The cost of SaaS tools around a company website in Poland

All 355 observations come from vendors' own pricing pages. Zero data under NDA, zero data from clients, zero prices from comparison sites — a stricter rule than the May edition, which allowed expert blogs and ranking profiles.

Each observation carries: vendor, plan, price exactly as published, currency, billing period, pricing metric, any renewal-jump information, and the date it was checked.

Three departures from previous editions, deliberate and named:

• Vendors are not anonymised. There, the observation was a contractor's quote for comparable work; here, the observation is a published price list, so the vendor's identity isn't metadata — it's data. "A certain platform charges CHF 140 plus CHF 0.80 per order" is useless.
• No medians or quartiles. The sample is price lists from seventeen categories, not quotes for the same thing. A median would blend CHF 1.94 for a mailbox with CHF 7,150 for an enterprise plan. Instead we give per-category ranges and break-even thresholds.
• The unit of analysis is the metric, not the price.

The exchange rates used are unchanged from the May edition and reflect the Polish-zloty rates behind the underlying study: 1 USD ≈ 4.00 zloty, 1 EUR ≈ 4.30 zloty. Prices are given in the vendor's own currency; conversions are made only where a scenario total is summed.

This report is the fifth in the collection. The previous four cover build costs: website costs, online store costs, web app costs and mobile app costs. This one covers what happens after launch.

Distribution of the 355 observations by what the charge is based on:

135 of the 355 observations have no price without a volume figure. That is the number the whole report is built around.

A fifth axis has emerged over the past year: AI credits. Canva gates plans behind an AI limit, Webflow sells credits at USD 20 per 2,000, IdoSell bills AI tokens and flags them "New!", and at Adobe Express the entire gap between the EUR 12.29 plan and the EUR 22.73 plan is, in practice, credits. Any cost article written before 2026 is structurally incomplete, not just out of date.

Wix. On the Polish storefront, Light costs CHF 26.27/month, and Core PLN 4,690/month. Core is cheaper and comes with 50 GB instead of 2 GB, five collaborators instead of two, and payments and eCommerce that Light doesn't have at all. There is no dimension on which Light wins. The reason: Core, Business and Business Elite carry a "50% discount", Light doesn't — the cheapest plan is the only one paying full list price.

home.pl. Hosting Biznes costs PLN 1,990 in the first year against PLN 3,990 for Hosting Start — and it renews more expensively: CHF 349.50 versus CHF 339.50. A buyer who chose based on the visible number pays more for the following decade.

Adobe Express. The business plan Teams is EUR 6.72 net per licence; the Premium plan for a single person is EUR 12.29 gross. Teams is cheaper per head and offers more: Admin Console, template locking, 1 TB instead of 100 GB. It requires two seats — so two colleagues each pay roughly half of what one of them would pay alone. Both prices sit on the same page, in different tax bases, with no warning.

Three independent companies, three categories, one shape. This stops being an anecdote.

The same store — 200 orders a month, a PLN 200 basket, PLN 480,000 in annual turnover. Everything except the platform held constant: domain, three mailboxes, email marketing, live chat, design, consent management.

The first column is the table every comparison article publishes. The second adds what the store actually uses to take payments: Przelewy24 at this turnover costs PLN 6,912 for everyone — plus the 2% that Shopify Basic charges when payments run through anyone other than Shopify Payments, i.e. PLN 9,600 a year.

The threshold is calculable: PLN 97,050 in annual turnover, about PLN 8,088 a month. Below that, Shopify is genuinely the cheapest. Above it, it stops being the cheapest and the gap widens — at PLN 3,000,000 in turnover it comes to CHF 53,556 versus PLN 49,052 at Shoper.

The second finding from the same table: Shoper beats IdoSell by PLN 3,666 a year, and the whole difference is the metric. Shoper charges CHF 1,625 a year and stops there; IdoSell charges PLN 3,348 plus CHF 0.80 per store order. There is no crossover point in IdoSell's favour — the subscription itself is already higher, so every order widens the gap.

This is a comparison of price, not capability. IdoSell bundles IdoPay, POS, couriers and returns into the subscription; Shoper Standard tops out at 50,000 products and 10 GB. A store that needs what IdoSell packs into its subscription isn't overpaying — it's buying something else.

Thresholds taken directly from the sliders on vendors' own sites:

Three independent companies in three categories advertise the number from the bottom rung of the ladder, and none of them say where the buyer actually stands on it.

It's sharpest with GetResponse: at 1,000 contacts, Starter is 4.05× cheaper than Marketer. At 10,000 contacts — 9.5% cheaper. The entire advantage of the entry plan evaporates, for the same feature set.

Zapier's "from USD 19.99" is a threshold of 100 tasks a month — three runs a day. Any real automation starts in the thousands — which means USD 129, six times the headline price.

The same mechanism can hide inside a single price list. Tpay advertises "from 0.99%", but that rate requires a service package at CHF 49.50 a month; the plan without the package costs 1.59% + PLN 40. The thresholds cross at PLN 149,434 in annual turnover — below that, the "pricier" plan is actually cheaper.

A price without a scenario is not a cost. Three named situations, all calculated from the dataset's rows, net amounts per year.

On the one-page site, 77% of the bill is human time, not software. Four tool line items cost CHF 349 a year together; site upkeep costs PLN 2,400. Arguing over whether hosting costs PLN 200 or CHF 340 moves the total by less than 15%. Removing the upkeep line halves it.

The two cheapest genuine decisions in the whole model don't touch the platform at all:

• Email. Zoho Mail Lite instead of Google Workspace brings S1 down to CHF 1,435, and S2 to PLN 6,380. Across three mailboxes, this one decision alone is worth CHF 498 a year — more than the entire hosting line.
• Live chat. Smartsupp Solo instead of Tidio is PLN 456 a year with a single operator.

The scenarios are averaged by design. To work out your own case, we have two tools: a website maintenance cost calculator and an e-commerce TCO calculator.

• KSeF has been mandatory since 1 April 2026 and costs PLN 0. inFakt KSeF and the Fakturownia integration cover the obligation for free. The free plan throttles rather than blocks: KSeF data is pulled once a week instead of every 15 minutes.
• CHF 140 net a month is the anchor of the Polish market. IdoSell Business, base.com Business and Sky-Shop Business — all three charge exactly that. And all three diverge on what they add on top: CHF 0.80 / CHF 0.50 / CHF 0.15 per order. They converge on the number the buyer compares, and diverge on the one they actually pay.
• Domain email is spread 8.1× wide — from CHF 1.94 (Zoho Mail Lite) to PLN 3,150 (Google Workspace Starter) per mailbox a month. The default choice is the expensive one: Google Starter costs 21% more than Microsoft 365 Business Basic, which gives you more.
• Only one registrar prints the renewal price next to the promotion. OVH states "Renewal CHF 29.50 net" for a .pl extension at CHF 8.35. home.pl, nazwa.pl and cyber_Folks advertise "from PLN 0" and don't list a renewal price anywhere on the page.
• Consent management is the cheapest recurring line item in the entire study — Usercentrics from EUR 2 a month. That most Polish websites run without a proper CMP is a decision about attention, not money.
• The vendor list overstates the number of independent vendors. Cookiebot and Usercentrics are one company after their 2021 merger. FreshMail is now MailerLite. BaseLinker is base.com. Hotjar is Contentsquare, with no published price. inFakt belongs to the Visma Group. Three vendors changed identity in eighteen months — a measurable rate of consolidation for any article that lists vendors without a checked-on date.
• "Local is cheaper" is true in one category and false in another. Reviews: TrustMate PLN 32/month versus Trustpilot USD 99/month per domain. CRM: Livespace CHF 39.50 and Firmao CHF 34.50 per seat versus Pipedrive from EUR 14 — the local option is pricier. That sentence can only be made category by category.

Transparent positioning section. Digital Vantage's own figures are not included in any tally in this report.

Digital Vantage is a premium Polish agency; stack Next.js + Payload CMS + MongoDB, rate from PLN 220/h (Poland) and EUR 75–95/h (EU). Base prices as published in the website cost report: PLN 3,500 for a one-pager, PLN 9,000 for a company site, PLN 15,000 for an extended one, PLN 25,000 for a corporate site, from PLN 60,000 for a portal.

Previous reports positioned the price of execution. This one covers recurring costs, so the right benchmark is upkeep — and our published hourly rate does something no subscription table can:

The entire Polish website-upkeep market fits between one and four hours a month. That's the same information as "PLN 200–900", but in a unit the buyer can actually use to judge whether they're getting what they need. A WordPress update with regression testing doesn't fit into 0.9 hours.

This closes the argument about free open source: WooCommerce with no licence fee costs PLN 6,000 a year in upkeep alone in the store scenario — that is, 27 hours of someone's work. The licence is free; the hours aren't, and the hours are the cost.

We don't sell an upkeep package off a price list, and we have no single number here to compare — we price by the hour. The consequence cuts both ways: a client who needs less than an hour a month pays us less than a PLN 200 package they wouldn't use up anyway; a client who needs four hours pays PLN 880, the same as the top market package, at a rate from the lower end of the premium segment.

• This is an inventory of price lists, not a market study. We don't know how many companies actually pay which rate. We give store counts where the vendor publishes them (Shoper 40,000, IdoSell 8,000) and say plainly that no single official study of market share exists for Polish platforms.
• We do not compare capability, only price. That caveat sits next to every table, not in a footnote.
• Exchange rates are an assumption, not data. They change every total expressed in EUR and USD.
• Four thresholds remained unread: Shutterstock's by-download-count pricing (the site blocks automated reading), and GetResponse's intermediate-threshold labels.
• Autopay's rates were not confirmed — the vendor's documentation contains not a single figure; it points to a price list it does not publish.
• Joomla holds 3.99% of the Polish CMS market and one row in the dataset. That is the one deliberate gap relative to actual market share.

The dataset accompanying this report is published under a CC BY 4.0 licence, in JSON format with Schema.org Dataset metadata, at the same address as the report.

Attribution format: "Source: Digital Vantage, The cost of SaaS tools around a company website in Poland, 2026 edition, www.digitalvantage.pl".

This edition is frozen — the content at /edycja-2026 will never change. Future editions will get their own addresses.

Unlike previous editions, the dataset is not anonymised: every row carries the vendor's name, plan, price exactly as published, currency, billing period, pricing metric and the date it was checked. The reason is explained in usage_notes and in the methodology section.