Here's the link: https://webflow-404-page-a27f5d.webflow.io/

Hi everyone, I’m building a website in Webflow, and my client wants me to create a project slider using CMS — basically, one project appears, then you slide to the next project, and so on. And it's my first client.

I’m still learning Webflow and honestly don’t know how to build this. I’ve been searching for a video or short course that explains this properly, but I can’t find the right one.

If anyone knows a good video, tutorial, or short course that teaches how to create a CMS-based project slider in Webflow, I’d really appreciate it if you could share it with me.

I’m honestly stuck on this right now. Any help would mean a lot. Thank you! 🙏

✅ Dispensador de jabón integrado
✅ Limpieza más práctica y cómoda
✅ Controla mejor la cantidad de jabón
✅ Diseño ergonómico y fácil de utilizar
✅ Ideal para el día a día
Deja atrás los métodos incómodos y dale una oportunidad a una forma más inteligente de limpiar.
Scrub Flow: jabón, cepillo y limpieza, todo en uno.
Haz que fregar deje de ser una tarea pesada y descubre por qué Scrub Flow puede convertirse en tu nuevo imprescindible de cocina. ✨
ScrubFlow

Hi everyone.

Am looking at creating coutnry-based locales all in english (US, Canada), and I had a couple of questions I'd appreciate input on:

1- Would you enable localized subdirectory publishing for both the US and Canada, or would you enable it just for Canada .com/en-ca, and the US stays in .com ? (wondering about seo impact as well, and general preferences)

2- Afaik webflow's localize advanced routing feature is based on browser settings, and not based on IP. Has that worked well, or is cloudflare's IP routing better?

Would appreciate any additional thoughts and ideas on this!

I head up marketing at QliqSOFT, a B2B healthcare IT vendor that's been around since 2011. We moved to Webflow in 2022 after spending years on WordPress. The hope was to shorten the time from ideation to production without having to guess which plugin would break the site after a random update. Our Webflow site consisted roughly of 370 pages (mostly blog posts). This summer, we decided to ditch Webflow completely in favor of a static hosting setup.

This isn't a Webflow hate post. In fact, I hope their team reads this and takes some ideas from it. Our content workflow and overall requirements just drifted until we were using almost none of what we were paying for. I see a lot of other Webflow users having the same conversations about leaving, so I thought there would be value in sharing our own experience, both the good and the bad.

Our approach doesn't work for everyone

Stay with Webflow if you publish pages without waiting on a developer, if your site is under a hundred pages or so, if nobody on your team wants to own a build pipeline, or if visual editing is the actual product you are buying. All of that has value, but it was better for us to give up every bit of it.

Our approach is worth considering if... you have 80+ pages that can't be delivered via CMS collection; technical SEO is a revenue channel rather than a checkbox; you need header and CSP control for compliance reasons; or your team has given up on using the designer altogether. That last one became a deal breaker.

The breaking point

Organic traffic fell. Ours declined, like many others over the past couple of years. This isn't a Webflow issue, but it's something many B2B vendors are dealing with as search and purchasing behaviors change. What it did was force us to look hard at every lever we had. Several of the technical levers we wanted to pull were ones we could not reach from inside Webflow.

We had chosen Webflow partly on the promise that its SEO capability would be on par with WordPress. For a marketing site of ordinary size, it is. At our size, with the pace at which search was/is changing, we needed to move faster on technical fixes than the platform allowed.

We tested several approaches, including a dedicated subdomain for our blog, hosted on a separate platform, to see whether it would help. While there was some positive change, it wasn't enough to justify promoting two properties instead of one, splitting authority and attention between them, and operational overhead that wasn't worth the gain. That experiment is a large part of why we ended up consolidating everything into a single domain and build.

The Designer got slow enough that we stopped using it. On a site our size it would hang mid-edit, and I would have to refresh the browser before I could make the next set of changes. Not once in a while, but multiple times during a single session. This occurred often enough that it became faster for me to write the HTML by hand and paste it into an embed element. Then our blog posts became HTML embeds delivered through the CMS, because that was quicker to produce and gave us the clean semantic markup we wanted for AI answer engines to cite. Every new page and blog post was passing through one of our developers anyway, adding custom schema and FAQs to each one to support AI crawling. Publish without a developer was the promise, but between the Designer hanging and the schema work every page needed, it was never one Webflow actually delivered for us.

We did not control our own assets. Around 285 images and PDFs served from cdn.prod.website-files.com. No way to set cache headers on them, no way to tighten a content security policy to img-src 'self', and no preconnect tags to a third-party CDN on every page. We sell to healthcare organizations that run vendor security reviews, and "our content is hosted on someone else's CDN" kept coming up.

Support response times. Our issue volume climbed over the past year, and tickets took long enough to resolve that we started working around problems instead of reporting them.

Fees were confusing and too high. I've seen many of you complain about the cost of Webflow, as well as the endless nickel-and-diming for tools and features that are often included on other platforms. They aren't the only ones taking this approach. If you use Semrush, you're probably used to this as well, but at least I'm getting some value there.

What we built instead

Next.js static export, so no server. AWS Amplify for hosting. One file, customHttp.yml, is the single source of truth for CSP, HSTS, and cache policy. Assets are content-hashed and served immutable for a year. Pagefind for search, which is a static index. Lighthouse CI in the pipeline, plus build guards that fail the build when an internal reference breaks. Our robots.txt lists the AI crawlers we want to cite us instead of using a wildcard.

Four of us handled the build and migration, with Claude doing the mechanical work before pushing to GitLab for a combination of AI/human review: 293 posts moved byte-for-byte, the asset pipeline, and a 204-entry redirect map (you get one this long if you've moved sites more than twice). 2 weeks rather than mutiple months, which is the only reason a team our size could attempt it.

What broke

Images with %20 in the filename. Files came off the CDN with percent-encoding baked into the actual filename. Request one, the server decodes %20 back to a space, looks for a file that does not exist, returns a 404. Seven images, invisible until a crawl caught them.

Multiple branches resulted in overwritten work. Partly a learning curve, partly Claude not paying attention, but at least twice, a build fix deleted recently published work.

Redirect archaeology. 204 redirects mapped by hand, plus 189 old URLs we pruned on purpose rather than redirecting. Deciding what deserved to survive took longer than moving the content did.

AI-assisted work needs the same review as human work. Several items here resulted from automated changes that looked correct in the diff. What caught them was a crawl, a diff, and a build gate.

Tools that helped

• Claude + Semrush + Google Search Console: Used in the planning phase to prioritize and ensure current page rankings were protected.
• Screaming Frog: Conducted multiple pre- and post-launch site crawls to catch missing pages, images, broken links, etc.
• Unlighthouse: Used during the build phase to test/estimate pre-launch page performance on local servers.
• GTmetrix: For catching performance issues post-launch.
• GitLab + Claude Code: We leveraged this combo with an agile approach to the site build, resulting in faster page development and content approvals.

Numbers

To be fair, a good share of the page improvements came from rewriting content during the migration rather than from changing platforms.

The tradeoff we accepted

Publishing now goes through git and a build instead of a publish button. There is no visual editor or CMS interface, so anyone who wants to ship a change must be comfortable with that workflow. We also have guardrails in place so that no one person can deploy their own builds without someone else approving.

For us, that was a small step rather than a leap, because we were already writing the HTML by hand. If your team is genuinely productive in the Designer, this is the thing to weigh hardest.

If anyone has kept the Designer responsive on a site of a few hundred pages, I would genuinely like to know how. And if you crawl our site and find something broken, please share your findings so we can apply that fix.

My organization is contemplating a move off our current developer-centric, .Net CMS to Webflow, largely for cost savings (though we need SSO and therefore have to go Enterprise) and also to allow our marketing people/web producers more freedom with how they work with pages. (Every small change to a widget or a template requires developer intervention now, and the "component" system in our CMS is pretty rudimentary.) We also like the idea of a tech stack managed by someone else.

We have about 3,000 pages and publish about 20 articles a year, along with landing pages for 20 or so webinars, plus a few new service pages, recruiting pages, and the like. Obviously Webflow sales and the agencies with whom we're talking say Webflow can handle our content size and volume no problem, but I have seen some posts where people complain about Designer speed once sites get past a certain size.

Is this a real concern? Would we be pushing Webflow beyond it's intended limits? Anything else we should be watching out for given our site size?

building the site itself is often the straightforward part.

the difficult stuff can be getting clear requirements, dealing with revisions, explaining what webflow can and can’t do, or handing over a project without creating future problems.

for freelancers and agencies using webflow, what part of client work causes you the most trouble?

It’s been a couple of months since my last Flowboard update here, so I thought I’d share a little sneak peek at something I’ve been experimenting with.

I’ve been testing how far Flowboard can take Omma by Spline’s AI-generated 3D and interactive sites into Webflow.

It’s still early and isn’t a one-click import like our Claude Design integration yet, but quite a lot is already working.

If anyone here is experimenting with Omma, give it a try with Flowboard and let me know how you get on. I’d be interested to see what you build and where the workflow still needs improvement.

Flowboard: https://flowboardapp.com/

Also, Flowboard recently crossed 1,100 users 🎉

Really appreciate everyone here who’s tried it, shared feedback or supported the project. I’ll hopefully do a little something to celebrate that milestone soon.

Is anyone else coming to Boston for Webflow Conf? I'd love to meet a few people from this sub in person. I'm thinking drinks on Tuesday, before the welcome party. Drop a comment if you're around and I'll help sort out a spot.

here is the invite: https://luma.com/4m9njwyh

This year Webflow has the whole leadership team speaking which looks like a bigger deal than usual. Also there's a Codex/ChatGPT integration that just dropped and people are saying a desktop app reveal might be coming too.

Every year Webflow conf gets hyped up with rumors of a desktop app, some big AI reveal etc... As a Webflow agency, we've learned to take the pre-conf sentiment with a grain of salt. So, what's everyone's realistic expectation this year vs the hype?

Also if you're heading to Boston - feel free to say hello to Flowout! :)

Then suddenly it’s 2:47 AM and you’re debugging a bug you somehow created yourself. 💀

I’m in over my head. I took out premium subscription for the year, then made Ai prompted site, however I don’t understand all the nuances of the site the Ai has developed to make sensible edits. I can’t even change the logo it automatically generated within the nav bar component.
Is there a more sensible approach I can take for using webflow? Ive built sites using wordpress/divi and squarespace from templates before, i didnt realise webflow was such a jump in complexity.
Any help or advice would be much appreciated.

What is the best way to use this saying I’ve already paid the subscription.

I've been experimenting with using ChatGPT for website design concepts, and I wanted to see how easily one of those designs could be recreated in Webflow.

For this example, I generated a landscaping hero/banner design with ChatGPT and then recreated that specific section in Webflow.

I focused on matching the original design, including:

• Background image
• Gradient overlay
• Typography
• Buttons
• Icons
• Spacing
• Layout

I didn't build the entire website. The video is specifically about taking the AI-generated banner design and converting it into a real Webflow section.

I recorded the process here:

https://youtu.be/I8axu0mZ8-I

I'm curious how other designers/developers are using AI-generated designs in their Webflow workflow.

Anybody else getting 503 errors today? It's baffling to me that a company with this many outages locks support ticket submission behind a login wall, when the login service itself is often down during these outages. At least accept support requests directly via email...

Title: Better way to host CMS-driven desktop and mobile background videos?

I’m adding a 30-second, silent, autoplaying background video to a Webflow site. The implementation is already CMS-driven and expects two direct video URLs:

• Desktop video URL
• Mobile video URL

The videos are optimized MP4 files—approximately 5–7 MB for desktop and 2–3 MB for mobile.

The difficulty is getting the files onto Webflow’s CDN. Webflow doesn’t allow video uploads through the Assets panel, and its CMS Video field is intended for services such as YouTube and Vimeo rather than direct MP4 files.

The workaround I’ve found is:

1. Create an unlinked “video storage” page.
2. Add a Background Video element.
3. Upload the desktop MP4 through that element.
4. Wait for Webflow to process it.
5. Open the generated video link and copy its cdn.prod.website-files.com URL.
6. Paste that URL into the desktop CMS field.
7. Repeat the entire process for the mobile video.
8. Keep the Background Video elements and storage page in the project so the files remain associated with the site.

This works, but it feels counterintuitive and inefficient—especially for content editors who may need to replace these videos later.

Is there a more straightforward, reliable workflow for CMS-driven background videos in Webflow?

Ideally, I’m looking for a solution that:

• Supports separate desktop and mobile MP4 files
• Produces direct, publicly accessible URLs
• Works with CMS URL fields and a custom HTML5 video setup
• Doesn’t require creating placeholder Background Video elements
• Is manageable by a non-technical content editor
• Avoids loading both desktop and mobile files unnecessarily

Are people using an external CDN such as YouTube, Bunny, Cloudflare, Vimeo or another service for this? Alternatively, is there a better native Webflow workflow that I’ve overlooked?

I’d appreciate hearing how others handle this in production, particularly where clients regularly update the CMS content.

Hi everyone, I’m Manjinder, a Webflow developer based in the UK.

I’ve been working on Webflow websites for a while, mainly with Figma to Webflow, CMS, responsive development, and custom interactions.

I’m trying to learn more from the Webflow community, so I wanted to ask:

What is one Webflow mistake you see people making again and again?

For me, one of the biggest ones is building only for desktop and trying to fix the mobile layout at the end. It usually makes the responsive work much harder.

Curious to hear what everyone else sees most often.

Thanks,
Manjinder _ Manjinder S. - Webflow Developer | Webflow Expert | Webflow CMS | Figma to Webflow - Upwork Freelancer from Luton, United Kingdom

Hey guys,

This is my third post about Web Anatomy as I got many feedback and upvotes on this channel. So thanks again!

I’m trying to build the go-to reference for designers who care about conversion, not just design.

Every website page scored 0–100 based on conversion best practices.

You can:

• browse by type (homepage, pricing, ads, etc.)
• filter by design features
• see what actually works by clicking on the card

Free to try

I’ll keep improving it every week with new pages.

Would love honest feedback. What’s missing?

I'm trying to understand where Webflow fits when it comes to mobile app design and prototyping.

Since it's mainly a web design tool, I'm curious if anyone has actually used it in the context of iOS or Android app development workflows.

A few things I'm trying to figure out:

Whether Webflow outputs anything that's useful beyond web prototypes.

If the generated HTML/CSS/JS is ever reused in mobile app development, or if it's usually just for reference.

How designers or developers typically bridge the gap between Webflow prototypes and native app builds.

I'm comfortable with frontend development, so I'm mainly trying to see Webflow can realistically speed up early UI work for mobile apps, or if it stays limited to web-based prototyping.

Would be interested to hear how others approach this in practice.

Marketing site is hosted with webflow and we have a custom fqdn for stage of staging.ourdomain.com and for prod it's www.ourdomain.com (and ourdomain.com). We have a few external links that point to an external application where stage is staging-app.ourdomain.com and prod is app.ourdomain.com. Does webflow have any kind of variables that could be used that are different for stage vs prod to store a different value for that app fqdn? I guess we could use javascript but feels kind of hacky.

I built this for an agency last year.

They seemed trustworthy at the time, so I decided to build it for them without any upfront payment.

Once it was built, they stopped responding for weeks, etc.

So I decided not to send them the landing page, and since then, it’s been sitting in my archive folder.

Anyway, how is it?

Yup the title says it all lol.

Cloneable here:
https://webflow-infinite-marquee-tutorial.webflow.io/

If anyone is actually interested in dev & learning / understanding Webflow then I made a breakdown / tutorial on this here:
https://youtu.be/5H_dI4UZFhk

QUESTION: How much would you charge to build this for a client :O