New Ecommerce Site Disappeared From Google: What to Check First
A sudden drop after early impressions on a new ecommerce site is a reason to investigate crawlability, indexation, templates, and page value—not simply wait for a supposed ranking sandbox to end.
· 14 min read
A new wine and spirits store earned roughly 250–350 Google impressions per day, with some bottle-name queries ranking between positions 5 and 20, then lost most visibility around 18 days later. When a new ecommerce site disappeared from Google, the practical payoff is knowing whether the loss is normal index churn or evidence of an issue that needs a fix before thousands more pages are crawled.
The situation comes from a Reddit post about a WordPress and WooCommerce store in Vietnam that launched with 5,977 product pages, 141 category pages, 116 brand pages, and 504 blog posts. Google had indexed only about 20% of products and reportedly made about 225 crawl requests daily. Bing, Yahoo, and ChatGPT were still producing visits and sales.
That combination does not prove a penalty, a sandbox, or a single technical failure. It does mean the right response is a structured investigation. Waiting can be appropriate after a clean audit; waiting without checking the evidence is not.
Why early rankings can vanish on a new store
The first point is that there is no official Google concept called a “new-site honeymoon” or guaranteed trial period. Google explains that crawling, indexing, and ranking are separate processes, and it does not guarantee that an eligible page will be crawled, indexed, or shown for a query. Early impressions can therefore be a normal byproduct of Google discovering pages and testing relevance, rather than proof that a site had established rankings.
For a store with nearly 6,000 product URLs, several things can happen in the first few weeks:
- Google discovers URLs through the XML sitemap, category pages, internal links, and external mentions.
- A product may appear for a precise bottle-name query because the title, brand, size, and product attributes match closely.
- Google can later reassess whether that URL is the best result once it has crawled more of the store, competing retailers, or alternative pages on the same domain.
- Bulk edits to titles, copy, canonical tags, stock status, or internal links can trigger reprocessing and change which version Google considers useful.
In the Reddit example, visibility began on August 23, rose through late August, and dropped around September 10. That is a short window. It is too brief to diagnose a sitewide quality problem from rankings alone, but eight additional weeks of low visibility is long enough to inspect Search Console data, crawl samples, and template output.
A site should not assume that a low Ahrefs Domain Rating—2.3 in this case—explains everything. Link authority can affect competitiveness, especially in commerce, but it does not explain basic non-indexing or a sharp loss of impressions by itself. The first job is to separate an indexing problem from a ranking problem.
Start with the Search Console evidence
Before changing product copy or ordering links, compare the period before and after the decline in Google Search Console. Use a date range that covers at least the first appearance, the drop date, and the following eight weeks. For this case, that means comparing late August with mid-September onward.
Identify whether the loss is sitewide or concentrated
Export clicks, impressions, average position, pages, and queries. Then sort the data three ways:
- By page type: product, category, brand, blog post, pagination, and any filtered URLs.
- By query type: exact bottle names, generic category searches, informational wine queries, and branded searches.
- By device and country: especially important for a Vietnam-based store with locally relevant availability and language signals.
If impressions fell across all URL types—including the home page and categories—the investigation should prioritize indexing, robots directives, server access, canonicalization, or a broad site-quality issue. If only product pages fell while category and editorial pages remained stable, duplicate or thin product templates become more plausible.
Google’s URL Inspection tool should be used on a representative sample, not one URL. Inspect at least 10 URLs from each major template: an indexed product, an excluded product, a category, a brand page, a newer blog post, and a URL that once had impressions. Record the canonical Google selected, the last crawl date, index status, mobile rendering result, and whether the live test can fetch the page.
This distinction matters because “Crawled – currently not indexed” points to a different next step than “Duplicate, Google chose different canonical,” “Blocked by robots.txt,” or “Soft 404.” A useful website ranking issue diagnosis starts by proving which of those states applies instead of treating every traffic decline as a content problem.
A 6,750-URL sitemap is not an indexing plan
The store in the source had approximately 6,750 URLs in its sitemap, but a sitemap is a discovery hint, not a command to index every URL. Google’s documentation describes sitemaps as files that help search engines discover pages and understand their relationships; they do not establish quality, uniqueness, or search demand.
The reported 225 crawl requests per day sounds encouraging, but it should not be interpreted as 225 new products indexed every day. Googlebot may recrawl the home page, categories, scripts, images, existing product URLs, redirects, feeds, or URLs that ultimately remain excluded. At that theoretical rate, 6,750 requests would take 30 days, but real crawling is not distributed evenly and requests are not equivalent to successful indexation.
Improve the sitemap before submitting it again
The sitemap should contain only canonical, indexable, live URLs that the site genuinely wants to rank. For a WooCommerce store, that normally means excluding:
- Internal search-result URLs.
- Empty category and tag archives.
- Parameter and sort-order URLs.
- Products permanently out of stock where the page offers no replacement or useful information.
- Redirects, 404s, soft 404s, and URLs marked noindex.
- Duplicate variant pages when a parent product is the intended canonical result.
Splitting the sitemap index by page type makes diagnosis easier. Separate product, category, brand, editorial, and image sitemaps provide clearer submitted-versus-indexed reporting. It also makes it possible to see whether only products are struggling rather than concluding that the whole domain is affected.
Google says crawl-budget management matters most for large or frequently updated sites. Nearly 6,000 products is meaningful for a new retailer, but it is not automatically a crawl-budget crisis. More often, the preventable issue is crawl waste: Google reaches filters, duplicates, thin archives, and unnecessary URLs before it has enough confidence in the priority product and category pages.
Check WooCommerce URL, canonical, and filter behaviour
WooCommerce stores can create many URL states without an obvious visual change. A product may be accessible through a clean permalink, a category path, a tracking parameter, an on-site-search URL, a product-feed URL, and several filtered collections. If those alternatives send conflicting canonical, noindex, or pagination signals, indexation can become inconsistent.
Google recommends crawlable ecommerce navigation and explains that it learns site relationships from links between pages. That makes category architecture more than a usability exercise. A product that exists only in an XML sitemap is weaker than one linked from a category, a brand page, a related-products module, and an editorial guide where the connection genuinely helps shoppers.
Test these template-level conditions
For a 20-URL sample of products across brands and categories, check the rendered HTML—not only the WordPress editor—for:
- A self-referencing canonical pointing to the preferred product URL.
- An indexable 200 status code for in-stock products.
- No accidental
noindexoutput from an SEO plugin, staging configuration, or taxonomies setting. - A single, consistent URL format for trailing slashes, www/non-www, HTTP/HTTPS, and category paths.
- Crawlable HTML links to primary categories and key products.
- Proper pagination that exposes deeper category inventory without requiring JavaScript interaction.
- Filter parameters that are controlled so they do not create large numbers of near-duplicate indexable pages.
The same audit should test the server response Googlebot receives. Slow responses, repeated 5xx errors, bot challenges, geo restrictions, or consent overlays can affect crawling even when the site appears normal to an administrator. Google’s crawl-error guidance specifically recommends checking availability issues, pages that should be crawled but are not, and whether changes are picked up quickly enough.
Bulk content changes shortly after launch deserve scrutiny here. If a plugin recalculated canonicals, rewrote product descriptions, changed status codes, or added noindex rules across thousands of products, the timing around September 10 may be meaningful. Compare deployment logs and plugin updates with the decline date.
Audit product pages for real differentiation
The source describes a well-considered topical map, keyword grouping, product specification checks against official brand sites, entity links, and internal linking between drink types. Those are useful foundations. They still do not guarantee that 5,977 individual product pages provide enough distinct value for Google to index and rank separately.
This is particularly relevant to bottle-name searches. Many retailers can publish the same producer name, alcohol percentage, bottle size, tasting notes, generic description, and manufacturer image. If the WooCommerce catalog reproduces supplier data with minor edits, Google may see limited reason to retain every product URL when established retailers offer similar pages plus stronger reputation, stock history, reviews, delivery details, and links.
Google’s people-first-content guidance asks publishers to consider who created content, how it was created, and why it exists. At scale, those questions should be applied to the product template, not just the blog.
Make priority product pages materially more useful
It is rarely efficient to enrich all 5,977 pages at once. Start with 50–100 products that have search demand, commercial margins, reliable inventory, or demonstrated initial impressions. Give each page details a shopper could not get from a standard distributor feed, such as:
- Availability and delivery conditions specific to the destination market.
- Original tasting notes or serving guidance where the retailer has genuine expertise.
- Clear differentiation between vintages, bottle sizes, gift packs, and variants.
- Food-pairing, cocktail, or occasion guidance linked to relevant editorial content.
- Accurate images, current price, stock state, returns, and merchant contact information.
- A short explanation of who selected, reviewed, or sourced the product where appropriate.
This is not an argument for adding 800 words to every bottle page. A concise page with unique commercial facts can be more useful than a long description assembled from common supplier wording. The goal is a demonstrably better result for a specific shopper intent.
Use structured data and Merchant Center as commerce signals
The Reddit post mentions schema linked to relevant Wikipedia entities. Entity references may aid semantic context, but they are not a substitute for ecommerce-specific product data. For retail pages, Google recommends Product structured data and also recommends submitting product data through Google Merchant Center where applicable.
Product markup does not guarantee rankings. It can improve Google’s understanding of price, availability, shipping, and product attributes, and may make pages eligible for richer shopping-related appearances. Merchant Center is also required for some Google surfaces, including the Shopping tab.
For each priority product, validate that visible on-page details match the markup. At minimum, check the product name, image, description, SKU, GTIN where available, brand, price, currency, availability, and offers. If shipping and return policies are included in merchant-listing markup, they must reflect the store’s actual policies.
Run representative URLs through Google’s Rich Results Test, then monitor the Merchant Listings report in Search Console. Template edits should be deployed in small, observable batches. Changing markup, page copy, canonicals, and navigation across 6,000 products at once makes it harder to determine which intervention helped or hurt.
Do not confuse AI visibility with Google recovery
The store received traffic and customers from Bing, Yahoo, and ChatGPT while Google visibility was weak. That is commercially valuable evidence: real people can discover and trust the store through channels other than Google. It is not, however, proof that Google should index or rank the same pages.
Different systems use different discovery paths, retrieval methods, freshness signals, user contexts, and ranking objectives. A strong citation or mention in an AI answer engine may coexist with poor Google index coverage. Likewise, a page that ranks in Google may never be mentioned in an answer engine.
That is why teams should measure the channels separately. An AI search visibility measurement guide can help define prompts, track citations, and distinguish brand mentions from referral traffic. The broader difference between AEO and SEO is practical: SEO work must still address crawlability, indexation, query competition, and search-result performance, while answer-engine visibility needs its own evidence.
For this store, the sensible course is to protect the pages and information that already earn AI-assisted visits while investigating Google-specific failures. There is no reason to remove helpful cocktail guides or product facts merely because their Google rankings are not yet stable.
A 30-day recovery plan for a new ecommerce site
The right plan is controlled and prioritized. A new store should avoid reacting to a visibility dip by rewriting every page, disavowing links without evidence, or submitting thousands of individual recrawl requests.
Days 1–7: establish the baseline
- Export Search Console data by page and query before and after the drop.
- Review the Page Indexing report and classify the top exclusion reasons.
- Inspect 40–60 URLs across products, categories, brands, blogs, and pagination.
- Crawl the site with a desktop crawler to find 4xx/5xx responses, redirect chains, canonicals, noindex tags, duplicate titles, thin content, and orphaned pages.
- Check server logs or hosting reports for Googlebot access problems around the decline date.
Days 8–21: fix the highest-confidence blockers
Resolve robots, noindex, canonical, redirect, status-code, broken-link, and sitemap errors first. Remove non-priority URLs from sitemaps. Improve category links to priority products, then enrich a limited set of high-opportunity product pages. Validate Product and Merchant Listing markup after each template change.
Days 22–30: measure rather than guess
Request recrawls only for important URLs that changed materially, such as the home page, major categories, and selected priority products. Google notes that recrawling and reindexing can take days, so the evaluation window should be measured in weeks, not hours.
Track four numbers weekly: indexed priority products, impressions for priority product queries, valid merchant-item markup, and organic sessions or orders by landing page. If those are improving while the total indexed count remains modest, the store is moving in the right direction. If they are flat after technical fixes, the next hypothesis should be page value and competitive differentiation rather than simply more crawling.
When waiting is reasonable—and when it is not
Waiting is reasonable when the audit confirms clean indexability, Googlebot can fetch pages, sitemaps contain only canonical URLs, key categories are internally linked, and the store is steadily gaining indexed priority pages. New domains often need time to earn recognition and links, and Google does not provide a fixed timetable for stable rankings.
Waiting is not a strategy when evidence shows major exclusions, incorrect canonicals, inaccessible product pages, broad soft-404 classification, unstable templates, thin duplicate copy, or a large volume of crawlable parameter URLs. In the Reddit scenario, 20% product indexation and a sharp post-launch fall justify investigation immediately, even if no single defect is found.
The practical conclusion is not that a 10-week-old store has failed. It is that the site should make its next 30 days measurable. Fix what Search Console, rendered-page checks, and crawl data demonstrate; retain the product and editorial work that helps real customers; then allow Google time to process a cleaner, more differentiated catalog.
FAQ
Why did my new ecommerce site rank briefly and then disappear from Google?
Early impressions can occur while Google discovers and evaluates new pages, especially for highly specific product-name queries. A later drop may reflect normal reassessment, but it can also expose canonical, crawlability, duplication, or quality issues. Compare indexed status, query groups, URL types, and deployment dates before assuming the site is in a sandbox.
How long does it take Google to index thousands of WooCommerce products?
There is no guaranteed timeline. Google can find new pages within days, but discovery does not ensure indexing, and individual products may be excluded if they are duplicate, low-value, inaccessible, or weakly connected internally. A 6,000-product launch should prioritize indexation quality for commercially important pages rather than expect every URL to appear at once.
Should a store submit every WooCommerce product URL in its sitemap?
Only canonical, indexable, live product URLs should be included. Exclude redirected, noindex, filtered, internal-search, duplicate-variant, and permanently unavailable URLs that offer no useful replacement. Splitting product, category, brand, and blog URLs into separate sitemaps makes Search Console reporting more useful and exposes page-type problems sooner.
Does ChatGPT traffic mean a site is doing well in Google Search?
No. ChatGPT referral traffic or citations show that the site can be useful in an AI-assisted discovery journey, but Google Search uses its own crawling, indexing, and ranking systems. The business should retain what is producing customers while measuring AI visibility and Google organic visibility separately.
What should be fixed first after a new ecommerce site loses Google impressions?
Start with Search Console and representative URL inspections. Fix pages blocked by robots or noindex, wrong canonical tags, server errors, soft 404s, redirects, broken internal links, and low-quality sitemap entries before rewriting content. Then improve a small set of priority product pages with distinctive, accurate shopper information and validate product structured data.
Sources
- https://www.reddit.com/r/bigseo/comments/1xocdm8/new_wine_and_spirits_store_had_google_visibility/
- https://developers.google.com/search/docs/crawling-indexing
- https://developers.google.com/search/docs/crawling-indexing/large-site-managing-crawl-budget
- https://developers.google.com/search/docs/specialty/ecommerce/help-google-understand-your-ecommerce-site-structure
- https://developers.google.com/search/docs/specialty/ecommerce/share-your-product-data-with-google
- https://developers.google.com/search/docs/fundamentals/creating-helpful-content
- https://developers.google.com/search/docs/crawling-indexing/troubleshoot-crawling-errors
- https://developers.google.com/search/docs/crawling-indexing/sitemaps/build-sitemap