Most ecommerce platform round-ups are written as if the choice were between logos. It isn't. You are choosing a model: rent someone else's system, install your own, or have one built from scratch. Everything else follows from that decision. Below is a breakdown of the three models and the platforms inside each, from a studio that launches stores on them every month.
The short version for anyone deciding right now. If the store is typical and you need to launch fast, take a SaaS builder. If you plan to live on this site for years and grow organic traffic, run a CMS on your own hosting. If your business has logic that no box contains, go custom — and the sooner you admit it, the cheaper it turns out.
The three models you are actually choosing between
| Model | What it is | You own | Ceiling |
|---|---|---|---|
| SaaS builder | Renting a ready platform for a monthly fee | Content and domain only | The platform's feature set |
| CMS on your hosting | You install the system and run it | Code, database, files | Your technical team |
| Custom development | The system is written around your processes | Everything | Budget |
The main mistake at the start is counting only the subscription. Real cost has three parts: launch, monthly payments and the cost of leaving. The last one is usually never discussed, yet it is exactly what determines the price of the mistake if in two years the platform stops fitting the business.
SaaS builders: fast, as long as you're typical
The model is simple: you pay monthly and get a working store. Updates, security and uptime are not your problem. The price of that comfort is no access to the code — you can only do what the platform's developers anticipated.
Horoshop
The most prominent Ukrainian SaaS platform and one of our working tools, so the numbers here are verified against the source rather than copied from other round-ups. Annual-billing plans: Basic 288, Standard 816, Pro 1,728, B2B 3,360 UAH/mo. Product limits are 1,000 / 20,000 / 50,000 / 50,000 respectively.
Its strength is local integrations out of the box: Nova Poshta, Ukrainian payment services, instalments, feeds to Rozetka.ua and Kasta.ua, and a broad set of Ukrainian CRMs. Western platforms have no such layer — you would be assembling it by hand.
Three details that are easiest to miss while choosing: the API only opens on the Pro plan (Basic and Standard have none), marketplace feeds work from Standard up, and the platform logo can only be removed from Pro. There are no dedicated connectors for BAS or 1C — out of the box the platform works with Dilovod, Torgovsoft, ERC, DNTrade and Drop Cloud. More detail: Horoshop pricing, integrations and platform reviews.
Prom.ua
It is often listed next to builders, but it is a different model — a marketplace. You sell inside someone else's venue and compete with other sellers on the same product page. That is fast access to existing traffic, but the brand and the audience stay with the venue, not with you. As a starting sales channel it works; as a replacement for your own site it does not.
Western builders: Shopify, Wix, Tilda
Technically mature and pleasant to use, but in Ukraine they all hit the same wall: local integrations. Delivery, domestic payment services, feeds to Ukrainian marketplaces — each is either a third-party app, a developer task, or simply unavailable. Add pricing tied to foreign currency. For a store selling inside Ukraine that is an extra layer of complexity for nothing.
Other platforms that come up in the decision
Shop-Express and Weblium are Ukrainian SaaS products that often sit next to Horoshop in comparisons. Ecwid is not a separate platform but an embedded store: a catalogue-and-cart widget you drop into an existing site, which suits the case where the site exists but produces no sales. Magento (Adobe Commerce) is a heavy open-source enterprise system: it handles very large catalogues but needs a dedicated team and servers, so for a small business it is overkill. We work with Horoshop and WordPress, so we describe the rest as options on the market rather than as our specialism.
CMS on your own hosting: slower start, higher ceiling
Here you install the system on your own hosting and own everything: code, database, template. Updates and security move to you as well — that is real work, not a formality.
WordPress + WooCommerce
The most widespread pairing in the world and our second working tool. The upside is unlimited flexibility: any logic can be written, any integration implemented, and URL structure and templates are fully under your control. It is the best SEO foundation among mainstream solutions precisely because nothing stops you from doing it properly.
The downside is that you are responsible for it. Plugins conflict, updates break layouts, and a site left alone for a year becomes a security hole. WooCommerce needs maintenance, and if the business has neither a developer nor a support contractor, this option will turn into trouble. We compared it with SaaS over a multi-year horizon separately: Horoshop or WordPress, and what our WordPress development covers is on its own page.
OpenCart
Lighter than WooCommerce and built specifically for retail. It comes up regularly where SaaS cannot handle non-standard inventory — for instance when a product is sold by length, weight or cut rather than by the piece. The trade-off is a smaller ecosystem and heavier dependence on one specific developer.
Custom development: when the box doesn't fit the business
This is not «the same thing but pricier» — it is a different task. Custom makes sense when business logic cannot be expressed as a standard catalogue: complex price calculations, dealer portals with separate price lists, deep exchange with an accounting system, unusual ordering scenarios.
Here is the mistake we see regularly: a business with those requirements takes SaaS «to save money», spends six months paying for a higher plan just to get the API, piles on workarounds — and ends up commissioning development anyway, only now with the budget spent and the time lost. If your store contains something the standard platform does not have, it is more honest to price the custom route from the start.
Payments and delivery: where Ukrainian services work out of the box
This is the layer where «I'll just take something global» most often falls apart. A Ukrainian buyer expects familiar payment methods and Nova Poshta at checkout. If the platform cannot do that, you either pay for custom work or lose the order at the final step.
In Horoshop the payment services are enabled on every plan, including the cheapest Basic: LiqPay, WayForPay, Portmone, Fondy, Mono and PrivatBank, plus instalments and pay-in-parts. The commission is charged by the payment service, not the platform — so moving to a higher plan does not change your acquiring cost. Nova Poshta delivery is likewise built in.
WooCommerce has all of this too, but through plugins: every Ukrainian service is a separate module to install, configure and keep updated. More flexibility, more work.
In Shopify, Wix and Tilda, Ukrainian payment services either come through third-party apps or are unavailable altogether, and pricing is tied to foreign currency. This is exactly where the «convenient global platform» turns into ongoing hassle.
The list of Horoshop payment services was verified against the platform's official site as of August 2026.
How large a catalogue the platform can hold
This is the thing people ask about least before signing and remember most a year later. A CMS has no catalogue limit at all — the ceiling comes from your hosting and database. In SaaS the limit is baked into the plan, and it is usually the limit, not the feature list, that forces the upgrade.
| Platform / plan | Products | Orders per month |
|---|---|---|
| Horoshop Basic | 1,000 | 1,000 |
| Horoshop Standard | 20,000 | 1,000 |
| Horoshop Pro | 50,000 | 1,000 |
| Horoshop B2B | 50,000 | 1,000 |
| Horoshop Pro / B2B by separate agreement | up to 250,000 | up to 10,000 |
| WooCommerce, OpenCart, Magento | no limit | no limit |
The practical reading is simple. A catalogue under a thousand items runs on anything. Between one and twenty thousand, Ukrainian SaaS handles it comfortably. Above fifty thousand products it stops being a question of plan and becomes a question of negotiating with the platform or moving to your own CMS.
A separate word on accounting integration, because nobody fills a large catalogue by hand. Horoshop has no dedicated 1C or BAS connector, despite how often those are named in advice articles. What works out of the box is exchange with Dilovod, Torgovsoft, ERC, DNTrade and Drop Cloud. If your accounting runs on 1C or BAS, the exchange goes through the API — and the API is available only on the Pro and B2B plans. Basic and Standard do not have it at all.
Magento technically holds catalogues of any size but needs a team to maintain it, which for a mid-sized Ukrainian business is almost always more expensive than a top SaaS plan plus a developer on call.
Horoshop limits verified against the platform's official site as of August 2026.
Features or price: what actually separates the plans
The question is framed wrong, and the pricing table shows why. In SaaS the price is the feature set: you are not paying for more of the same, you are buying the right to use things a store cannot operate without once it grows. So the useful view is not what a plan includes, but what it lacks.
| Horoshop plan | What is missing |
|---|---|
| Basic — UAH 288/mo | API, Rozetka / Kasta / Google Merchant feeds, hiding the platform logo; not sold monthly — annual payment only |
| Standard — UAH 816/mo | API, volume-based wholesale pricing, hiding the platform logo |
| Pro — UAH 1,728/mo | price types and customer groups, dealer cabinet, B2B API |
| B2B — UAH 3,360/mo | no restrictions within the platform |
Two rows usually come as a surprise. First, the platform logo can only be removed from the Pro plan up — on Basic and Standard it stays on your site. Second, Basic is not sold monthly, so "trying it for UAH 288" is not an option; that is an annual commitment.
Design is its own story. The platform ships over 350 ready themes, which genuinely covers most retail niches. But an approved brand book or a custom layout does not fit into the box: in SaaS you pick from what exists rather than build your own.
The good news is that getting in costs less than people expect: migration from WooCommerce, OpenCart, Bitrix, Tilda or marketplaces is free with annual payment on Standard and above, otherwise UAH 0–4,200, with products uploaded in two to three days. So moving into the platform is cheap while moving out means rebuilding the store from scratch. Factor that asymmetry in before you decide, not after.
The rule we give clients: budget not for the cheapest plan but for the plan the store will actually be running on a year from now. If marketplaces, accounting sync or wholesale customers are anywhere in the plan, the entry tier is not a saving — it is a postponed migration.
What to choose for your case
| Your situation | What to take | Why |
|---|---|---|
| First store, standard catalogue | SaaS (Horoshop) | Fast launch without a developer |
| SEO and organic traffic are the bet | WordPress + WooCommerce | Full control of the technical base |
| Selling on marketplaces in parallel | SaaS, Standard and up | Rozetka and Kasta feeds out of the box |
| Exchange with an accounting system | CMS or custom | In SaaS the API is on higher plans only |
| Products not counted by the piece | OpenCart or custom | SaaS cannot calculate that way |
| Dealers, wholesale prices, portals | Custom or a B2B plan | The logic goes beyond retail |
| Ukrainian payments and Nova Poshta needed | Ukrainian SaaS | Western platforms lack this layer |
| Approved brand book and mockups | CMS or custom | In SaaS design comes from ready themes only |
| Selling without holding stock | SaaS to start, CMS later | Critical — supplier stock feed handling |
That last row deserves separate attention. If you work on a direct-supply model, the platform choice comes down not to design or product count but to how often supplier stock levels update and in what format the feed arrives. How the model works and when a rented storefront stops being enough is covered separately in what dropshipping is and how it works.
Four questions that settle the choice
- Is anything about the store non-standard? An unusual unit of sale, your own price calculation, a special ordering flow. If so, SaaS is out immediately, and no plan will fix it.
- Where is your accounting? If it lives in BAS or 1C, budget either a higher plan with API access plus developer time, or a CMS where the exchange is built without platform limits.
- Who will maintain the site? No developer and no contractor — take SaaS. If you have them, a CMS gives you more for the same money.
- What will leaving cost? Exiting SaaS means a full migration: products transfer, design and functionality are rebuilt. With your own CMS you take the whole site with you.
What launching costs
The platform subscription is not the whole sum. The store still has to be built, filled and launched. Our development pricing: landing page from $350 (5–7 business days), corporate site from $750 (2–3 weeks), catalogue or complex store from $1500 (3–5 weeks). Instalments are available from $59, $125 and $250 per month respectively.
On top of that, SaaS adds the platform's monthly fee and a CMS adds hosting and maintenance. How the final figure comes together is laid out in how much a website costs in 2026, and the full price list is on our web development page.
Horoshop pricing was verified against the platform's official site as of August 2026. Other platforms' prices change and are tied to foreign currency — check them with the vendors before deciding.
Our conclusion
There is no perfect platform, only a fit to the task. SaaS wins at the start and loses over distance; a CMS is the other way round. Custom pays off only when the business genuinely does not fit in a box — which is rarer than owners think and more common than platform vendors admit.
A practical suggestion: don't pick the platform first. Start by writing down how an order moves through your business from click to shipment. Half the decisions become obvious at that step, and the other half stop being a platform question at all.