Promotion of an online store with a large catalog: we start with the number of pages
Promotion of an online store with a large catalog begins with arithmetic, not with semantics: how many pages do you have, how many of them reach the search engine and how many addresses per product card. These three numbers set the plan — which sections we lead by hand, which we close with template rules, and which we remove from the bypass altogether. An example from my own measurement on 07/31/2026: a cosmetics store with 5,561 positions, the site map gives code 200 with a body of zero bytes. The search engine does not receive a list of pages at all, and no semantics can compensate for this.
two independent calculations of the directory — a site map and a standard search engine
after a free audit
Price
30
Guarantee
days after project sign-off
two independent calculations of the directory — a site map and a standard search engine
Where do we start?
free, 3-6 business days
Measurement before the estimate
45-180
Full cycle period
business days including monthly lead times
5,561
From own measurement on 07/31/2026
items in the catalog of cosmetics; 8,358 addresses in the technology showcase map for 4,000 products
how many addresses does one product card generate
What are we counting before the start?
by sections of the catalog, and not by the site as a whole
How we report
the number of pages in the index and positions is the decision of the search engine
What we do not promise
Process steps
Transparent stages with approval at every step
Total duration:45–180 days
1
Measurement and starting numbers
3-6 working days
We count the catalog using two methods, we analyze the site map by address types, we record the speed on the main page, in the category and on the card. Everything that is measured goes into the document and becomes a point of reference.
2
The first turn: return to bypass what is already there
10-20 working days
Sitemap by page types, address in robots.txt, availability of deep sections, response time under crawling load. This is the shortest queue with the most noticeable effect, so it goes first.
3
The second turn: remove what is eating up the bypass in vain
10-25 working days
Gluing category paths, canonical addresses for filter parameters, traversal rules for sorting, language branch reconciliation. We warn you right away: the number of pages in the index may temporarily drop here.
4
Rules and markup for the entire catalog
8-20 working days
Templates for page titles, titles and descriptions by section types, Product and breadcrumbs markup. We write the rules so that new products pick them up themselves, without the involvement of the manager.
5
Priorities and boarding on demand
10-30 working days
We divide the catalog into three groups, collect the demand by priority sections, start landings where there is demand, and do not start where there is no demand. Texts are written only for this part of the catalog.
6
Monthly driving cycle
4-8 working days every month
A section of traversal and indexing by section, analysis of new finds, the next batch of landings. The report shows movement by department: the average for a large store does not explain anything.
Technologies & integrations
What we build on and what it connects to
Stack
Bypassing the sitemap index with deduplication — counts unique addresses, not records; shows the intent of the site, not the fact of indexing
The standard search engine is a second independent counting method; does not work where the search is broken or returns an empty output
Screaming Frog is a directory traversal on our end; sees what a crawler sees, not what a search engine bot does
Server logs are the only source of real bot visits; Not every hosting has it
Search Console — data from the search engine itself; are given with a delay and without details for each address
Template metatag rules in the theme code — close the entire directory; require developer access
JSON-LD Product and BreadcrumbList, rel=canonical — change the appearance of the snippet and paste duplicates, but not the position
GA4 — links the sections of the catalog with money; shows a trend, not an accounting truth
Integrations
Google Search Console
GA4
Google Merchant Center
Bing Webmaster Tools
Cloudflare
What's included
Complete list of work and what you get as a result
Measuring the catalog by two methods and fixing the starting numbers — then each change is measured from a reference point, not from sensations
Calculation of how many addresses one card generates and a plan to reduce this number - the bypass stops going in circles for the same products
Template rules for page names and titles for the entire directory - tens of thousands of pages get their own metadata without manual work
Prioritization of sections into three groups: lead by hand, close with a template, close from indexing — work goes where there is demand
Gluing category paths and working with language branches through canonical addresses — signals are collected on one page instead of three weak ones
Sitemap broken down by page type, with address in robots.txt and update dates that actually change - the file is visually verifiable
Landing pages for the demand that is visible in the data, instead of thousands of empty filter pages
Product and breadcrumb markup applied by template to the entire catalog - this affects the appearance of the snippet, not the position
A separate list of technical findings for your developer - formulations that are taken into the sprint without passing through the manager
A monthly report with the dynamics of crawling and indexing by sections of the directory — you can see which section has come to life and which is standing still
When this service isn't right
What's not included — so there are no surprises at delivery
Filling cards with descriptions throughout the catalog — for tens of thousands of items, this is a separate budget and a separate service page
Transferring and normalizing product data from suppliers' prices
Price aggregators, marketplaces and contextual advertising
Promises of the number of pages in the index - this is at the disposal of the search engine
Who it's for
Situations where this service delivers results
Scenario 1 of 5
The catalog is growing, and the transitions from the search stand still
Hundreds of items are added every month, nothing moves in the reports. Most often, new pages simply do not make it to the crawl: the bot spends its resources on repeatedly crawling the same products at different addresses. It is not more content that helps, but fewer addresses.
We'll review your situation in a free audit
Scenario 2 of 5
No one in the company will name the exact number of pages on the site
The admin says one thing, the site map the other, the webmaster's report the third. Until those numbers are tallied, any plan is a guess. We calculate the catalog using two independent methods and show the discrepancy between them; at the measured stores, it was kept within a few percent, and each percent here is hundreds of products.
Scenario 3 of 5
One card lives at several addresses
Category paths, language branches, filter parameters—each multiplier adds copies. In terms of the showcase of household appliances, 4,000 products and 170 categories gave 8,358 addresses in the site map, because each page is available in two language versions. This is still a normal case; it gets bad when the multipliers are three.
Scenario 4 of 5
You have already written texts on categories, and nothing has changed
On tens of thousands of pages, manual content does not scale: while writing texts for a hundred categories, the rest of the catalog remains without page names and headings. The order of actions is reversed - first the rules for the entire catalog, then the texts where the demand is visible.
Scenario 5 of 5
The assortment is uneven: a smaller part of the sections brings money
The normal state of a large store, and it is this that allows you not to do everything at once. Sections are divided into three groups — we guide with our hands, close with a template, remove from indexing. Without such a division, the budget is smeared with an even layer across the catalog, in which half of the items have no demand at all.
Maintaining a large catalog versus "we'll write texts on categories"
How this option differs from the alternative
Package "texts by category"Our approach
Where does the work begin?from the list of categories given by the adminfrom counting the directory by two methods and analyzing how many addresses are on the card
Directory coverageas many pages as they managed to write with their handsrules close all pages, including those that appear tomorrow
What is done with duplicate addressesnot included in the scope of workare calculated, reduced to templates, closed by canonicals and bypass rules
Procedureadd new immediately, out of bypass statefirst return the existing bypass, then remove the excess, then add new ones
What the report looks likethe number of submitted textsdynamics of browsing and indexing by sections of the directory
Free measurement of a large catalog
The first question at a big store is how many pages do you have. The second is how many of them the search engine bypasses. The difference between these two numbers is the working space, and we measure it for free.
What we measure
Directory size by two methodsThe site map and standard search engine are separate. The difference between the methods shows how much of the directory is left out of the list that the bot receives.
How many addresses does one card generate?Category paths, language branches, filter parameters. A real example from measurement: 4,000 products and 170 categories turn into 8,358 addresses in the sitemap, because each page exists in two language versions.
Is a site map provided at all?Let's look at the body of the answer, not the code. The most expensive find of a big store is a card that gives 200 with zero bytes; we encountered this in the catalog in 5,561 items.
Speed under directory loadThe time of the first answer on the main, in the category and on the card separately. In large catalogs, these figures differ the most.
Bypass priorityWhich sections the bot visits often, which — once a month. This is visible from the webmaster's reports and determines where it makes sense to put effort.
Analytics statusDoes the site have a working measurement property. On a large catalog, without it, it is impossible to determine which sections bring money.
What you get
Two independent numbers of catalog size and percentage of discrepancy.
Calculation of how many addresses are on one card and where the multiplication comes from.
The state of the sitemap, speed and analytics in numbers.
Conversation for 30 minutes on the document.
Timeline: 3-6 working days
Why is it free
Because it is impossible to even estimate the volume on a large catalog without measuring: the difference between 8 thousand and 60 thousand addresses changes the plan, the term, and the amount. We don't want to randomly name a number and you want to overpay for a guess.
What's next
After the measurement, we draw up a plan in three stages: return to the bypass what is already there; remove what eats up the bypass in vain; add new landings where there is demand.
Short form: your contact and site URL
Did not find your case?
Describe how it works on your side — we will tell you whether “SEO for an online store with a large catalog” fits and what it means in your situation. No brief and no call: one question, one answer.
Contract, act and 30-day warranty
Every project gets a written contract: scope, deadlines, amount, acceptance procedure. After delivery — act and invoice, then 30 calendar days of warranty.
Sole proprietor & bank transfer
The contractor is a registered sole proprietor. Payment by invoice with closing documents.
Rights & access — yours
Code, design and materials transfer to you after full payment. Domain, hosting, repository and analytics are registered to you.
Client portal instead of email chains
During the project you get access to a portal: contracts, invoices, acts and project status in one place.
European clients
Among our work — projects for Norway, Bulgaria, Moldova and Spain.
Verifiable numbers
Every case in the portfolio comes with a link to a live site and a technical measurement.
Audit first, then pricing
There is no price list on the site intentionally: the scope of the same work differs multiples between clients.
We say "no" when unsure
If the task isn't ours or the deadline is unrealistic — we tell you upfront.
What affects the price
Why two seemingly identical tasks are priced differently
How many addresses are on one cardThis is the main multiplier of the volume of work. One address per product — work goes through the catalog once. Four addresses through category paths, languages, and parameters — you have to describe the rule for each multiplier separately and check that they don't conflict with each other.
Homogeneity of the structureA store with 55,000 items with one address mechanism is easier than a store with 4,000 items with three different ones. The number of products in the price says almost nothing about labor intensity; the number of rules to be described says it all.
The number of language branchesEach language doubles the list of addresses and adds a separate check: are all the branches in the site map correctly connected to each other, or do they not compete in the same issue. It's not doubling the work, but it's not free either.
Who implements the rulesTemplate rules live in the template code, not in the admin text boxes. If you have your own developer, we give the specification and check the result. If there is none, implementation falls on us — and this is a separate line of work with its own deadline.
Driving depthHow many sections do we keep by hand every month. Ten priority sections and fifty are different lunar cycles, although the technical part is the same in both cases.
Status of accesses at the startReady-made Search Console with a year's history and access to logs shorten diagnostics to a few days. An empty webmaster account means that part of the conclusions will have to wait until the data accumulates.
Cases
Tasks and results in numbers — all metrics measured by us
online store of professional cosmetics and care products, catalog of over 5,000 items
Task
Calculate the real size of the directory and check what is available to the search engine.
Solution
Page-by-page walkthrough of the standard search engine, checking the site map and product feed, speed measurement, counter inventory.
Result
5,561 products in the live catalog - counted page by page: 55 pages for 100 items and 61 for the last one. First server response 801.2ms, full download 930.7ms with 153.8KB markup. At the same time, /sitemap.xml and the engine feed give 200 with an empty body — there is actually no site map. No counter was found on the main one: no tag manager, no analytics, no pixel, that is, conversions on the catalog of five and a half thousand items are not measured at all. The site is monolingual, there is no hreflang. Measured on 07/31/2026.
What we need from you
We can't start without this — best to prepare in advance
1Access to Search Console - Without crawl reports, working on a large directory becomes guesswork.
2Access to server logs, if available: This is the only way to see where the bot actually goes.
3CMS or Developer Access - Template rules are implemented in code, not in text fields.
4Business priorities: which sections of the catalog bring in money. Without this, prioritization will be formal.
5One negotiator on your side with the right to say yes to the directory structure.
If something is missing — let us know, we'll help you gather it or do it as a separate task.
FAQ
Most frequently asked questions — with concrete answers
How is this different from technical indexing of a large directory?
The technical page answers the question "how to make the bot reach all pages". This one is for the question "which pages should be kept in general and in what order". In a catalog with tens of thousands of items, it is physically impossible to manage them all in the same way, so prioritization here weighs no less than technology. Works usually run in parallel: priorities without a technical part rest on the fact that the bot simply does not reach the priority sections.
How many products are considered a large catalog?
It is not about the goods, but about the number of addresses and whether the bot has time to bypass them. We measured catalogs with 4,000 and 5,561 items, and in both cases the number of products explained little: in the first, 4,000 products and 170 categories gave 8,358 addresses in the sitemap. The practical threshold comes when one card begins to live at three or four addresses, and you can no longer name the number of website pages off the top of your head.
We have a site map that opens, so everything is fine?
Opening does not mean it is working. You need to look at the body of the response, not the code: in the cosmetics store at position 5,561 /sitemap.xml and the engine feed gave 200 with zero bytes, and outwardly it looked like a working map. Fair margin: the map does not raise positions and does not guarantee indexation, Google writes this directly in its documentation. It only gives the bot a list of addresses — but without this list in a large directory, some pages remain found only through internal links for years.
Is it possible to just close the excess in robots.txt?
Partly yes, but it's a crude tool. Closing in robots.txt does not remove the page from the index if there is a link to it, and at the same time prevents the bot from seeing the canonical link. Therefore, we combine: bypass rules for parameters, canonicals for options, redirects for dead branches. And immediately about the consequences: cleaning duplicates does not add traffic - it stops scattering it, and in the short distance the number of pages in the index may even drop. This is the expected course of events, but it is better to know about it before, not after.
Do you need texts for all categories of a large catalog?
No, and this is a rare case where the answer is unequivocal. Template rules for page names and headings cover the entire directory in a few days, and texts are written only for sections where demand goes - usually dozens of pages, not thousands. The main advantage of the rules is not savings: the new product picks them up by itself, without the involvement of the manager, so the catalog does not degrade as it grows.
Two language versions double the catalog. is it bad
By itself, no, if the versions are correctly connected to each other. Measured example: 8,358 addresses in the sitemap precisely because each page is present twice. It gets bad when only one branch gets into the map, and we have seen this more than once. Separately, it is worth saying honestly: the translation of the interface is not yet an entry into a new market, and the language version multiplies the volume of content, which must be constantly maintained.
We do not have analytics on our site. does it interfere with work?
It interferes, and more strongly than it seems. The 5,561-position measured store on the main had no counter — no tag manager, no analytics, no pixel. Without it, it's impossible to tell which section of the catalog is bringing in the money, and prioritization becomes a matter of taste. We put meters first, but without illusions: the meter itself does not improve anything, it only gives an opportunity to see. The data must be read further, and GA4 shows a trend, not an accounting truth.
How long does the work last and when are the first changes visible?
45-180 working days for a full cycle, and this is not continuous work, but queues: the technical part, then rules and templates, then landing, then monthly driving circles of 4-8 days each. The first changes in the scan reports are usually visible 4-6 weeks after the technical queue. We show the dynamics by sections of the catalog, because the average for a large store does not explain anything: one section doubled, the other failed, the total is equal.
Do you promise that the entire directory will be indexed?
No. The number of addresses in the sitemap is never equal to the number of indexed pages, and the decision here is made by the search engine, not the contractor. We are responsible for the rest: the page is reachable, not duplicated, delivered quickly and has its own metadata. A product card without a description, without demand and without differences from the neighboring one can be indexed and not receive a single visit - that is why we deliberately remove part of the catalog from the circulation, and not drag it into the index.
The response is two independent numbers of the directory size, the difference between them, the number of addresses per card, and a list of what is currently preventing traversal. If the catalog turns out to be small, we will say so in the first email.
From measured cases5,561 products in the live catalog - counted page by page: 55 pages for 100 items and 61 for the last one
There is no price list on the site on purpose: the same work differs several times over between two clients, and a “from” figure explains nothing in that case. First a free audit — we count your pages, duplicates and speed — then we name the sum and the deadline and fix both in the contract.