ebisumart: 5 checks when search can't find you

ebisumart
5 checks when search can't find you
| Category | Read Time | Topic | Description | Source | Copyright |
|---|---|---|---|---|---|
| Tips | 9 min read | Japan EC platform guides for search and AI search | ebisumart updates its platform automatically and runs BtoC and BtoB on one base. Both can hide your shop from search and AI search. Five checks: the before-record, five-area diffs, index boundaries, BtoB... | searchos.io/en/blog | © 2026 Search OS |
Shops on ebisumart report three kinds of trouble: rich results disappeared, pages stopped showing up in search or AI search, or the opposite, wholesale price pages turned up on Google. The causes fall into two groups. The first is the automatic version-up that the platform runs on its own. The generation rules for structured data, canonical, and meta tags don't break; they change quietly, so looking at the screen tells you nothing. The second is the index boundary accident that comes from running a BtoC shop and a BtoB wholesale site on the same base.
This article pulls five checks out of our two ebisumart articles and puts them in the order a shop manager should run them. The first two cover diffs after a version-up. The last three cover the boundary when BtoC and BtoB share a site. The detailed procedures live in the articles linked at the end.
1. Do you have a record from before the version-up
You can't tell whether something changed unless you know what it looked like before. Automatic version-ups are part of how ebisumart works, and you normally can't opt out. Instead of trying to stop them, you build an operation that detects change. That means the before-record comes before any checklist. The first time, record the following before the version-up runs.
The title, description, and canonical strings of your three main pages (top page, main category, main product)
The structured data types the Rich Results Test detects on those same three pages, and whether it shows warnings
The URL count in your sitemap
The LCP, CLS, and INP values from PageSpeed Insights
The contents of robots.txt
Keep them in a spreadsheet with a date. From the next version-up on, you only compare against this record. The same record settles whether a ranking drop came from the version-up or from a Google update. If the five areas show a diff right after the version-up, the cause is on the platform side. If they don't, the search engine side is more likely.
2. After the version-up, do any of the five areas show a diff
When the version-up notice arrives, check five areas on the next business day and compare them with the previous record. The whole pass takes about 30 minutes.
Area | What to check | Tool | Time | If it changed, suspect |
|---|---|---|---|---|
Metadata | Output of title, description, canonical | Source of the three main pages | 5 min | A change to the platform's title template |
Structured data | Detected types and errors | Rich Results Test (1 product, 1 category, top page) | 10 min | A change to the platform's default output, or a conflict with your template |
URL | Main page URLs and the canonical on parameter URLs | Browser | 5 min | A change in how parameters are handled |
Sitemap | URL count, lastmod, errors | Sitemap URL and Search Console | 5 min | A change of scope, or generation stopped |
Speed | LCP, CLS, INP | PageSpeed Insights (main product page) | 5 min | A script added on the platform side |
If robots.txt changed, suspect a change to the platform default. When you find a diff, read the ebisumart release notes for that version and decide whether it's an intended spec change or a side effect. For a spec change, adjust your template and settings to the new spec. For a side effect, contact support. Don't write "rankings dropped." Write the diff: "After version X.Y, the offer disappeared from Product structured data on product pages," and attach the before-and-after record. A customized template doesn't exempt you. The platform's default output for the parts you didn't customize still changes, so the check stays.
3. Have you split pages into four types and set public and index separately
ebisumart lets you run a BtoC shop and a BtoB wholesale site on the same base. That is exactly why wholesale price pages end up on Google. Draw the line by "do we want search engines to see this," not by "is it behind login." The two don't always match. Pages behind the login wall still get indexed through links or leftover settings from earlier.
Page type | Public | Indexed | Setting |
|---|---|---|---|
BtoC products, categories, content | Yes | Yes | Normal |
Public BtoB information (product overview, trading terms, contact) | Yes | Yes | Normal. Let it be indexed as the BtoB entrance |
BtoB prices, member-only products, quotes | No | No | Login required + noindex (both) |
Cart, my page, search results | Yes/No | No | noindex |
Member-only pages need both login-required and noindex. Login-required stops people from seeing the content, but it doesn't stop search engines from recognizing the URL, and sometimes only the title gets indexed. noindex stops indexing, but it doesn't stop someone who knows the URL from opening it. A robots.txt Disallow only blocks crawling; it doesn't remove what's already indexed, and with an external link the URL alone gets indexed. Check that both settings are present every time you add a page.
4. Have you closed the routes through which BtoB prices leak
Even with the boundary decided, prices leak through several routes. When the person in charge of BtoB products and the person in charge of SEO are different, a missing noindex is an everyday event.
Pattern | Cause | Fix |
|---|---|---|
Wholesale price shown as "member price" while logged out | Display setting for member-rank pricing | Hide the price or show only the standard price when logged out |
Wholesale site category pages are public | Member-only product categories aren't noindex | noindex the categories too |
Sitemap includes BtoB product URLs | Sitemap scope | Exclude member-only products |
BtoB product structured data contains the price | Structured data output conditions | Don't output price for member-only products |
Index remains from a period when pages were public | Removal missed after the setting change | Search Console removal tool + noindex |
If an indexed wholesale price page has to disappear fast, use the Search Console removal tool to hide it temporarily and set noindex and login-required at the same time. The removal tool alone expires, and the page shows up again.
5. Is the public page that serves as the BtoB entrance outside the login wall
Hide all of BtoB and new trading partners can't find you in search. They ask AI things like "which company wholesales X" or "where do I source X for commercial use." To be the answer, you need public pages outside the login wall. Hide the price, show the existence. That is the base of the index design.
Product overview (items, origin, specifications, no prices)
Trading terms (minimum lot, lead time, outline of payment terms)
Industries and regions served
The contact and account-opening flow
Industry-specific "how to choose" and case pages
Let these be indexed and include the BtoB business in your Organization structured data. Prices and stock stay behind the login wall. The boundary isn't finished when you decide it. Every new product, every new member rank, every template change can break it. Once a month, search Google for site:yourdomain plus "wholesale price" to see whether BtoB prices are indexed. At the same frequency, check that the sitemap has no member-only URLs and that the public BtoB pages show as indexed in the Search Console URL inspection. Whenever a setting changes, confirm noindex in the page source and open a member-only product in an incognito window to see what a logged-out visitor sees.
Search OS keeps watching whether the settings you decided still hold today
A checklist covers what a person can go and look at. If the change happened on the thousands of pages outside your three main ones, checking three pages won't find it. Whether AI crawlers can still fetch your pages after a version-up doesn't appear in Search Console or PageSpeed Insights. For sites on automatically updated platforms such as ebisumart, Search OS continuously verifies the state of metadata, structured data, canonical, and sitemap across every page and detects what changed before and after a version-up as a diff. For sites where BtoC and BtoB share a base, it continuously checks the list of indexed URLs, the noindex and canonical settings, the prices inside structured data, and the pages bots actually fetch, and turns a broken boundary into a fix list. It is not a layer that stops version-ups, and it doesn't decide the boundary for you.
Questions the person in charge should ask first
Is it decided who receives the version-up notice and who checks the five areas on the next business day
Do you keep a dated record of the title, description, and canonical of your three main pages and of the Rich Results Test results
Do you record both when rankings moved and when the version-up ran
Do member-only pages have both login-required and noindex
Does the sitemap contain any BtoB product URLs
When you open a member-only product in an incognito window, does the wholesale price stay hidden
Are the product overview and trading terms pages outside the login wall
Conclusion
When an ebisumart shop doesn't show up in search or AI search, the cause is usually one of two things: output the platform changed quietly, or a boundary between BtoC and BtoB that has come apart. Neither is visible on screen. You find them by comparing against the previous record and by confirming that both noindex and login-required are in place. Compare the five areas after every version-up, and walk the boundary once a month. Putting those two habits into operation is where to start.
Related reading
A checklist for confirming ebisumart's automatic version-up didn't change your SEO settings
How to design the index boundary on an ebisumart BtoC and BtoB site: member-only vs. public pages
References
Google Search Central: Page Indexing report
Google: Rich Results Test
Google: PageSpeed Insights
Google Search Central: Block Search indexing with noindex
Google Search Central: Remove a page hosted on your site from Google
Google Search Central: Organization structured data