Can Gamma REALLY Create a Website in 30 Minutes?

Full video transcript

Coming up to the top of the hour for day three of Gamma, which has been blowing minds for the last few days showing people how to use this powerful platform to get ideas out there, presenting them in presentations and full-blown websites. Something special today, even the host doesn’t know what Gamma is about to bring to the table. Hillary, Melissa, and Deo join for day three of the Gamma Lab.

Today’s session with Allison and Melissa is a design battle. The history of design battles at Gamma goes back to almost a year ago, November of 2023, when the team wanted to see if what they preach is actually true: can you build a really awesome pitch deck, presentation, or site in 30 minutes? There are three goals: can you build something good in 30 minutes, does it help test the product and find bugs, usability issues, and feature requests, and how far can the design functionality be pushed. Battles are done internally all the time. The format: a 30-minute timer, the audience votes on a few aspects, Melissa and Allison get a topic they’re not aware of, they design for 20 minutes, get a curveball at the 20-minute mark, then have 10 more minutes to incorporate it before pitching their result.

The audience first voted on site versus pitch deck; sites won by a landslide. Next the audience voted on the topic from three surprise options: an exotic animal daycare, a wizard’s beard grooming salon, or a space vacation rental agency. Space vacation rental agency won about six votes to one. The audience also suggested business names in chat, including Bottomless Martian Martinis, Starscape Escapes, Galactic Getaways, Zero-G Vacations, Cosmic Escapes, Star BNB, and Orbit Oasis.

With the timer started, Allison went straight into Gamma’s regular generate flow to get something down fast, not worrying yet about content quality, planning to refine it later. Melissa instead started in ChatGPT, drafting an outline there to paste into Gamma’s AI generator afterward. When Melissa pasted her ChatGPT outline in, it initially split into numbered fragments rather than clean sections; she used Gamma’s “plus” feature to add more content and chose “preserve” in the generator so her pasted content stayed close to as-is rather than being rewritten. Allison, meanwhile, adjusted AI image generation by deleting an existing prompt and using “enhance prompt,” which writes a new prompt based on the card’s current content — a feature she said she uses constantly.

By the six-to-seven-minute mark, Melissa was choosing between Aurora and Nebula themes for Orbit Oasis, eventually going with Nebula, and used Gamma’s remove-background tool (available on all plans) on an AI-generated logo. The host noted logo generation is not a substitute for hiring a real logo designer, but was fine for the demo. Allison, building out Bottomless Martian Martinis, converted a list of offerings into “package” cards and worked on ordering AI-generated images. A question came up about the “additional instructions” field in the generate flow: it’s used to describe color, style, or mood (for example browns and oranges) to further guide AI image generation, separate from the main content prompt.

Around the halfway point, Allison used the “condense text” feature to shorten overly wordy copy and converted content into toggles, using the pound-sign text shortcut to resize toggle text and the “::” emoji-menu shortcut to insert star ratings. With about seven and a half minutes left before the curveball, both designers were still refining images and layout; Allison added a “merch store” concept (a t-shirt idea suggested by the AI, “I Survived the Black Hole”) as an aspirational idea since Gamma doesn’t currently support e-commerce.

The audience voted on two curveball options; the winning curveball was to add a relevant diagram to the site, of the designer’s choosing. Allison, uncertain what a “relevant diagram” should be, opened Gamma’s freeform slash-draw-diagram tool and began building something from scratch, later adding a pyramid and a timeline. Melissa built tables and charts around traveler numbers and flight history, working with arbitrary figures. Audience suggestions for diagram ideas included a growth chart, an org chart, a projected profit margin, a history timeline, and a “joy map” showing happiness levels during the trip, which both designers liked as a concept. Melissa used Gamma’s “bar stats” smart layout, which auto-fills bars (or circles) based on entered numbers; Allison expressed uncertainty about how to map data logically into a matrix-style layout comparing planets on luxury versus temperature. With the clock counting down, both used AI suggestions (“make it shorter,” “add more content”) applied via the sparkles icon, with Command-Z available to undo changes that didn’t work.

When time was up, both presented. Allison walked through Orbit Oasis: private suites with their own bathroom and window, climate control, multiple travel packages, zero-gravity experiences, a curated photo service, reported 100% safe journeys and 90% new experiences, customer reviews, an FAQ section, and a merch store with a hat, shirt, and stickers. Her stated pain points: refining content took longer than expected because it was her first real read-through of AI-written copy, and the diagram step during the curveball was the hardest part, since she wasn’t sure which layout best fit the content. She also said she and Melissa would like an AI feature that reviews a finished draft and suggests concrete improvements.

Melissa then presented Bottomless Martian Martinis (BMM): scroll-linking navigation links, stellar destinations including Mars, the Moon, Jupiter’s clouds, and Europa’s icy caves, gravity-controlled beds, “Galactic room service,” guest testimonials, a booking guide, promotional offers including a “moon rock proposal package,” an FAQ section, reported figures of 2,590 flights and 9,491 happy travelers, and a merch section. Melissa’s stated pain points: she spent meaningful time generating images, and she wants a real logo generator in the product rather than the current AI logo output, calling the current process more like teaching a kid through feedback than getting a finished result.

After presentations, Melissa answered audience questions by demonstrating Gamma’s import and site-conversion features live. She showed that any existing Gamma presentation can become a site, and vice versa in spirit, though you cannot directly revert a published site back to a presentation — the workaround is to duplicate the gamma before publishing it as a site, so the original presentation is preserved. Publishing as a site assigns a default domain (customizable), adds a header where you can attach a logo and links, lets you link elements to specific cards so clicking scrolls to that section, lets you add additional linked sub-pages (via AI or blank), and lets you toggle each card between a bounded “card” look and a full-bleed “web page” look via page setup.

She also clarified the import workflow that had come up in earlier sessions: from Create with AI, choosing Import from URL and pasting a link scrapes that page’s content, whether it’s a Gamma-built site or a page built elsewhere, and pulls it in for a new type of content (for example, importing a presentation to build a webpage). Content can be split into sections manually with three dashes between blocks, or Gamma can be asked to split it automatically, though results vary. Alternatively, Import File accepts a Word doc, PowerPoint, or PDF upload, or a file from Google Drive, and pulls all of that content in; “preserve” keeps content close to the source, while other settings allow output in multiple languages (French, Arabic, and others were mentioned) and choice of format, such as website or social media carousels.

Closing out the session, the team thanked the audience for watching across all three days of the Gamma Lab and shared a promo code (AI Summit) for three months of Gamma Pro with unlimited AI features, noting it would expire soon. They mentioned there are no regular office hours currently, but encouraged submitting feedback and bug reports through the in-app feedback link (accessible via the question-mark icon), which goes directly to the team, includes a public board to upvote requests, and shows the product’s recent change log.