UI, AI, and new SaaS models at 37signals with Jason Fried

saas.unbound is a podcast for and about founders who are working on scaling inspiring products that people love, brought to you by https://saas.group/, a serial acquirer of B2B SaaS companies.

In episode #22 of season 5 of the saas.unbound podcast, Anna Nadeina sits down with Jason Fried, co-founder of 37signals and the visionary behind iconic products like Basecamp, Hey.com, Once.com, and the newest venture, Fizzy. This insightful conversation dives deep into product development philosophies, innovative UI experiments, challenges in SaaS business models, hiring practices, and how AI is influencing the future of software.

Building Hey.com’s Calendar: Experimentation Over Convention

Jason reflects on the lengthy journey of developing the Hey.com calendar, a feature that took roughly two years and around 18 to 20 six-week development cycles. Unlike merely recreating a traditional calendar grid, the Hey calendar was a novel product that required extensive experimentation to find the right ideas and interactions.

He emphasizes that much of the early work was exploratory, involving fluid, free-form ideation with a small team before transitioning into more structured development cycles. The complexity of calendar features—like accurately handling repeating events—and the high user expectations made this a challenging but rewarding endeavor.

When asked about involving customers early in development, Jason shares that 37signals typically builds the first version of their products for themselves, rather than seeking external validation upfront. This approach allows them to mold the initial product vision without outside noise. Only after launch do they gather user feedback, which they carefully absorb and integrate over time, letting users adjust to new paradigms before making drastic changes.

Building for Yourself vs. Building for Customers

Jason contrasts his approach with teams that rush into building products without validating the market. He stresses the importance of building because you want to build—not just because others are doing it or because you think “if you build it, they will come.”

He highlights that even a well-built product faces one of the toughest challenges: distribution. Getting people to notice, try, and switch to something new is often harder than the building itself. For small teams without large marketing budgets or audiences, this is a significant hurdle.

Jason also critiques the common practice of asking friends or acquaintances if they would pay for a product, noting that such feedback rarely predicts actual customer behavior. The ultimate test is real-world conversion—people discovering your product independently and deciding to pay for it.

Introducing Fizzy: A Fresh Take on Bug and Issue Tracking

Fizzy, 37signals’ latest product, began as a quirky idea inspired by bugs splatting on a car windshield during a road trip. The team initially envisioned a bug tracker represented by splats on a screen, symbolizing the mess of unresolved issues. However, practical UI constraints led them to evolve the concept from splats to bubbles, and finally to a two-column card-based interface.

This card system organizes tasks into “considering” and “doing” columns—an approach inspired by Basecamp’s card tables but simplified to focus on what matters most. Jason explains that while Basecamp traditionally tracks bugs internally, Fizzy offers a dedicated space to manage issues more immediately and clearly, without the distractions of other project elements.

One of Fizzy’s innovative features is the “golden ticket” system, which allows users to mark any task as a high-priority item that visually floats to the top of the list with a golden glow. This approach eschews traditional priority labels like high, medium, or low, favoring a simple, intuitive way to highlight what truly needs urgent attention.

Fizzy’s Broader Vision

Jason envisions Fizzy as a tool not just for software developers but for anyone who manages “issues” or “bugs,” whether in homes, rental properties, or other systems. Unlike complex, technical bug trackers like Jira, Fizzy aims to be approachable and useful for a wide range of users who need to track and resolve problems efficiently.

Future Integrations and AI Features

While Fizzy is still early in its development, it plans to include APIs and webhooks to integrate with tools like Slack or Campfire, allowing seamless notifications and updates. Additionally, the product experiments with a new UI paradigm featuring a command-line style interface called the “do bar” (fizzydo), enabling quick keyboard-driven commands and natural language queries to interact with tasks.

Jason shares that this command-line interaction will eventually extend to Basecamp, signaling a fresh way to navigate and manage data faster than traditional mouse-driven interfaces.

Reflections on Campfire and the Once.com Model

Campfire, 37signals’ self-hosted group chat product sold through Once.com, experienced an initial surge in sales but has since settled into a modest, steady stream. Despite making installation remarkably simple, Jason acknowledges that self-hosted software remains a tough sell for many users accustomed to SaaS convenience.

He sees the self-hosted model as promising but still early, with a niche of tech-savvy users embracing it. The team has also created Writeboard, a tool for publishing online books, which is free and open source, but both products are currently on the back burner as they focus on SaaS offerings like Fizzy.

There’s also an intriguing future possibility: starting with Fizzy as a SaaS product and later offering customers the option to host it themselves, blending the best of both worlds.

Hiring at 37signals: What Makes a Standout Candidate?

In a competitive hiring process for a product designer role that garnered over 1,400 applications, Jason shares insights into what sets candidates apart. The team quickly filters out unqualified applicants, especially those who fail to follow instructions or submit generic cover letters.

Key factors that elevate candidates include:

• Effort and Personalization: Candidates who invest extra effort, such as creating micro-sites or tailored applications, immediately stand out.
• Strong Writing Skills: Since written communication is fundamental at 37signals, well-crafted cover letters are crucial.
• Creativity and Design Imagination: The team looks for originality and the candidate’s unique design voice, not just adherence to existing styles.

After multiple interview rounds and paid design assignments, the final choice often comes down to intangible qualities—such as a palpable hunger for the role and a genuine desire to grow with the company. Jason notes that sometimes the best candidate isn’t the one who initially seems perfect but the one who continuously impresses throughout the process.

Measuring Developer Performance: Focus on Trajectory, Not Metrics

When discussing developer productivity and performance evaluation, Jason offers a refreshing perspective: rather than relying on metrics like lines of code or ticket velocity, the team focuses on trajectory and overall contribution.

Key takeaways include:

• Know Who’s Pulling Their Weight: In a small team, it’s clear who is contributing effectively.
• Focus on Improvement: Is the person growing, stagnating, or regressing?
• The One-Year Rule: After a year, the critical question is: “Would we hire this person again?” This simple question cuts through complex evaluation data.
• Acceptance of Different Growth Paths: Some developers may plateau at senior levels, which can be perfectly acceptable depending on the role and expectations.

Jason cautions against traditional priority labels and metrics, noting that physical order (like dragging tasks to the top) is often a clearer and more effective way to communicate importance.

Personal Hacks and AI Usage

Jason shares some personal productivity hacks, including a return to sketching ideas with pen and paper. He finds that limiting himself to a single sharpie marker helps focus on the core idea rather than perfecting details prematurely.

Regarding AI tools, Jason uses ChatGPT primarily for refining and simplifying his writing rather than generating content outright. He appreciates ChatGPT’s balance of conversational style and helpful suggestions, preferring it to other AI assistants like Claude or Perplexity.

He also shares a charming anecdote about how his 10-year-old son uses ChatGPT extensively, fueling his curiosity and learning, from baking recipes to making paints. The family enjoys interactive voice sessions during car rides, highlighting how AI is becoming an educational companion for the next generation.