Our own site used to run on Tilda. It did its job until the jobs got bigger: languages on separate domains, an interactive checker, unit selection demos, structured data for AI crawlers. Every next item was either impossible on the builder or possible through a workaround that broke something else. So we rebuilt on Next.js, and because we also ship client sites on builders when that is the honest choice, this comparison comes without religion.
What builders genuinely do well
Speed to first version. A competent designer publishes a decent landing in days, edits it without developers and pays a subscription instead of a build budget. For a campaign page, a test of a new offer, a small company site, that is exactly right. Recommending custom code there would be selling, not advising.
Builders also keep the editing loop in the client's hands. If your content changes weekly and the site is simple, that loop matters more than any technical elegance.
Where the ceiling is
The ceiling is product logic. A property site that needs unit selection with live availability, a store that needs subscriptions, a site in three languages with correct addresses in each: builders either cannot do it or do it through embedded widgets that look bolted on, load slowly and keep your data in someone else's account.
The second ceiling is search, in both meanings. Builders generate heavy markup, give partial control over structured data and no control over the server. You can rank with that on low competition. In crowded niches the technical layer becomes the difference, and on a builder it is capped. The same applies to visibility in AI assistants: the site must be readable without JavaScript, carry schema.org markup and answer fast. On our stack this is standard equipment; on a builder you get some of it on a good day.
The third ceiling is quieter: ownership. Export from most builders is partial or cosmetic. After years of content you do not really own a site; you rent the ability to show it.
What code costs, honestly
More, upfront. A coded site needs a developer to build and a process to change, and for many businesses that overhead is simply not justified. We tell that to clients directly, and we wrote a whole piece on what different site tiers cost in property specifically.
The economics flip when the site is an asset rather than a page: when it must integrate, scale, rank, be quoted by AI and carry product features competitors cannot copy in a week. From that point every workaround on a builder costs leads, and the code that looked expensive starts looking cheap.
How we decide
One question does most of the work: what should this site be able to do in a year? If the answer is "the same, maybe new texts", a builder is the honest recommendation and we say so. If the answer includes selection, accounts, languages, integrations or serious organic goals, we build on Next.js from the start, because rebuilding later costs more than starting right.
Unsure which side of the line your project is on? Send us the plan and we will tell you, including the case where the builder wins.
