What Does a B2B Website Cost? Web Design Prices for the Mid-Market in 2026, from Template to Custom Build

The short answer first. For a mid-market B2B company in the DACH region, a website costs up to about 2,500 euros as a template, 6,000 to 20,000 euros as an agency-built WordPress site, and 20,000 to 60,000 euros as a custom build with its own design, two languages and system connections. On top of that comes 10 to 20 percent of the initial investment per year for operation. Those bands come from our own projects, and they are wide for a reason: the word website covers a brochure and a sales system equally well. The rest of this article is about which band your case actually sits in, what moves the number inside it, and which line items are quietly missing from most proposals.

Why the question has changed

Ten years ago a B2B website was a credential. Prospects found you through a trade fair, a recommendation or a phone call, then looked you up to check you existed. The site had to be presentable. It did not have to work very hard.

That order has reversed. In a Gartner survey published on 9 March 2026 (n = 646 buyers who had completed a B2B purchase in the previous twelve months), 67 percent said they prefer a rep-free buying experience, up from 61 percent in the equivalent 2025 survey. Gartner also finds that buyers spend only about 17 percent of their purchase journey meeting with potential suppliers at all, and when that share is split across three or four vendors, any single provider gets a sliver of direct contact. The same 2026 survey found 45 percent of buyers used AI tools during a recent purchase.

Read those three numbers together and the budget question looks different. Most of the evaluation now happens where you are not present, partly through systems that read your site rather than look at it. That does not make design irrelevant. It does mean a website is no longer a cost centre in the marketing budget but part of the sales infrastructure, which is the frame the rest of these numbers belong in.

The three cost bands

Almost every offer on the German market falls into one of three approaches. They are not better and worse versions of each other; they solve different problems, and the expensive mistake is buying one while needing another.

Two remarks on reading this table honestly. The lower end of the WordPress band buys a good theme, careful configuration and your content in it, which for many companies is genuinely enough. The upper end of the custom band is not a nicer version of the same thing; it includes interfaces, multilingual structures and functionality that a plugin can only approximate. Where exactly the line between the two runs, and why platform detours eventually cost more than they save, is the subject of WordPress vs. custom development.

The six drivers that move the number

Inside each band the spread is large, and it is driven by six things. If you know your position on all six, you can estimate a proposal before you receive it.

Of the six, system connections and content are the ones that turn a fixed price into a moving one. The technical foundation is the one companies regret most often, because retrofitting server-side rendering into a site built as a client-side app is not an adjustment, it is a rebuild of the delivery layer.

The line items missing from most proposals

The initial price is usually the honest part of a quote. What is missing is the part that arrives later, and it is reliably the same four items.

• Operation and security. Hosting, domains, certificates, backups and, above all, updates. On a plugin-heavy WordPress installation this is not optional maintenance; it is the difference between a running site and an incident.
• Content upkeep. A site that never changes loses relevance in search and credibility with buyers. Someone has to own this, internally or externally, and that person costs money either way.
• Measurement. Analytics, consent management, conversion tracking and someone who actually reads the numbers. Without it you cannot tell whether the investment paid off, which makes the next budget decision guesswork.
• Small extensions. A new service page, a landing page for a trade fair, a form change. Individually trivial, collectively a recurring budget line, and much cheaper when someone already knows the codebase.

Add these up and you land at the 10 to 20 percent per year quoted above. Companies that skip the line entirely do not save it. They pay it in year three as a rebuild, which is the most expensive form of maintenance there is.

How to recognise a solid proposal

Comparing proposals on price alone is only possible when they describe the same thing, and they rarely do. Four checks make them comparable, and all four are things you can verify without technical knowledge.

Ask for measurable acceptance criteria. Adjectives like fast and modern are not criteria. Google's Core Web Vitals are: LCP under 2.5 seconds, INP under 200 milliseconds, CLS under 0.1, each measured at the 75th percentile of real page loads on mobile and desktop. A provider willing to write those into the contract is making a statement about their build quality.

Ask how the site is readable without JavaScript. Search engine crawlers and the AI systems that increasingly mediate buying decisions do not all execute scripts reliably. Server-side rendering or pre-rendering solves this, and it must be decided before the build, not after. We treat that as a baseline rather than an upgrade, which is also why our own pages ship as static HTML per route.

Ask for the redirect plan if this is a relaunch. Every existing URL needs a target. This is the single most common cause of a site that looks better and performs worse after going live, and it is entirely preventable.

Ask who owns what at the end: code, domain, hosting account, analytics property, image licences. Dependency is a price you pay later, and it does not appear anywhere in the quote.

Does it pay off? The calculation that matters

Conversion-rate rhetoric is not much use in the mid-market, because B2B volumes are small and cycles are long. The usable calculation runs backwards from order value.

Take a 30,000 euro investment, an average order value of 15,000 euros and a close rate of 25 percent on qualified enquiries. Two additional closed deals cover the investment in revenue terms, which is eight additional qualified enquiries per year. Calculate on contribution margin instead of revenue and the honest figure is roughly four to six deals, so sixteen to twenty-four qualified enquiries a year: one to two per month. That is the number to hold a proposal against. If a website cannot plausibly produce one to two additional serious enquiries per month for your business, either the investment is too large or the problem is not the website.

The same logic applies to the running cost. Two thousand euros a year is not an abstract overhead; it is roughly one qualified enquiry. Whether the site produces that is measurable, which is the whole point of the measurement line item above.

Conclusion

Website costs are not opaque, they are unspecified. Once you name the number of layouts, the languages, the interfaces, who writes the content and what the acceptance criteria are, the price band follows almost automatically, and proposals become comparable. For most mid-market B2B companies the honest answer is one of two: a well-built WordPress site in the six to twenty thousand range if the website supports sales, or a custom build from twenty thousand upwards if the website is sales. The decisive question is not which is cheaper, but which one you will still be able to work with in three years. If you want that decision made against your actual situation rather than a price list, that is what a first conversation is for, and how we work is described under services.

Frequently asked questions about B2B web design cost