Methodology
The Four Questions We Ask Every Builder
Instead of listing abstract criteria, it's more useful to think of our scoring as four direct questions we answer for every platform we test. The answers get weighted and combined into the single number you see on the rating page.
"How good are the templates, really?" — 35% of the score
Not how many templates exist — how many are genuinely distinct and well-designed, and how much real restructuring is possible before hitting a rigid section or a paywall. A builder with hundreds of dated, barely-different templates loses to one with a smaller but flexible, well-designed library.
"How fast can a stranger get to a published site?" — 30% of the score
We time the actual journey from account creation to a live, presentable site, paying attention to how forgiving the editor is of mistakes and how steep the learning curve feels for someone with zero design background. An editor that requires a tutorial video before anything makes sense loses ground here regardless of what it can theoretically do.
"Can you actually control your SEO, or is it locked away?" — 20% of the score
We check real, editable control over meta titles and descriptions, clean URL structures, redirects, and page speed out of the box. Platforms that gate basic SEO fields behind their most expensive tier score worse here no matter how good their templates look.
"What are you really paying, and what do you get when something breaks?" — 15% of the score
This one weighs price transparency across plan lengths against support quality — whether a genuinely useful free tier exists, and whether contacting support with a real problem gets you a real answer or a scripted non-response.
Why the First Two Questions Carry the Most Weight
Templates and ease of use together make up nearly two-thirds of the formula on purpose — they determine whether building your site is actually a good experience, which is the entire point of using a builder instead of hand-coding a site. Price and feature count can't compensate for a platform that's genuinely unpleasant to use.
What Doesn't Get a Number But Still Matters
Some things show up in the written review without moving the score directly: how deep eCommerce tools actually go past a basic cart button, how the mobile app compares to the desktop editor, and how a platform handles the basics of security like SSL and backups. We think these deserve mention even when they don't fit cleanly into the four questions above.
When We Revisit a Score
Not on a calendar — we re-test when something material changes. A major editor redesign, a pricing overhaul, or a meaningful shift in template quality triggers a fresh look. Every review shows its last-updated date so you're not guessing how current a number actually is.
What This Approach Doesn't Solve
No four-question formula can tell you whether a specific template will click with your particular taste, or whether an editor that feels intuitive to us will feel the same to you. Treat the score as a serious starting filter, then read the full review for anything you're seriously considering — the details a number can't hold are usually there.
Why We Don't Test Every Feature Equally
A builder can ship a long list of secondary features — AI content tools, built-in analytics, marketing automation — that never factor directly into our four questions. We mention genuinely useful ones in the written review, but we deliberately don't let a long feature list inflate a score that templates, ease of use, SEO, and price & support don't independently support. A feature nobody actually uses well isn't worth the same as one that changes the building experience.
How We Handle Disagreements Within Our Own Team
Testers occasionally land on different first impressions of the same builder, particularly around subjective calls like template visual quality. We resolve this by having a second person independently test any product where the initial score sits near a scoring-band boundary, rather than letting one person's judgment alone decide a result that close to the edge.
Where the Weighting Numbers Came From
The 35/30/20/15 split isn't arbitrary — it reflects which criteria most consistently separated builders people ended up happy with from ones they abandoned within a few months, based on patterns we noticed testing this category over time. We'd adjust it if we found a better-supported weighting, but we haven't yet, and we'd say so publicly if we did.