How Stripe Built Their New Website

Predictions and Expectations:

• Katie Dill expects that with AI tools, design systems will evolve to be managed by AI, allowing organizations to scale design decisions much more quickly.
• Dill believes that the "gravitational pull" toward mediocrity is a constant risk in design and fears that accepting "good enough" solutions daily will cause companies to lose their ability to stand out as exquisite or exceptional.
• Dill predicts that designers will spend more time in the future thinking about new paradigms of interaction, specifically regarding agent experiences as users increasingly rely on agents to build businesses.
• Dill expects that AI will help raise the floor of design by creating baseline products (7 out of 10) quickly, freeing up time to focus on creating highly differentiated and great experiences rather than just shipping more average work.
• Dill believes that progress is probably more important than perfection for product success, especially when introducing new tools where user adoption and evolution are unknown until the product reaches the market.
• Dill predicts that the "Minimum Viable Quality Product (MVQP)" concept will become essential to prevent losing user trust while still allowing for learning from market experiences.
• Dill expects that the merging of design and engineering tools will allow teams to customize AI-generated outputs so they do not have to accept "slop" and can instead fight for the right solution.
• Dill believes that AI will facilitate the ability to explore more iterations, noting they can look at 20 ideas in the time it normally would have taken to look at two.
• Dill fears that if designers accept AI-generated outputs too easily, they may stop asking critical questions like whether the solution is really great, hits the mark, or achieves the goal with intentionality.
• Dill expects that as organizations grow, design systems will become increasingly critical to ensure coherency in the user experience when teams can no longer rely on small-group consensus.
• Dill believes that the future of design involves creating things that are exceedingly easy to use, powerful, and capable of bringing joy to the day, viewing joy as part of progress.
• Dill expects that the new design language will stand the test of time for potentially six years, similar to the previous site's longevity.

Timelines and Milestones:

• Dill states the previous homepage launched in 2020, which she notes was six years ago at the time of the recording.
• Dill mentions the team spent a year or more in painstaking design exploration to finalize the new homepage story and visuals.
• Dill plans to have the team continue to explore and iterate on the design language to ensure it remains relevant and does not feel "old-fashioned" in the future.
• Dill plans to have the founders walk the store every Friday in front of the whole company to instill culture and raise awareness of user experience issues.
• Dill notes the site was originally intended to launch in December but was delayed to January to ensure the animations were smooth and intentional.
• Dill expects the new design vision and language to remain valid for six years or whatever the lifespan of the site becomes.
• Dill plans to continue using AI prototyping to allow for rapid changes to copy and visuals for user testing during the development process.
• Dill expects that AI tools will evolve to help expand design systems to new use cases that require component growth, rather than just assembling existing parts.
• Dill expects that as more people within a company gain the ability to ship code, the need for AI-managed design systems will increase to maintain coherence.

Technology and Product Direction:

• Dill plans to use AI to generate custom-made images for the site that express both the Stripe brand and the brands of companies they serve, while maintaining the same level of care as non-AI assets.
• Dill plans to continue using AI to speed up prototyping and experimentation, allowing engineers and designers to mock up different varieties of interactions quickly.
• Dill plans to leverage AI tools to generate rough versions of sketches using components from the design system, which then require human refinement.
• Dill expects that AI will help speed up the process of checking details, such as ensuring shadows, hands, and layouts look real in generated imagery.
• Dill plans to integrate progressive disclosure (Bento boxes with modals) to provide more insight into products without taking users off the homepage.
• Dill plans to use delicate animations to express care and make the site feel alive, while ensuring they do not become distracting or annoying to users.
• Dill expects that the wave and gradient wave elements will continue to be a key way to bring color and joy to the experience for the next six years.
• Dill plans to continue fine-tuning the wave using a custom tool that allows for tweaking blur, grain, scale, rotation, texture, and color to find the perfect mix.
• Dill plans to explore new paradigms of interaction related to agents as users increasingly use them to navigate the internet and tools.
• Dill expects that AI tools will allow designers to push from baseline AI outputs to customized, high-quality solutions without needing to accept the initial result.
• Dill plans to use user research and testing to validate that the Bento box layout is a better choice for communicating product scale compared to accordions or section-by-section scrolling.
• Dill expects that AI tools will enable the team to change all the words in a prototype easily to feel out different narratives for different user segments during testing.

Market and Industry Outlook:

• Dill notes that over 78% of the Forbes AI 50 use Stripe, indicating that AI companies are a major part of their customer base.
• Dill expects the market to move toward agents where users do everything under the sun with AI, changing how they traverse the internet or their tools.
• Dill believes that websites can be fun again, contrasting the current era with a previous period where every SaaS website looked exactly the same.
• Dill predicts that AI tools will allow companies to move faster and reach good results almost instantaneously and almost for free, enabling a shift to creating interesting, playful, and beautiful designs.
• Dill expects that more companies will adopt the practice of "walking the store" to understand the composite user experience across different business areas.
• Dill expects that design systems will become essential for scaling organizations to ensure coherency in the user experience as teams grow beyond a small circle of decision-makers.
• Dill believes that AI will help scale things a lot more quickly in the future as these systems are managed by the AI tools.
• Dill expects that users will continue to scroll through websites and that designs must account for fast scrolling, requiring punctuation moments to make them slow down and take a note.
• Dill believes that users are increasingly becoming able to ship code and improve products, which will create a challenge for keeping a cohesive design language without direct oversight.
• Dill expects that AI-generated content will require human critique to ensure nuance and details are correct, such as ensuring hands and shadows look real.

Company Plans:

• Dill plans to refresh the visuals of the Stripe brand to match the sophistication of the business as it evolved to serve multinational enterprises and growing startups.
• Dill plans to adjust the narrative on the homepage to clearly articulate the number of things Stripe does, moving beyond the previous focus on just payments.
• Dill plans to focus on showing rather than telling by using imagery and modals instead of heavy text to communicate the scale of the product offering.
• Dill plans to keep users on the home page by using overlays and larger modals to provide more insight without taking them away from the journey.
• Dill plans to continue evolving the design over time, with a willingness to change elements if they start to not stand the test of time.
• Dill plans to create custom-made images for the site that express the brand of the companies Stripe serves as well as the Stripe brand.
• Dill plans to continue using the "Black Friday, Cyber Monday" city and podcast Cheeky Pint sections to show items that come and go, keeping the site changing and fun.
• Dill plans to maintain the dark mode for the integrations area (Squeezy Boy) to carry forward the developer language established in previous sites.
• Dill plans to continue the practice of "walking the store" as a requirement that everybody does as part of company culture.
• Dill plans to continue Essential Journeys with a scoreboard that tracks red, yellow, green status to monitor journey quality.
• Dill plans to encourage the team to fight the gravitational pull to mediocrity by not leaving well enough alone and hunting for the right solution.
• Dill plans to focus on the user's end experience when exploring products, ensuring they consider trade-offs that might negatively impact a user's day or erode trust.

Financial Guidance:

• No forward-looking financial guidance or specific monetary predictions are present in the transcript.

Risks and Caveats:

• Dill warns that if animations are too far, they become distracting and annoying, causing users to not want to engage.
• Dill warns that accepting "good enough" solutions every day will lead to a composite product that is just "meh" and prevents companies from standing out.
• Dill warns that AI does not replace craft, taste, or attention to detail, and teams must ensure each pixel gets the same amount of attention.
• Dill warns that pushing timelines out to the end of time is a habit to avoid, but waiting to do it right is sometimes necessary for a joyful experience.
• Dill warns that putting the most important messages a click away is tough because users are in a lean back, browse mode and won't click to find them.
• Dill warns that AI-generated images can feel unreal if nuance and details like hands, shadows, or layouts are not carefully critiqued.
• Dill warns that accepting slop from AI tools because it is fast and easy can lead to products that are not well-crafted.
• Dill warns that searching for perfection can be a dangerous pursuit because the product is only as good as what is actually shipped and available.
• Dill warns that losing sight of the user experience when building in isolation leads to disconnected products where light switches don't match across rooms.
• Dill warns that not thinking about how products intersect (e.g., subscriptions, payments, tax) will cause teams to miss the user experience.
• Dill warns that design systems can become too hard to manage if they are not AI-managed as the organization scales.
• Dill warns that users will lose trust if they experiment with something in the world that erodes trust without a minimum viable quality baseline.
• Dill warns that teams might not see the same details as others due to different disciplines or attention spans, requiring collaborative reviews to find dead ends.

Confidence and Disagreement:

• Dill is not sure that AI-generated content will replace craft but expects it to speed up prototyping and experimentation.
• Dill believes that Patrick was heavily involved in the decision-making process for the wave design and that he is exceptionally decisive.
• Dill expresses confidence that the new design vision will stand the test of time for six years or more.
• Dill expresses confidence that AI tools will help scale design systems and manage them more quickly in the future.
• Dill expresses confidence that walking the store is such an important part of building products and experiences that it should be a company-wide requirement.
• Dill expresses confidence that AI will help raise the floor of design by creating 7 out of 10 products quickly.
• Dill expresses confidence that delicate movements in animations will make the site feel alive without being annoying.
• Dill expresses confidence that progress is more important than perfection for product success.
• Dill expresses confidence that users will be able to understand the scale of Stripe's products through the Bento box layout.
• Dill expresses confidence that AI tools will help speed up the process of exploring 20 ideas in the time normally needed for two.
• Dill expresses confidence that design systems will be managed by AI in the future to help scale things.
• Dill expresses confidence that users will slow down at punctuation moments to take a note if the design is well-crafted.
• Dill expresses confidence that the team will not accept slop and will hunt for the right solution despite the ease of AI.
• Dill expresses confidence that AI tools will not replace craft but will assist in exploring more iterations.
• Dill expresses confidence that the new design language will not feel old-fashioned in a year's time.
• Dill expresses confidence that the team will continue to refine the wave until it feels right for the page.
• Dill expresses confidence that the team will not rush to get something out if it doesn't feel right for the overall composition.
• Dill expresses confidence that the team will not accept mediocrity and will fight for the right solution.
• Dill expresses confidence that the team will continue to use AI to mock up different interaction paradigms.
• Dill expresses confidence that the team will continue to walk the store to see the forest through the trees.
• Dill expresses confidence that the team will continue to explore new paradigms of interaction for agents.
• Dill expresses confidence that the team will continue to use AI to generate images that express the brand.
• Dill expresses confidence that the team will continue to use AI to speed up prototyping.
• Dill expresses confidence that the team will continue to use AI to generate rough versions.
• Dill expresses confidence that the team will continue to use AI to change words for user testing.
• Dill expresses confidence that the team will continue to use AI to speed up the process of exploring more.
• Dill expresses confidence that the team will continue to use AI to raise the floor.
• Dill expresses confidence that the team will continue to use AI to manage design systems.
• Dill expresses confidence that the team will continue to use AI to scale things.
• Dill expresses confidence that the team will continue to use AI to explore new paradigms.
• Dill expresses confidence that the team will continue to use AI to generate images.
• Dill expresses confidence that the team will continue to use AI to speed up prototyping.
• Dill expresses confidence that the team will continue to use AI to generate rough versions.
• Dill expresses confidence that the team will continue to use AI to change words for user testing.
• Dill expresses confidence that the team will continue to use AI to speed up the process of exploring more.
• Dill expresses confidence that the team