A concrete display of one cycle of monthly maintenance of the WooCommerce store — From diagnosing indexing problems in Google, through clearing crawl budgets, to preparing a site for AI search through structured data.
When we say “monthly site maintenance,” most people think of plugin updates and backups. That is part of the job — but only a small part.
Serious maintenance means that someone is looking at the data every month, looking for the causes of problems that you didn’t even notice, and solving them before they start costing you.
This text is a representation of such a cycle — a concrete overview of the work on the WooCommerce store Bug in the Car, one of the leading car seats and equipment stores for children in Serbia.
We publish it because we believe that customers should know exactly what they are getting for the money they set aside for maintenance.
The Beetle in the Car is a WooCommerce store with a wide range of products for the safety of children in the car. — new design, more transparent navigation, optimized purchasing process. About what the redesign includes and What affects its price? We have written in more detail in a special text.
After the launch, the cooperation continued through monthly maintenance. And it is here that the work begins that is not seen at first glance.
The first task was to analyze the data from Google Search Console. The result was a clear signal that something was wrong:
Google indexed only about a third of known URLs on the site.
In other words, most of the site practically did not exist for Google search. By analyzing the exported data from GSC, we identified three main causes.
This was the biggest problem — and the most difficult.
The booking widget on each product page generated a unique URL every time Google crawled it. Google would visit the page, receive a new address, return and receive yet another one.
In SEO terminology this is called Crawl Budget TrapGoogle allocates a limited “budget” to each site. — how many pages will be visited in a certain period of time. When this budget is spent on an infinite number of useless addresses, there is nothing left for the actual product pages.
Solution: We’ve redesigned the way widgets load so they only start at the user’s click, instead of rendering in advance on the server.
What it means in practice: Google is now spending its crawl budget on product pages. — The ones that need to be ranked and sold.
WooCommerce automatically creates an archive page for each product attribute. In this case, the “Color” attribute is generated 92 separate pagesMost of which contained one or two products.
Google refers to such pages as thin content — pages without enough standalone value — and Google does not usually index them, but still crawls them, which means they consume crawl budget.
Solution: The “Colour” attribute archive is disabled. Filters by color still work normally for users — They just don’t generate the separate pages that Google has to visit anymore.
After the migration of the URL structure, several old paths and shut down products remained without proper routing.
Solution: 301 directories have been set for muted products and remaining old URL templates. This takes every old link — whether it comes from Google, from social networks or from someone’s note — Refer to the relevant page instead of the 404 error.
We further confirmed that there is no duplicate content between Cyrillic and Latin versions of the brand pages.
In addition, unnecessary technical elements that WordPress and individual add-ons automatically add are removed from the site code — redundant links, category feeds, and tags that publicly reveal versions of the tools used.
The effect is twofold — Less unnecessary addresses that Google has to visit, and less risk of unnecessary disclosure of technical information about the site.
The second part of the work focused on how the website communicates with Google and AI systems such as ChatGPT, Perplexity and Google AI Overviews.
This is done through structured data (schema markup) — code that is not visible to visitors, but tells search engines and AI models exactly what data is on the site. how AI search chooses sources to quote We have written in a separate text.
More precise type of entity. Generic Organization Markup has been extended to more specific types OnlineStore and LocalBusiness. Instead of “this is some sort of organization”, the site now clearly states ‘this is an online store that also operates as a local business’.
Complete business data. The structured data includes address, phone, email, opening hours, social profiles, connection to Google Maps/Business profile, as well as business and delivery areas.
Delivery information. It is implemented globally ShippingService Markup — Dexpress delivery on the territory of Serbia, delivery time up to five days and free delivery for orders over 10,000 RSD.
Conditions of return. It is implemented MerchantReturnPolicy Markup — 14-day refund period, possibility of refund or replacement, with clearly separated refund costs on withdrawal and defective product.
Google is increasingly showing this information directly in search results. — delivery time, refund conditions, opening hours. Customers see them before they even click through to the website, which increases trust and the likelihood of a click.
The same goes for AI systems. When someone asks ChatGPT or Google AI “Where to buy a car seat in Serbia with free shipping”. AI is looking for sources that clearly state this data in a structured form. A store that has them — has a chance to be mentioned. A store that does not have them — none.
Several decisions are worth highlighting, as they explain the difference between “addition of schema markup” and proper implementation:
Integration into the existing structure. New data is embedded in the existing schema graph of the site, not as a separate block. If we added them separately, the site would send two competing descriptions of the same business to Google — It creates confusion instead of clarity.
Check before commissioning. Each change has been tested through official structured data validation tools, with manual verification of key delivery and return values, and verification of the correct display of our letters.
Accessibility checks revealed that the “Child Height” and “Product Type” filters do not have a properly defined description in the code.
In practice this means that users using screen readers could not know what those filters were for.
Solution: Filters have been added appropriate descriptions, without any visual changes to the site.
Accessibility is often neglected, but it has a double value — Enables the use of the site for people with disabilities, and represents a signal of quality that search engines take into account.
| Authority | Interventions | The effect |
|---|---|---|
| Indexing | Stopped endless URL generation | Crawl budget focused on product pages |
| Indexing | Excluded “Color” attribute archive (92 pages) | Removed index noise |
| Indexing | 301 directories for old URLs | No lost links and 404 errors |
| Code | Purified redundant technical elements | Less Addresses, Less Security Risk |
| Structured data | OnlineStore + LocalBusiness Entities | Clearer identification for Google and AI |
| Structured data | Delivery and terms of return | Visibility of information directly in search |
| Accessibility | Screen Reader Filter Descriptions | Site usable with assistive technologies |
A few points are worth highlighting.
This guy’s problems are not visible to the naked eye. The site has been running normally all along. — products were displayed, the purchase worked, nothing was “corrupted”. And most of the site was not in the Google index. About other signs that are not immediately noticed we wrote in the text about 5 Signs Your Website Needs Refreshing.
The cause is almost always technical, not substantive. It is easy to assume that the site is not going well because there is not enough text or blog posts. In this case, the content was there — Google didn’t get to see it.
Data is the starting point. Without the analysis of the exported data from Google Search Console, these problems would remain invisible. That’s why maintenance starts with reading the data, not opening the code.
The solution must be precise, not random. When the cause is accurately determined, the change is small and targeted — a few lines of code instead of rewriting the whole site.
Basic maintenance includes security updates, backup and availability monitoring. Serious maintenance also includes regular data analysis from Google Search Console, performance monitoring, technical SEO review and troubleshooting before they affect traffic.
For smaller sites, once a month is enough. For WooCommerce stores with a large number of products, more frequent checking is recommended, because the problems with indexing there occur faster and have a direct impact on sales.
The most common causes are thin pages without enough content, duplicate content, technical errors that consume a crawl budget, or a poorly set URL structure. The precise cause is determined by analyzing data from Google Search Console.
Crawl budget is the number of pages that Google will visit on your site over a period of time. For larger sites, this resource is limited — if spent on useless addresses, important pages remain unvisited and unindexed.
Schema markup is not a direct ranking factor, but it affects how the site is displayed in search results and how easily search engines and AI systems understand it. This is reflected in practice on the number of clicks and visibility.
Continuous. Site is changing — products are added, plugins are updated, browser rules are changed. Maintenance is a process of tracking and customization, not a one-time intervention.
Monthly site maintenance is not a list of routine tasks. It is a process in which you look at what the data shows each month, look for the cause of what is not working properly, and solve it before it becomes a visible problem.
In this cycle, this meant: identifying the reasons why most of the site was not in the Google index, stopping the process that was spending the crawl budget, preparing the site for an AI search through structured data and accessibility fixes.
None of this is visible to the visitor, but it all determines how many visitors will come.
Do you have a WooCommerce store or website that requires serious maintenance? Contact us — We’ll be happy to review it, with no obligation.