When people talk about building an online store, Tilda is still often seen as a solution for a small range. A few dozen products? Fine. Several thousand? Better look for another platform right away.
That view no longer matches what modern Tilda can do. The catalog holds tens of thousands of products, supports variants, import, sync with accounting systems, filters, payment and delivery. Product count is only one parameter. What matters more is how the range is structured, what happens to products after launch and which processes around them need automation.
I develop and promote commercial websites. When choosing a platform, I look at both how quickly sales can start and how the store will live and grow a year later.
Why product count isn’t the main question
First, the numbers everyone starts with. According to Tilda’s docs, you can add up to 5,000 products on Personal and up to 50,000 on Business, with up to 1,000 categories on both plans. The platform also recommends splitting products into sections of about 100–200 items for buyer convenience. The catalog is only available on paid plans. If you downgrade to Free, products are kept for 30 days and then deleted permanently — worth knowing early, if only to export a backup in time.
Those figures say little about real complexity. Imagine two stores: one sells 300 home décor items, the other 300 types of auto parts. Same count, radically different site requirements. In the first case the buyer picks color and size and checks out. In the second they may need to find a part by make, model, modification and specs.
A large catalog and a complex store are not the same thing. A clothing shop may run thousands of SKUs with a small set of standard attributes. A specialized industrial equipment shop may have only a few hundred lines yet need a configurator, custom pricing and integration with internal systems. The second project can be harder even with fewer products.
Catalog and product data: what happens after launch
The catalog is the core of any online store. How easy it is to manage drives how much time staff spend adding products, updating prices and handling the range. In Tilda, products can be created manually or imported from CSV and YML. Products can have variants — one shoe model in several sizes and colors — with separate images per variant.
Things get interesting when products need updating. Import docs state that each product has a system Tilda UID. If you ignore it on re-import, products won’t update — they’ll load as duplicates. The site suddenly has two copies of every line and someone has to clean up by hand. The same docs describe linking variants via Parent UID and external code. Dry documentation, but that’s the difference between “updated prices in five minutes” and “lost a day”.
Filters depend on data quality just as much. Picture a technical fabrics store where buyers choose by weight, roll width, color and use case. Consistent attributes enable sensible filtering. Mixed text formats, different units and missing parameters break even a good filter. Without preparation, a convenient admin UI quickly becomes a messy pile of cards.
So store development starts with a product model: which attributes are mandatory, which feed filters, which affect price and which matter for SEO. Filter docs allow filtering by variant properties, characteristics, stock or category. Price filters show items in a range with a ±1% tolerance — a small detail buyers notice when “up to 1,000 rubles” still shows 1,010.
Accounting sync: which system is the source of truth
For a store with a frequently updated range, manual edits become painful. If the company keeps books elsewhere, data exchange makes sense. Tilda’s docs mention catalog sync with 1C, MoySklad and SABY, and an external code field on the product card to tie items to those systems.
Being able to connect 1C doesn’t mean every business process runs on its own. Before integration you need to decide which system is the master source, how often stock updates, what happens when a price changes and how exchange errors are handled. Without answers, integration just spreads existing confusion across the site faster.
Multiple warehouses, suppliers and price types are a separate story. One store may only need stock from a main warehouse. Another must show availability in several cities and let the buyer pick a pickup point. Those scenarios need their own analysis even when accounting sync is supported.
Products first, storefront second
If the store already lives in spreadsheets, accounting software and the manager’s head, start with the product model and exchange rules — card design comes on top of solid data.
Cart, payment and wholesale: where the standard flow ends
For ordinary retail, Tilda has what you need. The buyer picks a product, adds it to the cart, leaves contacts, chooses delivery and pays if required. Payment systems, promo codes and upsells are available. The catalog also supports reservation: per the docs, a product can stay reserved until payment is confirmed or for a set period.
Buying scenarios differ. In retail the price is known upfront; the buyer places the order and the system sends data to a manager or accounting. In wholesale, price may depend on batch size, client terms, delivery method and stock at a specific warehouse. Checkout becomes a request for a quote and negotiation.
Some non-standard flows can be built with forms, JavaScript and external services — custom development territory. Each integration needs maintenance. If much of the store runs outside the platform, the constructor’s initial simplicity fades.
What changed in ST340 — and what it doesn’t promise
A fresh reason to revisit the topic is the ST340 update, the main block for catalog product grids. Tilda’s September news and an October 28, 2026 post describe flexible layout for price, buttons and image, new card styles and two block variants, ST340G and ST340H.
For the storefront it’s a clear win: easier to match brand style without heavy custom work. But the update is about how a product looks in the tile. It doesn’t change data exchange, price update rules, multi-warehouse logic or the buyer journey. A pretty preview still shows exactly the data that feeds it.
The same applies to search. Built-in filtering doesn’t automatically create full SEO landing pages for every attribute combination — a buyer may search for “PVC tarp fabric 650 g/m²” and you need a dedicated category or selection page. You can also link a catalog item to a custom landing: the card has a field for a description page, and the tile leads there. I covered indexing and SEO tools in SEO on Tilda: what works and where the limits are.
Tilda or another platform: how to decide
Start with money. Tilda’s subscription is relatively cheap: on the pricing page as of November 8, 2026, Personal is 500 rubles/month and Business 1,000 rubles/month when billed yearly. A professional store costs far more than the plan. My ballpark for development starts at 90,000 rubles; the final budget depends on complexity — a few dozen products with a standard cart vs thousands of attributes and stock sync are different jobs. Factor in ongoing costs: accounting, integrations, catalog upkeep, support and marketing.
I compare platforms against the specific store. Tilda fits shops with a clear range structure, standard checkout logic and high demands on custom design. Creatium is worth a look if you need dynamic pages and tabular data, but evaluate commerce separately. Classic CMS and dedicated e-commerce platforms are often better when you need heavy server logic, rich data management and deep ties to internal processes.
Switching platforms doesn’t solve problems by itself — any complex system needs sound architecture and careful build. A broader view of Tilda’s strengths and limits is in Tilda website development: pros and cons.
I start store projects by studying product structure, sales processes and search demand. Product count is on the list too, but far from first. That order helps pick technology you can grow on without rebuilding the site every quarter.