Three people asked me some version of the same question this month: should I get off Webflow?
That's new. I've been building on Webflow for years, across my own brands and client work, and it's never been the kind of no-code visual development platform people asked me to defend. It just worked, and worked well enough that the question never came up. Now it's coming up a lot, and I think it deserves an honest answer instead of a hot take.
I still use Webflow. I still recommend it for certain projects. But I'm not going to sit here and pretend the last few months haven't given me real pause, because they have, and if you're paying attention, they've probably given you pause too.
What Actually Happened, and Why Everyone's Suddenly Asking
If you haven't been following it closely, here's the short version of what changed in the last two months.
On May 13, 2026, Webflow rolled out a pricing restructure it called a "simplification." CMS and Business plans got merged into a new "Premium" tier, CMS item limits went up, and AI credits got bundled into every subscription workspace. Sounds fine on paper. But the base bandwidth on that new Premium tier dropped from 100GB to 50GB, and for most freelancers and small teams on monthly billing, it works out to a net increase once you run the numbers, not a decrease. It's the second pricing change in six months, and it's the kind of shift that makes you start thinking seriously about total cost of ownership instead of just the headline price.
Then on May 27, CEO Linda Tong published a blog post titled "Evolving Webflow for the Agentic Web," announcing a company restructuring. It was the second layoff round under her leadership in under two years. What made it sting for people watching wasn't really the layoff itself, companies restructure, that happens, it was how it was handled. Employees found out because their laptops locked at 7am. No email, no meeting, no calendar invite first. One affected engineer, on a closed work visa with his family's relocation on the line, posted publicly asking Tong directly if he still had a job, hours before the company's official explanation landed as a blog post about "the agentic web." He'd defended her that same morning, saying he didn't think she'd handle it that way again after the last round.
I'm not sharing that to pile on. I'm sharing it because it's the clearest example I've seen of the thing a lot of us in the Webflow community have been feeling without being able to name: it doesn't feel like anyone's really talking to us anymore. Not the freelancers, not the agencies, not even, apparently, a lot of the employees. And every time a platform this size goes through a stretch like this, without a confirmed acquirer, without a clear roadmap, and without much internal transparency, it's fair to ask how much of this is normal operational risk and how much is something bigger.
The Community Isn't Imagining This
I've spent a fair amount of time lately reading what other agency owners, developers, and freelance Webflow developers are saying, and it's a consistent picture, not a fringe reaction.
The recurring theme is a quiet structural decline that's been building for close to a year: more dashboard lag, more interface bugs, and a few high-profile hosting outages that used to be rare and now feel routine. If you build client sites for a living, that's not an abstract complaint, that's your reputation on the line every time a client's site hiccups and you have to explain why.
Layered on top of that is a strategic direction that a lot of longtime users don't recognize themselves in anymore. Webflow spent years building its identity, and its ecosystem, around designers, freelancers, and small agencies, people who wanted visual control without giving up front-end quality. The "agentic web marketing platform" language in Tong's restructuring post points somewhere else entirely: enterprise teams, AI-native workflows, governance at scale. That's not automatically a bad direction. But it does mean the platform's own leadership seems to be deprioritizing the exact community that made it what it is, and doing it while pricing keeps climbing for the people getting deprioritized. That combination is what's driving the "is it still worth it" conversations I keep running into.
What Webflow Still Gets Right: CMS, Custom Code, and Technical SEO Fundamentals
I want to be fair here, because it's easy to pile on a platform mid-news-cycle and lose the plot.
Webflow still outputs real HTML, CSS, and JavaScript under the hood, which puts it closer to actual front-end production than most no-code tools. You can still use custom code, API connections, and third-party integrations. It's not a closed toy, and the technical foundation hasn't degraded just because the business side is going through a rough stretch.
The Webflow CMS is still one of the best reasons to use it. Fast publishing workflows inside the Webflow Designer, a client-friendly editor, direct control over presentation, that combination is genuinely hard to replicate elsewhere without giving something up. For content-driven marketing sites, resource hubs, and local SEO-focused small businesses, none of that changed on May 13 or May 27.
Technical SEO is still solid at the platform level too: automatic sitemap generation, 301 redirect management, canonical tag support, and clean robots.txt configuration. Those fundamentals matter because they protect indexation, they keep search engines from wasting crawl budget on the wrong pages, and they're the foundation everything else, Core Web Vitals, Lighthouse scores, structured data, gets built on top of. The tools are there. Whether a site actually performs well still comes down to how it's built and maintained on Webflow hosting, that was true before this year and it's true now.
Local SEO, Internal Linking, and Structured Data: Where Execution Still Beats Platform Debates
This is the part of the conversation that gets lost when everyone's arguing about whether a platform is "dying." Most of what actually drives results has nothing to do with which logo is in your footer.
Solid URL structure, internal linking that reflects real information architecture instead of navigation convenience, and clean redirect mapping during redesigns will make or break a site's indexation long before platform choice does. Schema markup and structured data help search engines understand what a page actually is, which matters even more as AI-driven search results lean harder on that signal. And for businesses trying to scale content, programmatic SEO is genuinely possible in Webflow, but only with disciplined collection design; sloppy structure creates index bloat that looks like growth in the CMS and decay in Google Search Console.
Measurement is what separates confidence from assumption here. Google Search Console tells you whether pages are actually indexed. Google Analytics tells you whether that traffic converts into leads. Neither of those tools cares what platform you're on, they just tell you the truth about execution.
AI-Assisted Design vs. AI-Assisted Everything: Where the Market's Actually Moving
Here's the piece that worries me more than any single pricing change or layoff. AI-assisted development is genuinely changing how fast people can go from idea to shipped page, and that pressure isn't coming from Webflow's old competitors, Wix, Squarespace, WordPress. It's coming from a whole new category of tools that let people prompt their way to a working site in an afternoon, using an LLM and something closer to a GPT-style conversation than a design canvas.
Webflow knows this, it's literally why they're pivoting toward "agentic" workflows and leaning harder into AI-assisted design and AI content generation across the product. But knowing the threat and outrunning it are different things, and right now those AI features feel like they're arriving carefully and incrementally while the market outside is moving in leaps. Personalization, in particular, is trending toward being integration-driven rather than page-builder-driven, meaning the intelligence increasingly lives in APIs and external logic rather than inside any single visual tool. I don't think the gap closes itself. It closes because Webflow ships something that actually changes the calculus, or it doesn't, and more of us drift toward faster, more composable alternatives one project at a time, not because we hate Webflow, but because the next deadline doesn't care about brand loyalty.
They Are Shipping Updates, So Why Doesn't It Feel Like Enough?
To be fair to Webflow, it's not like they've gone quiet on the product side while all this plays out. Check the Webflow updates page and it's a steady drip: AI code components, an AEO push now available for Enterprise, one-click app deploys on Webflow Cloud without spinning up a full site, a dedicated Localize panel, and, notably, a new setting that logs every change in your Site Activity Log as either human, Webflow AI, or MCP-connected tool. That last one tells you exactly where their head's at.
Here's my honest reaction to all of it: I see the releases, and I still don't fully trust them yet. Not because the ideas are bad, but because a lot of what ships lately feels half-finished the moment you actually try to use it for real work. You get the announcement, you get the demo, and then you get into it and find the edges that weren't sanded down.
The MCP server is the clearest example, and it happens to be the exact thing people are asking for: connect an AI agent to Webflow and prompt your way to a built site, the same way you'd work in Replit or Cursor. On paper, that's the future everyone wants. In practice, right now, it's rough. It can't touch interactions at all, you still have to build those manually in the Designer. Style updates through the MCP tools are reportedly timing out entirely for some developers, to the point where property changes just don't apply. It needs a live Designer tab open and pinned in your browser for a chunk of its capabilities, or it silently stops working. That's not a finished product, that's an early beta wearing a "shipped" badge.
And that's really my concern with the broader release cadence, not any one feature. A lot of what's coming out lately reads like it exists to be announced more than it exists to be used. You get the update, the blog post, the changelog entry, and then when you actually build with it, the value isn't fully there yet. That matters for Webflow ecommerce too, where merchandising, filtering, and checkout complexity already push a lot of stores toward integrations rather than native tools, so half-finished AI features layered on top don't inspire confidence. Maybe it gets there. Maybe six months from now the MCP server is genuinely reliable and half of this critique ages badly, and honestly, I'd be glad to be wrong. But right now, testing has to happen before I'd trust any of these AI-adjacent features on a client project, and that gap between "announced" and "actually works" is exactly the kind of thing that erodes content velocity and confidence over time.
What I think people actually want is simpler than anything on that updates page: the ability to prompt their way into a working site the way they already can in Replit, Cursor, Claude, or Lovable, with Webflow's design quality and CMS underneath it. That's the pitch of the "agentic web" language in Tong's restructuring memo. It's just not clear yet whether that's genuinely the plan, or a good phrase attached to a roadmap nobody's fully seen, because right now, with the communication issues we've been talking about, we don't actually know what they have planned. If Webflow lands that simple version of prompt-to-site well, execution matches ambition, it's a real turnaround story. If it stays a collection of half-working integrations announced faster than they're finished, it just adds to the uncertainty instead of resolving it.
Vendor Lock-In, Design Systems, and Governance: The Boring Stuff That Actually Protects You
Vendor lock-in in no-code platforms is rarely about your text and images being trapped. The bigger issue is that Designer-specific layouts, CMS schemas, components, and Marketplace dependencies don't translate one-to-one into another system, so migration cost usually comes from structure rather than content.
That said, Webflow is less risky than a lot of visual builders because parts of the stack are still portable if you plan properly. A documented design system, consistent class naming conventions, a clear component strategy, and disciplined custom code boundaries make a future migration slower than people want but far more manageable than people fear. If you ever do need to move to a headless CMS setup or a fully custom build, clean information architecture is what makes that possible without starting over.
Governance matters just as much once more than one person touches a project. Permissions, role-based access, and change control keep the visual layer from turning into technical debt inside your publishing workflow. The practical mitigation is boring and therefore often skipped: keep exportable assets where appropriate, document your CMS schema, preserve redirect mapping, and maintain a running list of every integration, every script, every place custom code touches Webflow, so portability is a habit instead of a panic project.
Pricing, Total Cost of Ownership, and Whether the ROI Still Adds Up
Webflow pricing is only the visible part of the cost equation. Total cost of ownership includes the subscription itself, plus maintenance, SEO work, integrations, governance overhead, and developer time, and that math just got more complicated with the May pricing changes.
ROI should still be judged through speed-to-launch, conversion rate, lead generation, and the cost of making changes without breaking things, not through the headline subscription price alone. That's why a lot of businesses still do well on Webflow even when another platform looks more flexible on paper, flexibility that requires constant technical intervention often becomes expensive in practice. Webflow templates and the broader Webflow Marketplace can lower cost when a site's structure is straightforward, but custom builds still require teams specialized in Webflow development with high-level execution, and that extra investment only makes sense when the business gains enough from stronger architecture and ongoing iteration to justify it.
So, Is It Still Reasonable to Build on Webflow? A Business Owner's Decision Guide
Honestly, yes, for the right project, with eyes open.
If your main need is a fast, visually controlled marketing site, a content-driven small business site, or something local SEO-focused where speed of iteration matters more than deep application logic or enterprise-grade governance, Webflow is still a strong choice. That hasn't changed.
If you're depending on it for heavy app logic, or you need airtight predictability on pricing and roadmap, or you're watching bandwidth costs closely as a freelancer on monthly billing, I'd go in with a real cost model and a migration plan sitting in a drawer, not because you're leaving, but because betting blind on any single platform right now feels riskier than it did two years ago.
A business-owner lens changes this evaluation for me. As a Las Vegas native with deep roots, bringing hands-on experience from owning multiple local Las Vegas businesses, giving me a business-owner's perspective on SEO and strategy, I care less about platform fandom and more about whether a site keeps producing leads without becoming harder to operate every quarter. That's the standard I hold Webflow to, and it's the standard I hold my own work to. I'm specialized in Webflow development with high-level execution, but the platform is only half the equation. Local SEO and broader digital growth to drive visibility and leads depend just as much on ongoing Webflow support and maintenance to keep sites evolving as they do on which builder you picked. Combining technical Webflow expertise with results-driven business growth strategies is how you avoid vanity metrics, and it's the whole point: we ensure your site converts, focusing on tangible business outcomes like leads, bookings, or sales, not on which platform sounds more future-proof in a blog post.
Where I Land
I'm not ready to say Webflow's future is settled, in either direction. What I can say is that the platform I recommend without hesitation and the platform going through a rough, poorly-communicated stretch right now are the same platform, and that's an uncomfortable place to sit as someone who builds on it for a living.
But I've also seen platforms come back from stretches like this. It genuinely could take one strong, well-communicated update, something that closes the AI gap without alienating the community that got it here, to turn the sentiment around. Webflow still has the technical bones, the CMS, and a massive existing base of sites to build from. What it needs to prove now is that the "agentic web" pivot is additive to its ecosystem, not a replacement for it.
I'll keep building on it. I'll also keep watching closely, and I'd encourage you to do the same instead of taking either the doom takes or the marketing at face value.