Interview
Brand Design Tips From Linear Founder Karri Saarinen
Brand Authenticity & Stage Alignment
- Kari Saarinen advises startups to avoid mimicking mature companies (e.g., Stripe, Linear) early on, as it creates misaligned user expectations.
- Over-polishing a website before the product is ready sets false expectations regarding the product's maturity and feature set.
- Website branding must evolve with the company stage; Linear's site shifted from a simple, one-page launch in 2019 to a complex enterprise platform as the customer base grew.
- Startups should distinguish between their investor pitch (ambitious, long-term vision) and their website copy (specific, immediate utility).
Target Audience Specificity
- Early-stage copy should use specific technical jargon (e.g., "issue tracking") to resonate with ideal users and filter out irrelevant traffic.
- Linear's original site explicitly used "issue tracking" to attract engineers, avoiding vague terms like "work platform" common on enterprise sites.
- Saarinen suggests that startups should identify one core attribute where they outperform competitors (e.g., Linear's speed) to attract early adopters who value that trait.
Visual Strategy & Distraction
- Effective landing pages often use faded or obscured product screenshots to build mystery without overwhelming users with unready features.
- Animations should serve a functional purpose; excessive movement (e.g., looping backgrounds) can distract users from core value propositions.
- Elements that resemble standard web patterns (like floating cookie banners) risk being ignored by users conditioned to dismiss them.
- Enterprise-facing sites require more text, security details, and customer stories, whereas consumer sites can rely on visual abstraction.
Linear Website Evolution (2019–2025)
- 2019 Launch: A single-page site built in two days, featuring a "faded" product screenshot and explicit "issue tracking" copy to drive waitlist signups.
- Purple Phase: A second iteration utilized a purple aesthetic that was widely replicated across the industry.
- Current Iteration: Features a black-and-white design to signal maturity while retaining playful effects; clearly segments features into "Issues," "Projects," and "Product Roadmap" for technical teams.
Feedback on Submitted Websites
- Sprites.ai:
- Critique: Highly stylized graphics and floating elements distract from the value proposition, making it unclear who the target audience is.
- Recommendation: Group workflows into specific user categories (e.g., "for analysts," "for YouTubers") rather than a generic list.
- UI Note: Floating prompt elements resembling cookie banners risk being ignored; they should be integrated into the main content flow.
- Giga ML:
- Critique: Low conversion rates are attributed to the enterprise sales cycle, where users rarely click "Book a Demo" directly from a site.
- Opportunity: The headline "AI agents for enterprise support" is too generic; it should specify the persona (e.g., CTO, VP of Support) and the specific problem solved.
- Technical Details: The AI voice agent's description lacked clear guidance on what questions to ask, potentially confusing users.
- Unreal Milk:
- Strengths: Highly memorable, hand-drawn aesthetic successfully communicates an "organic" brand identity despite the lab-grown product.
- Weakness: The site functions more like a narrative announcement than a sales page, lacking a clear Call to Action (CTA) or "buy now" options.
- Recommendation: Add an email capture CTA framed around a social mission ("Join the cause") to build a community of interested early adopters.
- Confident AI:
- Clarity Issue: The relationship between "Confident AI" and the open-source "Deep Eval" tool was not immediately obvious, requiring users to click external links to understand the offering.
- Visuals: Dark purple links with underlines evoke outdated browser aesthetics; headlines like "Build in a weekend" are too vague to convey specific benefits.
- Strategy: The site effectively leverages GitHub stars and open-source adoption to build trust but lacks differentiation from competitors in the "evaluation" space.
- Dropback:
- Audience Mismatch: The "chaotic" design with moving grids and bright yellow is overly complex for non-technical users (college coaches).
- UX Recommendations: Remove unnecessary animations to improve legibility; use darker yellow for buttons to ensure contrast.
- Structure: Suggest creating separate pages for specific personas (coaches vs. athletic directors) rather than forcing all information onto one long scrollable page.
- Trust Signals: Non-technical audiences require social proof (e.g., coach testimonials) rather than feature lists to establish trust.
- Sprites.ai: