Motiff UX Teardown: A Sales Page With a Shutdown Notice on Top

There's a red bar across the top of Motiff's website. It says the product has been discontinued and that you can export your data until October 31, 2026. Underneath it, the page carries on as if the bar isn't there: an enormous headline claiming "The best UI generator," a benchmark section arguing that the claim is measured rather than boasted, a gallery of generated interfaces, a wall of happy users, and a pricing page that will take sixteen dollars a month from you right now. Nothing below the bar acknowledges the bar. That gap is the most instructive thing on the site. It's also not a mistake, exactly. It's what a site looks like when one component was given the job of announcing the ending and every other page was left holding the argument.

The whole product, reproduced in the hero

Start with what Motiff got right, because it's genuinely good. Most AI design tools put a screenshot of a canvas in the hero: panels, layers, a cursor caught mid-gesture, and an invitation to picture yourself in there. Motiff skips the picture. It drops the actual input into the page — a big empty box reading Tell AI what you want to design …, with four chips underneath naming what you're allowed to put in it. Text to UI. Wireframe to UI. Screenshot to UI. PRD to UI.

Here's the funnel across three pages — step through it:

That box is an argument about scope. It says the tool has one surface, one gesture, and no learning curve worth mentioning. You don't have to recall what a design tool's chrome means or map your intent onto a toolbar; you recognize a text field and you type (NN/g on recognition vs. recall). The four chips do the rest of the work, converting an intimidating open prompt into a list of things you already have lying around — a paragraph, a sketch, a screenshot, a PRD.

It's a confident reduction, and it's worth pausing on how much got thrown away to reach it. Motiff started as a full browser-based design editor going directly at Figma, with an AI layer on top of a real canvas — component inventories, design-system consistency checks, the whole apparatus of a tool you move into. That product is gone. The only trace of it anywhere on the site is a footer link: Export old Motiff editor files.

The gallery is the product page

For a generator, the sales pitch is unfalsifiable until you see output. Motiff clearly knows this. Use cases skips the customer-logo ritual entirely and just opens the catalog: two dozen finished interfaces, full size, no signup between you and any of them.

Look at how the tiles are captioned. Not by feature, not by industry problem, but by taste: high-end gallery ambiance with warm bronze accents and spotlight effects; timeless Scandinavian design crafted for everyday life; wireless charging, perfected. The category chips across the top let you filter by domain, but the descriptions are all aesthetic. That's a deliberate answer to the accusation every AI generator faces — that it makes one competent layout over and over and swaps the copy. By making style the axis you browse along, the page invites you to check for exactly the thing you suspect. It's an unusually brave way to organize a proof gallery, and it only works if the range is real. The page lets you decide.

The paywall is a door, not a quota

Pricing is where the strategy gets stated in numbers, and it's the clearest page on the site.

Two cards. Free gives you 100 credits, up to 10 UIs, and export as images — PNG, JPEG, WebP. Pro, at $16 a month, gives you 1,000 credits and then the three lines that actually matter: copy generated UI to Figma, export production-ready code, download IDE-ready project ZIPs, with a row of little tool marks underneath naming where that ZIP is expected to land.

Read those tiers again. The free plan is not a smaller generator; it's the same generator with the exits sealed. You can make the UI and look at it, and a flat image is where it stops. Every paid line is a door out of Motiff into the tool you were always going to finish in. Most SaaS meters how much you can do and charges you to do more. Motiff meters where the work can go, which is an admission that it never expected to be the place your work lives. It's a tool you pass through.

Which is a coherent, even principled position for an AI generator to take. It is also a strange thing to be charging monthly for, on a page whose top edge announces the product has been discontinued and gives you a date to get your files out.

The page that never mentions its own ending

Read the homepage top to bottom and the disagreement becomes structural rather than incidental.

The white hero gives way to a pale band where four claims sit beside four screenshots of the tool doing the thing being claimed. Then the page goes fully black to argue that "best" is measured rather than asserted, backed by a "comprehensive dataset," daily expert review, and stated scoring criteria. Then white again, for a mosaic of generated screens. Then black again, for testimonials. It's a well-paced page. Every section of it was written by someone who assumed there would be a next quarter.

And running above all of it, on every single page, is the bar.

The red strip is doing textbook work in isolation: it's persistent, it's the highest-contrast element on the site, it states the fact and gives the deadline, and it offers the one action that matters — Export now (visibility of system status, the first of the ten heuristics). If you were grading the component, it would pass. The problem is that a shutdown isn't a component-level event. It's a fact about the product, and the product's own pages haven't been told. The hero still recruits. The pricing page still asks for a subscription with no mention of what happens to it after October. The testimonial wall still speaks in the present tense.

Most teams will meet this eventually, and most will meet it the same way Motiff did: sunsetting gets scoped as a banner. One component, one sprint, ship it. What's left is a site where the bar tells the truth and every page under it keeps making a promise the bar has already withdrawn. Nobody is lying. It's just that the reconciling never got done, so the visitor does it, scrolling.

What Motiff got right on the way out

There's a version of this teardown that files Motiff under "company stopped caring." The site won't support it. These are the pages of a team that had worked out precisely what their product was — something you pass through on the way to Figma or an IDE, never somewhere you move in — and had gotten every surface to say so.

Which makes the ending unusually clean, in a way that's worth stealing. A tool built for you to take your work elsewhere is a tool that's easy to leave. The export link in the footer isn't a concession bolted on at the end; it was the business model. Most products that shut down have to invent an exit under deadline pressure, and their users can tell.

So the banner isn't the failure here. Motiff shipped it everywhere, high-contrast, with the date and the one action that matters. The failure is that shipping it was treated as finishing. An ending changes the tense of every promise a site makes, and no component can do that on its own — someone has to go back through the pages and reconcile them. That's the expensive part, it happens when the team is smallest, and it is the part nearly everyone skips.