How Stripe operationalises design quality at scale

Executive overview

Most companies treat quality as optional — something to pursue after features ship. At Stripe, beauty and quality are treated as direct drivers of growth, not aesthetics for their own sake.

Katie Dill, Head of Design at Stripe, shares the frameworks she has used across Airbnb, Lyft, and Stripe to build quality into how teams work. The core mechanism: structured "walk the store" reviews of 15 critical user journeys, scored on a shared rubric, calibrated across product, engineering, and design leaders quarterly.

Beauty enhances functionality — and quality is growth, not its opposite.

Why beauty and quality matter for business

• Beautiful products increase trust: visible care in details signals care in details users cannot see.
• Penn Station vs. Grand Central: environment shapes how people feel, and feeling shapes behaviour.
• 99% of top e-commerce sites have errors in their checkout flows that hurt conversion.
• Stripe improved checkout quality and measured a 10.5% increase in merchant revenue.
• Quality improvements to onboarding directly lifted activation rates.
• "Quality is growth" — making a product easier to use drives adoption, retention, and word of mouth.
• The gym analogy: skipping quality today makes it easier to skip tomorrow; it compounds either way.

Levels of quality

• Level 1: Does it work? Does it deliver the basic value proposition?
• Level 2: Is it error-free and well-rounded?
• Level 3–5: Does it exceed expectations and do something users did not know to ask for?
• The right quality level depends on user expectations, not just internal ambition.
• Companies that treat quality as non-negotiable stop playing the "can we measure it?" game.

Operationalising quality: walk the store

• Stripe tracks 15 critical user journeys — the things that matter most to users and must be excellent.
• Each journey has named engineering, product, and design leaders accountable for its quality.
• Teams walk the journey as a user would — from Google search through docs, dashboard, and beyond.
• They friction log what they find: good moments, rough spots, bugs, inconsistencies.
• Journeys are scored on a rubric covering usability, utility, desirability, and "surprisingly great."
• Scores use a colour system (not numbers) to avoid false precision in a qualitative judgment.
• Teams meet quarterly in a Product Quality Review (PQR) to calibrate scores and share findings.
• Leaders across functions attend PQR: design, product, engineering, product marketing.
• The quarterly cadence allows enough time for material changes while staying frequent enough to catch regression.
• Walking the store also happens informally — Katie and CTO David Singleton do ad hoc walk-throughs regularly.

Quality as a group effort: key principles

• Quality degrades when teams are too focused on their own area to see how pieces fit together.
• Like renovating one room in a house: it makes every other room look worse.
• Vision and alignment matter: even the best individual contributors will diverge without a shared picture of the ideal.
• Editing is essential — someone (a "GC" or "architect") must see the whole and say "not yet."
• Courage to reject work that is "almost but not there" is one of the hardest leadership acts.
• Understanding quality requires understanding it from the user's journey, not in isolation.

Performance formula and team leadership

• Performance = Potential − Interference (learned at Airbnb).
• Increase potential: hire well, develop talent, help people grow.
• Decrease interference: remove blockers, fix broken processes, redesign org structure.
• At Lyft, designers sat behind a locked door, separated from engineering and product — leading to wasted work, misalignment, and slower cycles.
• Solution: break down the wall physically and structurally; embed designers with their engineering and product partners while preserving a shared creative space for crits and exploration.
• Teams wear two t-shirts: their discipline team (design) and their product team (marketplace, driver, rider, etc.).
• Every org redesign requires listening first — trust is the prerequisite for change.

Building awareness across large design teams

• Every few weeks, designers add a screenshot or prototype of their current work to a shared Google Slides deck.
• The deck is sent to design, product, engineering, and leadership — not as a status check, but as visibility.
• Benefits: surfaces overlapping work early, sparks cross-team collaboration, shows work in progress before it is too late to change course.
• Kept low-maintenance intentionally — currently monthly; previously bi-weekly.
• Key rule: label it clearly as work in progress, not a finished artefact.

Hiring designers

• Taste and character are harder to teach than tools and process — prioritise them.
• Look for a strong hit rate for great judgment, even in junior candidates.
• Humility is essential: it signals curiosity about users and willingness to collaborate.
• Hustle and courage matter: creation from a blank page is scary; the best designers fight for great anyway.
• For early-stage companies: combine a senior design advisor (for strategic lens) with a hands-on doer.

Reach for the stars, land on the moon

• Fear of bold ideas leads to incrementalism that never meaningfully improves the whole experience.
• Vision work is not about beautiful decks that go on a shelf — it is about knowing the ideal end state.
• Sketch the 11-star version of the experience; then step toward it with measured, sequenced bets.
• Without a north star, teams increment in different directions and never converge on something great.