Marketing Against the Grain · Hubspot Media

I Replaced a $5,000 Designer with Google Stitch

March 31, 2026·15 min·1 clip

Matt Wolfe shows how Google's Stitch can replace a $5,000 designer by generating landing pages from a simple prompt.

The episode opens with host Kieran introducing Matt Wolfe, a recurring guest and AI tools expert, to walk through Google's newly released Stitch design tool. Wolfe frames Stitch as Google's answer to the 'vibe coding' trend, describing it as a Figma-like canvas tool that generates website designs from natural language prompts. He contrasts the current workflow with the old approach of spending a week building landing pages in tools like ClickFunnels, LeadPages, or OptimizePress. Wolfe demonstrates how he prompted Stitch with his own Future Tools website URL, and the tool automatically reverse-engineered the site's color palette, fonts, and layout to create a style sheet. From that starting point, Stitch generated three distinct design variations — bold dark, warm energetic, and cool professional — each with different layouts and color schemes. Wolfe then shows how users can iterate further by voice, asking for new hero images or headline variations for split testing, all within seconds. He attempts to build a second project: a stats dashboard comparing the top ten large language models, asking Stitch to pull live data from the web. The tool produces a visually polished dashboard called Neural Core but fails to retrieve current data, instead drawing from its 2024 training data. Wolfe tests whether providing a direct URL helps the tool pull accurate model data, but finds it cannot actively scrape web content despite being able to match color schemes from a URL. He notes the limitation clearly: Stitch can visually analyze a page but cannot extract structured data from it. The episode then moves to the export workflow, showing how a completed Stitch design can be sent directly to Google AI Studio, which recently received an overhaul. AI Studio imports the HTML and markdown files and presents them in a simplified IDE environment. Wolfe explains that Stitch is intentionally kept simple for the design phase, while AI Studio is where functionality is built out. He demonstrates that AI Studio can generate a predictive heat map showing where users are likely to click, a feature that previously required paid analytics tools. Wolfe walks through building additional pages — a models page and a compare page — directly in AI Studio, showing that the tool preserves the original design aesthetics while adding new functionality. He briefly covers publishing, noting that Google Cloud hosting is required and that the app gets a public URL without needing a custom domain initially. Wolfe compares Stitch to competing tools like Lovable and v0, noting that Stitch appears to be free with no obvious usage limits at the time of recording. He also speculates that AI agents like OpenAI's operator-style tools could eventually automate the entire workflow — visiting Stitch, generating a design, exporting to AI Studio, and delivering a live website autonomously. The episode closes with Wolfe summarizing the recommended workflow: prompt in Stitch, export to AI Studio, then iteratively build pages. He reiterates that despite being desensitized to new AI releases, Stitch genuinely impressed him.

As heard by us

A clear, practical test of Google Stitch as a fast concepting tool for marketers, with honest limits on data grounding and output readiness.

Google Stitch comes across as a useful shortcut for marketers who need design concepts before they can get a designer or developer involved. The episode works because Matt Wolfe does not just praise the tool; he runs it through a real prompt-to-dashboard exercise and lets the…

Read the full review in PlayNext →

Why you'd press play

Need faster mockups without waiting on a designer?

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Marketing Against the Grain