WEBFLOW AGENCY CHECKS TO RUN PAGE BY PAGE BEFORE a WEBSITE GOES LIVE
Pre-launch checklists are usually written for the site as a whole: the redirects, the analytics, the forms, the sitemap. Those matter and they are well covered. What is almost never checked is each page against every other page, which is where a different class of problem lives. These are faults that cannot be seen on any single page, only across several, which is exactly why reviewing page by page in isolation misses them every time.
TL;DR
Ask the team on the build to run a cross-page pass as well as a per-page one.
Some faults only appear when two pages are compared.
Check that the same control is named the same way everywhere.
Check that repeated navigation sits in the same order on every template.
Do it on one page per template, not on every page.
The Faults That Hide Between Pages
A single page can be perfect and still be wrong. The contact button says Get in Touch here and Contact Us two templates over. The footer has five columns on the service pages and three on the blog. The breadcrumb is present on products and missing on resources. Each of those looks fine when reviewed alone, which is how they survive a thorough review and reach launch.
A page review cannot find a problem whose definition requires a second page. The cross-page pass is the one that catches them, it takes about an hour at twenty templates, and almost nobody schedules it. Most webflow agency checklists inherit the per-page structure from design review and never add the other axis.
Name the Same Thing the Same Way
The first cross-page check is naming. The consistent identification criterion is Level AA and requires that components with the same functionality within a set of web pages are identified consistently, according to the W3C's understanding document, and its examples include a document icon labelled differently on different pages. Build a short table: every repeated control down one side, every template across the top, and fill in what it actually says. The gaps and the disagreements show up immediately. It is dull work and it is the kind that a webflow design agency can hand to anybody, because it needs no design judgement at all, only attention.
The table is the deliverable, not the opinion that everything looked fine. A webflow agency that hands over the filled-in grid has given the client something checkable, which is a different thing from a reassurance in a status call.
Keep Repeated Things in the Same Order
The second check is order. The consistent navigation criterion is Level AA and requires that navigational mechanisms repeated on multiple pages occur in the same relative order each time, unless the user initiates the change, per the W3C's understanding document. Campaign and landing templates are where this breaks, because trimming the navigation is somebody's idea of reducing distraction.
Check the header, the footer and any in-page section navigation across every template, in order, not just for presence. A missing item is obvious; a reordered one is not, and it is the reordered one that makes visitors feel the site is unfamiliar without being able to say why.
One Page per Template, Not Every Page
The practical scope matters, or this never gets done. Faults of this kind live in templates rather than in pages, so checking one instance of each template finds everything that a full crawl would, in a fraction of the time. Pick the busiest page of each template, list the templates first so nothing is forgotten, and work through the two tables above. A site with forty pages usually has eight or nine templates. An hour spent on nine templates finds more than a day spent on forty pages.
The exception is any page built outside the template system, which should be checked individually because that is precisely where the system was not applied. Those pages are usually easy to name from memory: the ones built in a hurry, the ones inherited from a previous agency, and the landing pages nobody in the current team commissioned. Add them to the list explicitly rather than hoping the template pass covers them, because by definition it will not.
Three Ways to Schedule This
A cross-page pass in the pre-launch window: works for a new build and catches everything before anyone sees it, with the limit that it competes for time against content sign-off.
A quarterly pass on the live site: works to catch drift from new pages and costs an hour, though you risk problems living for three months before anybody notices.
In the definition of done for new templates: works to prevent recurrence and needs no separate scheduling, but the cost is that existing drift never gets found.
Final Verdict
List your templates, not your pages, and check one instance of each. That reframing is the single thing that makes this work affordable, and it is why most teams who try to do it page by page abandon it halfway.
Run the naming table and the order table, in that order, since naming problems are more common and faster to fix. Then put the same pass on a quarterly repeat so new templates do not reintroduce what you just corrected. Any webflow agency should be able to hand this over as a filled-in table rather than a verbal assurance, and a webflow design agency that treats it as unnecessary has not had to explain the inconsistencies to a client later.
FAQs
Q: Is this not what a design system is for?
A: A design system prevents the problem for anything built through it, and most of these faults come from the pages built around it: the campaign template, the page a different agency made, the one someone assembled under deadline. The check exists because the system is never applied universally in practice.
Q: Can this be automated?
A: Parts of it can. Extracting every button label across a set of pages and listing the distinct values is a script, and worth writing if the site is large. Judging whether two differently worded labels mean the same thing still needs a person, but the script narrows what the person has to look at considerably.
Q: When in the launch sequence should it happen?
A: After content is final and before the final client review, so that what you find can still be fixed without renegotiating a date. Running it earlier means re-running it; running it later means filing the findings as post-launch tickets, which is how they end up unfixed six months on.
0 comments
Log in to leave a comment.
Be the first to comment.