How We Score: Inside the 40-Point Rubric

How we score website builders is a single sentence: six criteria, 40 points, headline score equals the total divided by four, rounded to one decimal with half rounding up. Everything else on this site is downstream of that sentence. A ranking is only as trustworthy as the weighting behind it, so this page defends each weight rather than presenting the rubric as self-evident.

Weighting is where every comparison in this category smuggles in its opinion. A rubric that awarded ecommerce 12 points would put Shopify first, and a rubric that awarded price 12 points would put Hostinger first, without either result telling a reader anything except what the author had already decided. The weights below are fixed in advance, applied identically to all six builders, and published so a disagreement can be located precisely rather than argued in general terms.

Why ease and design carry 8 points each

These two are the joint-heaviest because they are the only criteria that affect every buyer, every day, for the entire life of the site. Ease of use is weighted at 8 for a reason that is easy to underestimate: the maintenance burden outlives the build. A tool that takes weeks to learn imposes that cost again on every new person who touches the site, and the difference between a founder editing a page in four minutes and a founder abandoning the attempt is not a matter of polish.

Design carries 8 because it measures two things at once and both are permanent. The first is the quality of what a builder hands you before you touch anything. The second is how much control you keep afterwards, which matters more than template count, because on two of these platforms the template cannot be changed after publishing. A design decision made in the first hour of a Wix or Hostinger build is effectively a decision about the site's whole life, which is precisely the sort of consequence a heavy weight is for.

Why SEO and value carry 7

SEO and performance sits one point below the top pair rather than at the top, because a meaningful minority of sites do not depend on organic search at all. A restaurant with a strong local reputation, an agency that sells entirely on referral, an internal portal: none of them needs a canonical tag. For everyone else this is the criterion that decides whether the site works, and it is the only criterion in the rubric supported by verified third-party field measurement, which is why it also serves as the tiebreak.

Value for money also carries 7, and its definition does the work: true cost across two years, renewal year included. Weighted this way, an introductory rate earns nothing on its own. Two of the six raise plan prices structurally at renewal, so scoring value on year one alone would produce a ranking that is wrong for every buyer who keeps their site longer than twelve months, which is very nearly all of them.

Why ecommerce and support carry only 5

Ecommerce and extensibility is capped at 5 because most people buying a website builder are not selling products. Weighting selling capability as heavily as ease of use would hand the ranking to Shopify on behalf of a majority of readers who will never open a store. The rubric handles this by scoring the criterion honestly at a low weight and stating clearly that where selling is the purpose of the site, this criterion should override the headline total outright. A 5-point criterion that governs the entire project is still the deciding factor for that project.

Support and lock-in is capped at 5 for a different reason. Support is a fallback that most users touch a handful of times a year, and lock-in is a tail risk: migrations are rare, and pricing a rare event as heavily as a daily constraint would distort every score. The word tail is doing real work there, though. Rare and catastrophic is not the same as unimportant, which is why lock-in is scored at all rather than mentioned in passing, and why Webflow's code export earns it a point that four rivals cannot match.

A worked example you can check

Wix is the highest-scoring builder in the current revision. Its card, criterion by criterion, with the reason each figure was awarded:

• Ease of use, 7 of 8. Genuinely capable drag-and-drop with AI onboarding, marked down because the dashboard sprawls and free-positioning can wreck the mobile view.
• Design and templates, 7 of 8. An enormous library with real variety, marked down for uneven quality and the inability to switch template after publishing.
• SEO and performance, 6 of 7. The deepest hosted SEO toolkit in the set, plus second place on Core Web Vitals at 74.86% in the November 2025 report. The seventh point is withheld on depth of schema and head-level control.
• Value for money, 5 of 7. Mid-market pricing from $17 to $159 with no renewal cliff, marked down because mailboxes and applications are all charged separately.
• Ecommerce and extensibility, 4 of 5. A competent store and a 500-plus app market. Not Shopify, but nothing else in this set is either.
• Support and lock-in, 3 of 5. Round-the-clock coverage without a direct inbound phone line, operating a callback model, and zero site export.

The arithmetic: 7 + 7 + 6 + 5 + 4 + 3 = 32. Then 32 divided by 4 equals 8.0. That is the whole calculation, and there is no adjustment, no editorial bonus and no weighting applied after the fact. Every card on this site can be added up the same way in under a minute, and any card that does not reconcile is an error we want reported.

Each revision is also cross-footed before publication. The six row totals and the six column totals must reach the same figure, and no individual cell may exceed its criterion's weight. In the current revision both directions sum to 170 across the six builders, which catches transcription errors that reading alone would miss.

The tiebreak rule, and why SEO breaks ties

Where two builders finish on the same total, the higher SEO and performance score ranks first. The rule is published in advance so it cannot be applied selectively after a result is known. It decides fourth and fifth place in the current revision: Shopify and Hostinger both total 28 of 40, and Shopify ranks above Hostinger on 5 of 7 against 4 of 7.

SEO holds the tiebreak for two reasons. It is the most consequential of the non-cosmetic criteria, since a site that cannot be found frequently fails outright rather than performing slightly worse. And it is the only criterion resting partly on external field measurement rather than capability inspection alone, so a tie broken on SEO is broken on the best-evidenced axis available.

What the rubric deliberately ignores

Popularity is excluded. Market share measures adoption, not quality, and the two come apart visibly: the platform that led the November 2025 Core Web Vitals table at 84.87% accounts for 0.7% of all websites in the W3Techs survey of 21 August 2026. Unshipped feature announcements are excluded, because a roadmap is not a capability. Uptime service levels are excluded, since none of the published figures for these six could be verified at source, and an unverifiable number is worse than a missing one. Commission rates are excluded, which is set out in full in the affiliate disclosure.

Two exclusions are about method rather than relevance. The rubric does not publish hands-on load-time or Lighthouse measurements of builder output, because a handful of test sites is not a sample and dressing it as one would be worse than citing nothing. And growth statistics sourced from secondary aggregators with unpublished methodology are excluded, however quotable they are. What remains is documented capability, verified pricing, and published third-party measurement.

How scores move between revisions

The standard is revised monthly, and every page carries the revision it was scored under, currently 2026.8. A score moves only on one of three triggers: a documented capability changes, a verified price changes, or new third-party measurement is published. Preference is not a trigger, and neither is a redesign of a vendor's marketing site.

Two recent movements illustrate the mechanism. Webflow's native canonical tags and noindex shipped in June 2025, removing a long-standing custom-code workaround, and that capability change is what supports its current 7 of 7 rather than a lower figure. Webflow's pricing restructure on 13 May 2026 merged the old CMS and Business site plans into a single Premium tier, which changed the inputs to its value score without changing the criterion. When a figure moves, the criterion that moved it is named on the scorecard.

The formal statement of scope, criteria and revision procedure is published in the methodology, and the rubric applied to all six products, with every card added up in full, is published in the ranking of the best website builders for 2026. Corrections to any figure are welcome and take priority over new work.