file, nothing changes, and a saved address stays. Press Remove to stop using a feed. When you use a feed, it replaces the public product list for these rows.
- Your Shopify store, when you connected it. If you connected Shopify in Settings, Integrations (to publish fixes), each check of your store also reads your active products from Shopify: each product's title, brand, product type, description (NameIO reads at most its first 5,000 characters, as plain text, and only counts their length) and address; the price, stock, SKU and barcode of its variants; its images' addresses, sizes and whether they have alt text; and the Google Shopping fields that Shopify's Google & YouTube app adds (google_product_category, gender, age_group, condition and custom_label_0). NameIO reads Shopify only because you connected it, and asks for no new access. It never stores a description and never reads your customers or orders. Image size comes only from Shopify, and Feed matches the pages compares Shopify's prices and stock with your product pages. When NameIO could read your feed, the other feed rows keep using it; otherwise Shopify replaces the public product list for them.
The box says which one NameIO uses, for example "Feed: https://yourstore.com/products.xml · read Nov 3 · 1,204 items", "Feed: uploaded CSV file · 310 items · read Nov 3" or "No product feed yet". When the latest check read your Shopify store, it also says "Shopify connected: 200 products read Nov 3". Point at a product data cell to see where its result came from: "From your product feed", "From your public product list" or "From your Shopify store". For competitors, these rows always come from their public product list. NameIO never reads a competitor's feed.
NameIO never stores your feed or the text of its items; what it keeps is under What NameIO reads and keeps.
With no feed and no public list, the feed rows read "Not checked: add your product feed". When NameIO couldn't read your feed, they read "Not checked" with the reason, such as "The feed didn't answer". When NameIO couldn't read your Shopify store, the rows Shopify answers read "Not checked" with the reason, such as "Shopify was busy, so NameIO read no products this time". When Shopify gets busy after the first 25 products, NameIO keeps the products it already read. A feed or a Shopify store NameIO couldn't read never counts as a fail, and the rest of the check still runs.
These are the checks for each feed field:
- Google feed fields: each item has an id, title, description, link, image link, availability, price, and a brand, gtin or mpn. An item that is used or refurbished says so in its condition.
- ChatGPT feed fields: each item has the nine fields ChatGPT shopping requires, from OpenAI's product feed spec, read on 11 Nov 2026: item_id, title, description, url, brand, seller_name, image_url, availability and price. Google's names count too: id for item_id, link for url and image_link for image_url. availability is in_stock, out_of_stock, pre_order, backorder or unknown (Google's "in stock", "out of stock" and "preorder" count too), and price is an amount and its currency, such as "79.99 USD". A price without a currency is wrong. seller_name has no other name, and NameIO never takes it from your brand. Stock counts and the search and checkout flags (is_eligible_search, is_eligible_checkout) are optional, so they never count against an item.
- Valid barcode numbers: each gtin has a correct last digit.
- Clothing details: each item in clothing (by its google_product_category) has a color, size, gender and age_group.
- Variants grouped: items that are versions of one product share one item_group_id.
- Feed matches the pages: each item has the same price and stock as the product page NameIO read at the same address in the same check. It needs the feed's address, because an uploaded file isn't compared. An item whose page NameIO didn't read isn't compared, so the count is of the products NameIO read. With Shopify connected, Shopify's prices and stock are compared instead: each product with the page NameIO read at its address, or at /products/ and its handle, using its first variant's price and stock.
- Extra answer fields (for information): how many items have fields that give AI answers more to say, such as questions and answers or reviews.
- Image size: each product's main image (its first) is at least 500 by 500 pixels, by the size Shopify gives. NameIO doesn't download images, so this check needs your Shopify connection. A product with no image, or with an image Shopify gives no size for, isn't counted; when no product has an image with a size, the row shows "—".
The correction file. When NameIO has read your feed and some items fail, the box shows Get the correction file. It downloads a CSV file with one row for each failing item and four columns: item_id, missing_fields, invalid_fields and note. It has item ids and field names only, never your product data. When NameIO can't tell which of a few fields one item lacks, the note names them, such as "Check also: gtin". Fix those items in your platform or your feed app; NameIO reads the feed again in the next check.
Fix what fails
For a check that fails on your store, the panel's What to do has a button. Get the file gives you something to add to your site. See the steps gives you a note and the steps for a change in your platform or your network. The same button is on the check's row in What to do next on the dashboard. NameIO builds the fix from what its latest check of your store found and from your project's own details. No AI model writes it, and it never has a rating, a review or a number NameIO made up. NameIO never writes to your store, your platform or your network: you make the change.
The button opens the fix's own page. It shows the check, the fix's status and a line on the check's result on your store, such as "Failing since Oct 30; check again after you apply it" and later "Failing on Oct 30, passing since Nov 3". Then comes what NameIO built, with Copy and Download, and the steps for your store's platform. When you have made the change, press Mark as done, or press Dismiss if you won't make it. Check again checks your store now, as on this page. A fix is Resolved once a later check of your store passes the check. NameIO keeps one open fix for each check and page. Pressing the button again opens it, built again from NameIO's latest check of your store and the facts you added since, such as the facts of a page fix or the store facts you ticked. Once you mark a fix as done, the button opens it as it is. Fixes are free on every plan.
What NameIO builds for each kind of check:
- robots.txt lines (AI search crawlers allowed, AI assistants can open pages, Sitemap listed in robots.txt): the lines to add, naming each crawler the check found blocked, and the numbers of the lines that block them. NameIO tests the lines with the rules crawlers use when it has your file. It keeps only the numbers of the lines that block, not your whole robots.txt, so today the fix page says it couldn't test them against the rest of your file. The file is robots-lines.txt.
- Structured data (product data on your product pages, shipping and returns in it, and your store's own data and profiles): the code to add. For a product page it uses the facts NameIO read on that page for a page fix (name, image, price, stock and product code); for a page it has no facts for yet, it says to start a page fix for that page first. Shipping and returns come only from the store facts you ticked on a page fix. Your store's data uses your home page's share image as the logo and the profile links found on your home page. What NameIO doesn't know is a blank to fill in. The file is structured-data.html.
- A correction file (the product feed checks): one row for each failing item with its id and the fields to fix, and the steps in your feed app. The file is feed-corrections.csv.
- Page templates (Policy pages): a plain text template for each missing page (shipping, returns, privacy and contact) with blanks such as "[your return window in days]". NameIO fills in no numbers, and the templates are not legal advice. The words are yours to write, or your AI assistant's with a page fix for a new page. The file is policy-pages.txt.
- Notes and settings (pages kept out of search or not quoted, the main address, text that isn't in the page, sitemaps, the secure address, AI crawlers blocked by your network, home page speed, mistakes in product data, and product pages that need more facts or detail): a note naming the pages NameIO found and what to change, with the steps for your platform or network. Where your host decides, the steps say to ask your host or developer. The text of a product page is written with a page fix.
Checks for information and your competitors' cells never get a fix. Public product list, Shopping agent file, IndexNow key and Image size have no fix yet.
Under the table, Site fixes lists your fixes that are open or done, each with its status and dates. It opens each fix's page.
ChatGPT for business
ChatGPT lets a business claim its profile in ChatGPT's Business Hub, add its product feed there and put a Business Agent live. NameIO can't see inside Business Hub, so the ChatGPT for business box under the table lists three steps that you mark yourself:
- Business claimed and verified: you claimed your business in Business Hub and finished the steps that show it is yours.
- Product feed connected: your product feed is added in Business Hub, sent by SFTP or through your store platform. Check the feed here first: the ChatGPT feed fields check looks for the nine fields ChatGPT needs in each item. If NameIO doesn't read your feed yet, export it as a CSV or TSV file and upload it in the Product data box. Shopify stores are already connected to ChatGPT, so on Shopify this step says "Already connected through Shopify".
- Business Agent live: a Business Agent for your store is set up and turned on in Business Hub.
ChatGPT offers the feed and agent steps to stores in the US first. Each step has a short list of what to do and a link to Business Hub, which opens in a new tab. When you have done a step, press Mark as done: the box then says when it was marked, such as "You marked this done on 12 Nov". Press Undo to take it back. These steps are not site checks, so they are never in the count at the top of the page. Until the first step is marked, What to do next on the dashboard shows Claim your business in ChatGPT, and its See the steps opens this box. The box is on NameIO at app.getnameio.com.
How often NameIO checks
- Your store: NameIO checks it by itself at most once a day: after an AI check finishes, or when your store has not been checked for a day.
- Your competitors: once a week. NameIO checks up to 5 competitors on the free plan and up to 20 on Pro. When you track more competitors than that, NameIO starts with the ones named in the most answers of your latest AI check. On Pro, the others take turns from week to week.
- Check again: press Check again at the top of the page to check now. You can check your store again once an hour and your competitors once a day. When you can't yet, the button can't be pressed and the page says when you can, such as "You can check again at 3:30 PM." While a check runs, the page says when it started and shows the results when they arrive, usually within a few minutes.
Checking your site costs no credits and asks no AI model anything.
What NameIO reads and keeps
For your store, one check reads robots.txt, up to 4 sitemaps, the home page and its plain http:// address, up to 10 product pages, one collection page when the sitemaps list too few products, up to 4 policy pages, and the small files llms.txt, agents.md, the shopping agent file and the IndexNow key file. NameIO looks for product pages first in your product sitemap, then in the product links of the feed NameIO reads from your feed's address or of your public product list (an uploaded file is not kept, so its links are not used), then among your sitemap's other addresses, and last in your home page's links. It looks for policy pages in your home page's links, then in your sitemap; on Shopify, it uses Shopify's own policy addresses in place of the sitemap. Where your store's platform publishes a public product list, NameIO reads one page of it. When you give NameIO your feed's address, the check also reads the feed once, wherever it is hosted; when the feed is on your store's website, robots.txt applies to it as to any page. A check of your store reads at most 25 MB in all, or 45 MB when it reads your feed too. When your Shopify store is connected, the check also reads your products from Shopify's own system, which doesn't count toward these limits. On your store only, NameIO also opens the home page once with the name of each AI search crawler, to see whether your network or firewall blocks them. For a competitor, NameIO reads only what any visitor can: robots.txt, up to 3 sitemaps, the home page and its http:// address, up to 3 product pages, up to 3 policy pages, llms.txt, agents.md, the shopping agent file and one page of its public product list.
NameIO keeps the results: for each check, pass, fail, doesn't apply or not checked, with counts such as "3 of 4". With the 3 latest checks of each site it also keeps a few details: up to 5 links, short values and robots.txt line numbers. NameIO never stores the text of a page or of your feed: for a feed it keeps its address, its format, how many items it has and when it was read, and with the results the ids of the items that fail each check. It doesn't keep an uploaded file or its name. From Shopify it keeps only counts, when it read them and the Shopify ids of the products and variants that fail. Results are kept for 24 months.
The checks
These are all the checks. Also on competitors says whether NameIO runs the check on your competitors too. A check for information only is shown in its own group and never counts as passing or failing, because it is your choice or no AI engine says it uses it.
| Check | Group | What it checks | Also on competitors | For information only |
|---|---|---|---|---|
| Home page opens | Access | NameIO can open the home page and read it, and no bot-protection service answers instead. | Yes | No |
| AI search crawlers allowed | Access | The robots.txt file lets the crawlers of AI search and answer engines read the home page and product pages. | Yes | No |
| AI assistants can open pages | Access | The robots.txt file does not block the agents that open a page when a person asks an AI assistant about it. | Yes | No |
| AI training crawlers | Access | Shows whether robots.txt lets AI training crawlers in, which is your choice and does not change AI answers. | Yes | Yes |
| Sitemap listed in robots.txt | Access | The robots.txt file has a Sitemap line that crawlers can follow. | Yes | No |
| AI crawlers not blocked by the network | Access | The store's network or firewall does not block or challenge AI search crawlers that robots.txt allows; a challenge page counts as blocked. | No | No |
| Bot protection answers for pages | Access | A bot-protection service answered instead of the page on some reads; those checks show Not checked. | Yes | Yes |
| Products in the sitemap | Access | The sitemap lists the store's product pages. | Yes | No |
| Product pages open | Product pages | Product pages answer with a normal page, not an error or a redirect. | Yes | No |
| Pages can be in search | Product pages | Product pages do not tell search engines to leave them out of results. | Yes | No |
| Pages can be quoted | Product pages | Product pages do not stop search engines from quoting their text. | Yes | No |
| Main address is set | Product pages | Each product page names its main address, and that address opens. | Yes | No |
| Text is in the page | Product pages | Product pages send their text in the page itself, so crawlers that run no scripts can read it; challenge pages are not judged. | Yes | No |
| Product facts in the page | Product pages | A product page shows the name, price, stock, a description of at least 50 words and product data before any script runs. | Yes | No |
| Product data on the page | Structured data | A product page has Product structured data that NameIO can read. | Yes | No |
| Required product data | Structured data | The product data has a name, an image and an offer with a price above zero and its currency. | Yes | No |
| Stock in product data | Structured data | Each offer in the product data says whether the product is in stock. | Yes | No |
| Brand and product code | Structured data | The product data has a brand and a valid barcode number, a part number or a product code. | Yes | No |
| Shipping in product data | Structured data | The product data or the store's organization data describes shipping. | Yes | No |
| Returns in product data | Structured data | The product data or the store's organization data describes the return policy. | Yes | No |
| Variants in product data | Structured data | When a page sells a product in more than one version, the product data lists each version. | Yes | No |
| Reviews in product data | Structured data | When a page shows reviews, the product data has the average rating and how many reviews there are. | Yes | No |
| Product data matches the page | Structured data | The price and stock in the product data are the same as the ones the page shows. | Yes | No |
| Structured data can be read | Structured data | Every block of structured data on the page is written correctly. | Yes | No |
| Feed matches the pages | Product data | The price and stock in your product feed are the same as on the product pages. | No | No |
| Policy pages | Trust | The store has shipping, returns, privacy and contact pages, linked from the home page or listed in the sitemap. | Yes | No |
| Store described | Trust | The home page has organization data with the store's name, web address and logo. | Yes | No |
| Store profiles linked | Trust | The organization data links to the store's profiles on other sites. | Yes | No |
| Detail on product pages | Product pages | A product page has a long description, headings, a table of details, visible questions and answers, and image descriptions. | Yes | No |
| Public product list | Product data | The store platform's public product list answers, so NameIO can see the products without a feed. | Yes | No |
| Google feed fields | Product data | Every item in your product feed has the fields Google requires for shopping results. | From their public product list | No |
| ChatGPT feed fields | Product data | Every item in your product feed has the nine fields ChatGPT shopping requires. | From their public product list | No |
| Valid barcode numbers | Product data | Every barcode number in your product feed has a correct check digit. | From their public product list | No |
| Clothing details | Product data | Clothing items in your product feed have color, size, gender and age group. | From their public product list | No |
| Variants grouped | Product data | Items that are versions of one product share a group id in your product feed. | From their public product list | No |
| Extra answer fields | Product data | Shows how many items in your product feed have the extra fields that help AI answers, such as questions and answers. | No | Yes |
| Image size | Product data | Every product's main image is at least 500 by 500 pixels. | No | No |
| Shopping agent file | Agent files | The store publishes a commerce profile that shopping agents can read. | Yes | No |
| llms.txt file | Agent files | Shows whether the store has an llms.txt file in the usual format; no AI engine says it reads one. | Yes | Yes |
| agents.md file | Agent files | Shows whether the store has an agents.md file for AI agents. | Yes | Yes |
| Home page speed | Speed | The home page starts quickly, is not too large and gives its images a size. | Yes | No |
| IndexNow key | Access | The store publishes an IndexNow key file, so it can tell search engines when pages change. | No | No |
| Secure address | Trust | The home page opens on a secure https address. | Yes | No |
Our crawler
NameIO reads sites with its own crawler. It names itself with the user agent NameIO-Audit/1.0 and a link to this section. The only requests without that name are the tests of your own store's home page with the names of AI search crawlers, described above. NameIO never makes them on a competitor's site. The crawler works like this:
- It reads robots.txt first and obeys it. It never reads a page that robots.txt closes to NameIO-Audit. When robots.txt closes the home page, it reads nothing more on that site.
- It sends one request at a time to each site, at least a second apart. When robots.txt sets a
Crawl-delay, it waits that long between requests, up to 10 seconds. - When a site answers "too many requests" (429), it stops reading that site for a while.
- It only reads. It never stores the text of a page, only the results described above.
To stop NameIO's site checks from reading your site, add this group to your robots.txt:
User-agent: NameIO-Audit
Disallow: /
NameIO then reads only your robots.txt, sees the rule and stops. The same name is used by NameIO's Site audit, which a NameIO user starts by hand and which follows your robots.txt rules for all crawlers (User-agent: *).
To ask NameIO to leave your site out completely, email chris@getnameio.com with your site's address. We add it to a list of sites NameIO's site checks never read.
What's not done
- NameIO reads a feed of up to 20 MB, and its first 10,000 items. An uploaded file can be up to 10 MB. The results list up to 200 failing items for each check.
- The store connection is Shopify only, so on other platforms Image size isn't checked. NameIO reads up to 200 active products from Shopify, with up to 10 variants and 5 images each, and doesn't read stock counts or color and size options.
- NameIO can't read ChatGPT's Business Hub, so it doesn't know whether a ChatGPT for business step is done until you mark it.
- NameIO doesn't read your Google Merchant Center or Microsoft Merchant Center account, or Google's speed data from real visitors (the Chrome UX Report).