Product comparison method for an England web brief: evidence table, WCAG-EM tests and pass or fail gates

Which evidence should an England organisation examine when replacing a small information website? This method compares two named candidates, WordPress.com Business and a Webflow site plan, with the existing system as a documented baseline. Each gets the same table, tests and gates. No winner is declared before the tests are run.

What to take away

Treat the existing system as a documented baseline before assuming a replacement is better.
Supplier documentation explains product mechanics but does not prove compliance or performance in England.
Run the same synthetic content and authorised roles through both plans and the baseline.
Keep privacy, security, licensing, procurement, contract and exit as separate gates that cannot offset each other.
Hold the decision until exact proposals, qualified review and controlled tests exist.

One case and one evidence rule

The case is one public information site, run by a small editorial team, with recurring content revisions and accessibility needs for visitors and editors. Personal data is collected only if separately approved. E-commerce, membership and bespoke server code are excluded.

The named plan must be the exact edition and term. The two candidates here are a WordPress.com Business plan and a Webflow site plan, named as examples of a hosted builder and a site-plan builder; no claim about either is made beyond the evidence rows. Write down the product, edition, contracting entity and date of each record.

The baseline is not a product claim: document the present hosting, editor workflow, contract and known defects first. The tools and supplier guide shows how to record them.

Like-for-like evidence table

Fill the same table for every candidate and for the baseline. Leave a cell as unknown when the evidence is missing. An unknown stays open and never receives a midpoint score.

Field	WordPress.com Business plan	Webflow site plan	Existing-system baseline
Service boundary	Features stated in the supplier's plan record	Features stated in the supplier's plan record	Buyer records current components and owners
Accessibility evidence	Supplier statement, configured output untested	Supplier statement, configured output untested	Test representative visitor and editor tasks
Data terms	Contract and privacy review needed	Contract and privacy review needed	Inventory parties, purposes, transfers and contracts
Export and exit	Documented export boundaries	Documented export boundaries	Authorised backup, restore and handover test
Support and full cost	Unknown until the exact order exists	Unknown until the exact order exists	Buyer-owned evidence, not assumed to be zero

These records show what each supplier publishes, not that a configuration meets the brief. Supplier documentation explains product mechanics. It does not establish legal compliance or performance in England.

Accessibility tests with WCAG-EM

The W3C WCAG Evaluation Methodology is an approach for determining how well a digital product conforms to the Web Content Accessibility Guidelines. It sets out five steps: define the scope and conformance level, explore the product, select a representative sample, evaluate the sample and report the findings.

Applying it needs knowledge of WCAG, accessible design, assistive technologies and how people with different disabilities use digital products. Ask an independent specialist, not the supplier. The method is a supporting resource and does not define additional WCAG requirements.

Government accessibility requirements, which also satisfy the Public Sector Bodies (Websites and Mobile Applications) (No. 2) Accessibility Regulations 2018, include level AA of WCAG 2.2 as a minimum, work on the most commonly used assistive technologies and publish an accessibility statement, as the Service Manual accessibility introduction sets out. A private buyer can borrow the same three tests.

Control tests before a decision

Use the same synthetic content set and authorised account roles for every candidate. Record publishing, rollback, export, restoration, account recovery and support escalation outcomes in the same log.

Control of data is a core test. Make sure you have full control of any data you store, and reduce the chance of being locked in to long contracts for specific tools and providers. An export that cannot be restored is not an exit.

If the site will process personal data, you must determine your lawful basis before you start using it, and you must document it. Review data terms, roles and transfers with qualified advice. Read the commercial contracts records before signing anything.

Gates and verdict

Keep privacy and PECR, security, intellectual-property licensing, claims, procurement, contract, continuity and exit as separate gates. A strong result in one cannot offset a failure in another. The review methodology explains how a verdict is reached from gate results.

Gate	Pass if	Fail if	Named reference
Privacy and PECR	A lawful basis is documented before personal data is used, and cookie and marketing rules are mapped	Personal data is collected with no recorded basis	ICO lawful basis guide, ICO PECR guide
Data terms	Processor terms cover the Article 28(3) details	The supplier handles personal data with no processor terms	ICO guidance on what the contract must include
Accessibility	A WCAG-EM sample passes level AA and editors can work with assistive technology	Key tasks fail or no independent test exists	W3C WCAG-EM, WCAG 2.2, Equality Act 2010 sections 20 and 29
Security and supply chain	Sub-suppliers with data access are listed and meet your standard	Sub-suppliers with data access are unknown	NCSC cloud security principle 8
Export and exit	An authorised export restores into a test site	The export cannot be restored	GOV.UK Service Manual on control of data
Contract and procurement	Term, renewal and cancellation are written and approved	Terms allow lock-in you cannot accept	GOV.UK Service Manual on lock-in

For a private buyer, the Equality Act 2010 applies to service providers. Section 29 bars a service provider from discriminating against a person requiring the service, and section 20 sets the duty to make reasonable adjustments. The 2018 accessibility regulations apply to public sector bodies, so a private buyer borrows their tests, not their duty.

Until exact proposals, qualified review and controlled tests exist, the verdict is hold. Run the selection checklist at each gate. Repeat the comparison with fresh plan records before any 2027 decision, because supplier plans change.

For a worked contrast between a hosted builder and an open platform, read the guide to WordPress and website builders.

Before you act

Confirm the exact edition and term of each plan before comparing anything.
Document current hosting, editor workflow, contract and known defects.
Ask an independent accessibility specialist to test representative tasks.
Record publishing, rollback, export, restoration and support escalation outcomes.
Review data terms, roles, transfers and contracts with qualified advice.
Rerun the comparison with fresh plan records before any 2027 decision.

Common questions

Why does the comparison refuse to declare a winner?

Without exact proposals and controlled tests, any winner would rest on supplier marketing. The table records what each supplier publishes and leaves unknowns open until evidence exists.

What does the like-for-like table show?

It shows what each supplier publishes, not that a configuration meets the brief. Support, continuity and full cost remain unknown until the exact order and support evidence are obtained.

What must happen before replacing the existing site?

Run the same synthetic content and authorised roles through every candidate and the baseline, have an independent specialist examine representative tasks, and close each gate separately.