Jeffrey Veen

On Coding Agents and the Future of Design

2026-01-15T00:00:00+00:00

In his seminal 2010 article on responsive design, Ethan Marcotte described a new way of thinking about web user interfaces that moved beyond a fixed frame. Reacting to his clients’ requests for an “iPhone version” of their websites, he suggested instead a single experience that could respond to the user’s device. Page elements could use a different set of design properties for large screens and small, transforming their form to match the capabilities of the hardware. Crucially, on simpler devices, elements could be hidden entirely if deemed non-essential.

As this technique became popular among designers and product teams, they quickly realized that it was more challenging to start with the “full” version and pare it down. Instead, by starting with the least capable device, you could boil the product offering down to its essence and present those primitives in the limited screen real estate available. This soon led to an acknowledgment of the irony of responsive design: That the least capable devices often had the most effective user experience, as they were most closely aligned with user needs.

We came to realize that responsive design wasn’t just about layouts, it was about forcing organizations to confront what actually mattered. But organizations, it turns out, cannot resist the temptation to use available UI space for promotions or expressions of their org chart, all of which almost always get in the way of what users are trying to do. That this still remains true all these years later is astounding; the difference in the UX of apps versus websites for banks, airlines, and ecommerce is stark evidence¹.

Towards a More Radical Responsiveness

I believe we’re at the beginning of a new era of further crystallization, and it feels both more disruptive and more enabling than the shift to responsive design 15 years ago. Earlier last year, foundation model providers released coding agents that developers could install and use directly from their command line terminals. These agents, including Claude Code and OpenAI Codex, introduced two seemingly small advances that would turn out to be incredibly significant.

First, they could iterate. Unlike earlier chatbot interfaces, these agents would continue to work and refine their output, looping over a plan until its goals had been achieved. Second, they were given access to tools. They could now call standard Unix utilities like curl for loading web pages or grep for searching file contents. A few months later, the underlying models were updated specifically to take advantage of this kind of tool use and to stay more focused on plan-following. With those pieces in place, developers on the frontier of AI-assisted engineering began sharing new workflows that felt genuinely different and everything sped up again.

Recently, many non-developers, myself included, have found that using Claude Code with files locally can be an incredibly effective way to get work done. My social feed is filled with people sharing their use cases: setting the agent to work on an Obsidian vault, managing email and calendars, finally getting value from smart home devices.

Primitives All the Way Down

The key to all of this is the tools: small, simple command-line apps with clear, direct documentation. Despite the promise of MCP servers, Claude Code with access to iCalBuddy is a much faster and more effective way to ask for help organizing next week’s schedule. Codex is an expert with gh, the CLI app for GitHub, and can quickly summarize and organize your next sprint’s outstanding issues. Both can quickly navigate through 1Password with the op command to avoid spewing access tokens everywhere². These apps so clearly expose the primitives of the systems behind them. An agent looping iteratively while stringing dozens of these composable tools together starts to feel like a very new way of working.

This is the next step in responsive design and the blueprints are lurking all around us. The coding agents now let you add skills — simple descriptions of how to accomplish a task written in natural language. You might explain to your agent how to get data from an internal API to help you create your quarterly update, for example. But if you really want a great look at what the future may hold, try this: on an iPhone or Mac, open the Shortcuts app and start searching through the available commands exposed by each app you have installed. There, you’ll find a remarkably clear visualization of all the atomic components of an app’s capabilities. All the nouns and verbs.

None of that works with agents quite yet³, but I think this is the clearest glimpse into what apps become in a world of advanced agentic workflows. Here in three screens are the actions we can take when communicating with friends, planning our vacations, and managing our money. This is what it looks like when apps are honest about what they can do. And a step further: all the APIs for all the enterprise SAAS and personal productivity apps, lurking behind developer documentation, waiting to be unlocked. What is a “seat” and what is a “user” when we send our agents to ask questions and make requests? This isn’t just app design, but what businesses become as we collectively lean harder into agents.

Clarity as Competitive Advantage

This may seem counterintuitive, but my career in user experience design makes me even more excited about a potential future like this. If agentic workflows strip back applications to their bare essence, then it makes sense that they’d present what they find in UI components crafted on the fly. Perfect customization, whether you’re accessing the capability on a large display or through your AirPods, in English or Cantonese, wrapped in shadcn or Chakra UI — your choice!

An agentic future elevates design into pure strategy, which is what the best designers have wanted all along. Crafting a great user experience is impossible if the way in which the business expresses its capabilities is muddied, vague or deceptive. So the best have always pushed for more authority, more access to the conversations where the real decisions are being made in their organizations. To gain that access while maintaining advocacy for humans using the products has always been my goal as a designer.

This takes a pretty big leap of faith for those practicing design. I see the same happening for engineers waking up to the fact that they’ll soon stop typing code, and not long after that, stop reading it. Many of us are imagining futures in which aspects of the labor we love are quickly slipping away. But I’ll argue it’s what we’ve been after all along: Responsive designs that mold to our customers and patients and citizens in ways that ask businesses and institutions to express to us exactly what they have to offer.

If an agent used your product tomorrow, what truths would it uncover about your organization?

I have known Ethan for a very long time and have much admiration of his work and the powerful and positive influence he’s had on the design community. I’m also aware of Ethan’s clear and reasonable commentary on the ethical implications and environmental impacts of LLMs. Citing his work here is done with the utmost respect. ↩
Let’s not discount the security implications here. They are dramatic. But we need use cases to solve for, rather than blanket bans. ↩
You might not find better proof of Apple’s missteps with Apple Intelligence. The raw materials are all there; they’ve already shipped the hardest part. ↩

Small Acts Build Great Cultures

2025-11-11T00:00:00+00:00

This morning as I walked to the cafe my phone alerted me that I’d left my luggage at home. This and other bugs have been happening off and on since I upgraded to iOS 26. My AirTags are randomly forgetting where they are. Some of my friends are showing up as mere phone numbers in iMessage. Nothing terribly bad, just a string of little paper cuts.

It reinforces my belief that teams need a culture that values attention to detail when building products. Tiny annoyances so often get neglected as we rush to ship, but the consequences accumulate, souring the whole brand. It’s not a long journey from “Ugh, these AirTags…” to “Apple has lost their way…”

But in my experience, those rough edges seldom go unnoticed by someone, somewhere, who was unable to stop the momentum of a product release for such an “insignificant” flaw. Or, even more consequentially, they did not feel it was safe to do so.

This resonates with the concept of jidoka — a principle from Toyota’s legendary production process that gave a switch to every worker allowing them to stop the assembly line if they noticed a problem. The innovation was not that they wired the factory to be controlled this way, it was that everyone on the team had a shared sense of quality. Each worker had a constant reminder that the organization trusted their judgement.

This is where your values become tangible. We can talk about believing in quality, caring about users, or the psychological safety of our teams. But how does that boil down to a junior designer filing a P0 bug the night before launch?

I’ve always felt that culture is made of the accumulation of small acts of gracious leadership: acknowledging moments of bravery during a retro, teasing out a reticent comment during a product review, and on and on. It can come from other places too, but it is most effective when it comes from the top.

If you’re leading a team remember: Never criminalize pride in craft.

Vibing with Metcalfe’s Law

2025-10-27T00:00:00+00:00

Earlier this year, we brought the entrepreneurs from our portfolio companies together at Founder Camp, a gathering designed around the premise of the ‘unconference’: that the audience will connect and self-organize into conversations that are most relevant to them. More facilitation, less programming. For startup founders, we’ve found this is intensely valuable. The nature of what they’re struggling with is, by definition, out ahead of where conventional advice can help. The best solution is to connect with others who are at the coalface with them.

I hosted the welcome session, and wanted to illustrate the value of these connections from a first principles perspective. As I thought about how I could do this, I remembered Metcalfe’s Law — a way of calculating the value of network effects named after Bob Metcalfe, the creator of Ethernet in the very early days of the internet. The idea is that as a network gets larger, the value grows proportional to the square of the number of nodes. A telephone network with a single phone isn’t very useful, but as you add devices it becomes essential. Same holds for people on a social network, or in this case, founders in a hotel ballroom¹. I wanted to show the math: the 209 people in the room represented over 21,736 connections. I told them, “If you’re struggling with a seemingly intractable problem, odds are high that the answer is in the room.”

Feeling the algorithm

Two days earlier, on the flight to the event, I sat looking at the slide I would use to explain this. I was using the image from the Wikipedia page: telephones connected in a graph diagram. It was accurate but uninspiring. What I really wanted to drive home was that value grows much faster than our intuition would have us believe. Perhaps an animation would help us feel how it expands, rather than grappling intellectually with a formula. I could show the network growing like a time lapse of a plant unfurling from a seed. I started playing with a more abstract version of the visualization and quickly grew frustrated. I could see this ending up with hundreds of elements on the slide, each needing to be tediously positioned and animated by hand. This needed to be generated, not drawn.

I left Keynote for Claude. I gave it the Wikipedia diagram and described how I envisioned the animation working, emphasizing that the output would need to be embeddable in a presentation. It gave me a simple web app that output a movie. A few minutes later I had a crude start. We iterated a dozen times in about 30 minutes and ended with the animation I had envisioned. There was applause when the video played onstage.

Vibe coding is a design superpower

The experience I had preparing for Founder Camp was a step towards working in a very different way.

First, it’s a good marker for how fast things are changing. That was in March. We’re now in October and it already feels like my vibe coding story comes from a different era. Today, as I write this, multiple Codex agents work in tiled windows on my desktop, refactoring one of my projects. I would say that I am struck by how far we’ve come, but that feeling hits me weekly now. The model releases of the past few months have been an astonishing leap forward in the efficacy of code generation, design prototyping, and UX of increasing powerful agents. I made the animation in a Sonnet 3.5 chat window, setting everything up by hand in the terminal and pasting back and forth as we iterated. That feels like an ancient way of working but it was only seven months ago!

And second, this is just such a great example of how the changing nature of our tools can unlock our ambition for what’s possible now. I’ve spent a career working under constraints, always able to imagine far more than my teams could realistically develop and running up against budgets. Even more so, I’ve always struggled to ask developers to abandon their code because a feature I asked for wasn’t quite right. But this experience, and others since, has me thinking about tools in a very different way now. Our abilities are compounding at a rate that feels even faster than what my diagram illustrated, and traditional software isn’t keeping up.

The tools we can summon are limited only by our imagination.

The Answer is in the Room

Back in that welcome session at Founder Camp, after showing this animation, I asked everyone to think of their current intractable problem. Then I asked them to stand and move through the room, physically traversing the connection points in that diagram, looking for the node that could help. It worked so well they wouldn’t stop, and I had to abandon most of the rest of my talk. Which was fine. There was something bigger than the sum of the people in the room or the quadratic growth of the network. We were creating community and connections and sense that the often fraught entrepreneurial journey is not something you need to travel alone.

I probably would have abandoned the Metcalfe’s Law point had I not been able to visualize it in a way that made it come alive. It sparked the idea to have the crowd act it out, and that led to this wonderful moment. With the wind of AI at my back, I could dissolve long-held constraints and be more ambitious. And that led me to not just be more productive, but to find entirely new ideas.

Try the app I made! It’s online at https://veen.com/metcalfe-generator/

I am generalizing. Metcalfe’s Law counts potential connections and assumes equal edge value. Networks of people have uneven ties: knowledge, power, trust, distance all weigh on relationships. Normalizing these are why we did this exercise in the first place. ↩

30 Years of Building

2025-10-10T00:00:00+00:00

You can open up a terminal on your computer and type the following at the command prompt:

$ whois veen.com

Your terminal will fill with a few pages of mostly-redacted information. This command queries a database of publicly available domain registrations. Years ago, you’d see actual email addresses and phone numbers for the people who owned and managed domain names across the internet. But in today’s world, all of that is hidden.

However, if you pick through the whois output, you’ll still find these lines:

Domain Name: VEEN.COM
Creation Date: 1994-12-02T05:00:00Z

Thirty years! I missed the big anniversary in the year-end bustle last December, but it feels like quite a milestone. I remember registering the domain when I was building webpages at Wired Magazine. Back then, you needed to send email directly to the InterNIC, with an explanation of what you intended to do with the address. Thankfully, they granted me the URL, and I set up an old Mac IIci in the server closet to serve a simple site. Eventually, I started using early blogging software and started updating the site regularly.

Skip ahead to the 2010s. For many reasons, I let the site sit. I was writing on the Typekit blog; I wrote on Medium for a spell. Then I mostly went quiet in public as part of general retreat from social media, which was healthy and refreshing for me.

Recently, GitHub started sending me Dependabot alerts for this site’s repo with the forbidding message, “Known security vulnerabilities detected.” In the past, fixing something like this would have been just enough effort to ignore. I’d have to dig through the code to remember how everything worked, and what I was thinking back in 2012 when I last updated the site. But we live in the future now, so I pulled the repo down, asked Claude to update the libraries, and it happily did so, modernizing the CSS and making everything more accessible to boot.

But before I checked it back in, I paused. Why not start posting again? And in doing so, why not make it less precious? Looking through some of the older posts, I can see that was only posting if something Very Important had happened, like the launch of something new or a change in my career.

But something big has been happening lately and I thought maybe I could just clean up all these notes I’m keeping for myself and start dropping them here. And the thing that has been happening is rekindling my love of building things. You can see it here:

That’s a Github heat map showing commits I’ve made over the last 12 months. During that time I have been steadily shipping code that actually works, with coding agents that get more powerful seemingly every day. Some of this code is just for me: Codex chewing through a directory of PDFs with test results from my doctor to build a dashboard for me. Other projects are work related: a system for extracting co-investors and valuations from archived cap tables at True.

What a revelation that has been. After a long career in product and design, most of which was spent in management and leadership, it feels incredibly empowering to just have an idea, kick that idea around with a virtual developer who has read the entire internet, and get the code back in a few minutes. Test, iterate, think, repeat. The dopamine rush of shipping happening multiple times a day.

Thirty years ago, on 1994-12-02T05:00:00Z, I felt like something big was happening. The Web was blossoming in front of our eyes. The future seemed limitless. There are similar vibes today. I’ve been quietly exploring the new future, and I think I’ll share a bit about that here.

With luck, I might get another 30 years out of this domain.

2015-02-05T00:00:00+00:00

Recently I had the privilege of being interviewed by Om Malik at his Roadmap Conference. We were talking about about design in general and Adobe in particular, and he asked me about the project I had completed earlier in the year, in which we redesigned the entire Adobe.com experience from the ground up. It was an almost absurd undertaking, and Om rightly pointed out what had been at stake. “How do you convince a big company like that to make such a big change? Isn’t design risky?”

I thought for a moment, and then told him I disagreed:

“Design isn’t risky – change is. Good design mitigates that risk.”

I’ve felt this way for a long time. It was one of our core values 14 years ago when we founded Adaptive Path. More recently, it was part of our pitch when we started working on Typekit. And it’s been at the core of the transformation we’ve done at Adobe. Now I’ll have the opportunity to practice these values in a much different venue: As of this week, I’m leaving Adobe and joining True Ventures as a Design Partner.

In my new role, I’ll be focusing on helping True’s portfolio companies cultivate the best product and design work they possibly can, and to seek out new companies who have that same passion. To that end, I’ll be taking advisory roles with a variety of companies, including a wonderful opportunity to serve in a temporary capacity to guide product design and strategy with the amazing team at about.me.

The Internet has matured into an age of design. We spent the ’90s getting the technology to work, and the following decade mainstreaming and scaling this magical new platform. But as we’ve slowly climbed up our version of Maslow’s Hierarchy, we’re reaching a point where the overall experience someone has with a product is quickly becoming the biggest competitive advantage a company has. John Gruber expressed this change nicely in his coverage of Apple’s record-breaking revenue numbers this week: “[D]esign can matter in the mass market. For decades the industry’s conventional wisdom held that design wasn’t important. The industry’s leaders created shitty software and shitty hardware. Apple’s success has upended the industry’s value system.”

That value system has always interested me. In my work, I’ve focused on putting human needs ahead of technical achievement; empathy ahead of processing power. Working in venture capital with True will give me the ability to practice that more broadly, and at a more influential time in a company’s growth. Working intensely with a company like about.me will let me hone my skills more deeply. It should make for a rewarding balance.

Last week, I said goodbye to my team at Typekit.

Yes, I’m excited about my new opportunities, but they come with bittersweet emotions. I’ve made some amazing friends at Typekit, some of whom I’ve worked with with since the product’s inception. Together, we grew the germ of an idea into an astonishingly large service, serving billions of fonts to millions of web sites. We’ve had a significant impact on how the web looks and works, and provided solid revenue for the craftspeople who shape the world’s type. But what I’ll miss the most are the bonds forged between all of us and expressed through the culture of trust and respect. I’m proud of our work together, and confident they’ll continue to do amazing things at Adobe.

As one chapter closes, another opens. My new job will give me the opportunity to contribute from a different perspective than the entrepreneurial world I’ve been in.

Still, I feel like I’ll be doing the same thing I’ve always been trying to achieve: Figuring out what people truly need and presenting a solution that satisfies those needs as clearly as possible. It’s as relevant now as it ever has been.

Building Typekit on Relationships

2011-11-17T00:00:00+00:00

There are lots of reasons to do a startup. Sometimes, there’s an idea you just can’t stop thinking about – a thing you absolutely want to exist in the world. Or sometimes you see a gap in how an industry is evolving, and with a small team of talented people you think you can fill that gap much faster than the big companies can.

Both of these things were true when we started Typekit three years ago. Web browsers started implementing @font-face and a lot of people started wringing their hands over the issue of intellectual property and typefaces. A debate sprang up: Web designers were embracing new CSS features like never before, but font designers worried that their craft would go the way of Napster and BitTorrent. It was a recipe for disruption and opportunity, and we jumped in.

For us, so much was uncharted. The four of us who founded the company had worked together and built products before, but we’d never been down the venture capital path. We scratched up a little cash putting on a conference and doing some work with our friends at Twitter. Still, it was clear that if we wanted to build Typekit at the scale we imagined, we’d need some real money.

It takes a team

Cloud computing has really made it inexpensive to try out new ideas. I mean, seriously, it’s crazy cheap now. Case in point: when we started Measure Map in 2005, the table stakes were about $25,000 for a rack of Dell servers just to see if it would work. The machines were shipped to us and we took them to a data center and opened all the boxes and plugged them in and spent a few days configuring everything. Six years later, that feels as antiquated as starting your car by turning a crank.

So yes, it’s true that cloud computing has changed the startup world. It’s cheap to build stuff now, and that makes things less risky. But a great team costs money, and that hasn’t changed – nor should it. Experienced designers and engineers can’t pay their kids’ tuition for 10 cents an hour or spin down health insurance when it’s not utilized.

You need a team. I don’t believe you can do it all yourself, or even with a co-founder camped out in your parents’ garage. Examples to the contrary are the exception. It’s rare that one person can write all the code, craft an exceptional user experience, communicate transparently with customers, and manage the financial health of the business. Yes, you can build tangible proof of your idea. You can even launch it and get traction. But to really build something that has broad reach and significant impact, most of us need the diversity of talent and experience that comes with a team of collaborators.

So when we started Typekit, we did some math. Given our track record and how much we believed in the idea of fonts on the web, how much of the theoretical future value of the company should we take in advance? With that up-front capital, could we actually put together the team we wanted and move fast enough to get a compelling product to an audience that would pay for it? And even if we could, would it generate enough revenue to both share with our partners (since we didn’t actually have any fonts for our font service) and grow into a sustainable business?

When I say “math” above, I run the risk of implying some rational process. Sometimes, I wished there was a simple algorithm or an actuary table for startups – plug in your numbers and out comes the answer. Unfortunately, it’s nothing like that. Financing a startup has as much to do with timing as it does with the track record of the founders or the attractiveness of the idea. The first browsers implemented CSS font linking just around the same time as the 2008 financial meltdown. Was that good for us? Was that bad? Tough call, but it certainly was relevant. We iterated our business model as much as we did our interface.

There are other models for getting started, of course. We bootstrapped Measure Map with the profits of Adaptive Path’s vibrant consulting practice, following the model we saw 37signals forging. When you’re a consultant, you trade your time for money and once you’ve done the work, you can pretty much do whatever you want with the cash. But it’s distracting. For me, doing my best possible work requires complete focus on a single problem. I couldn’t solve other people’s problems to make money and build a great product at the same time. To do that, I needed – again! – a great team. So at Adaptive Path, some of the people earned the cash, while others burned through it building a product using – again! – the same sort of undocumented math as a VC-backed startup.

Ultimately, we did what might be called a traditional round of funding for Typekit: A VC firm led the round that included a number of angel investors.

A True story

Our lead investor was Tony Conrad. I first met Tony when he was one of Adaptive Path’s clients a bunch of years ago. He was on the entrepreneur side of things back then as the CEO of Sphere, and was looking for help with visualizing his product concept through Adaptive Path’s user-centered design process. This was during the time I was transitioning to Google, so we didn’t get the chance to do day-to-day work together, but we did spend a lot of time talking about the product and how he was designing the business to support it.

It was immediately clear that Tony put a lot of work into surrounding himself with very talented people, regardless of whether he could convince them to actually take a job at his company. In every project he’s taken on, I’ve watched him build up a network of advisors, and then trust what they would tell him. Later, when he launched about.me, the product’s initial growth was driven through Twitter. His group of advisors collectively had millions of followers to help him get the word out.

It was through Tony that we meet everyone else at True Ventures. He’s a partner there now, and when I left Google and started thinking about doing something with Bryan, True was our first call. They were one of the early VC funds, about five years ago, to focus exclusively on seed-stage companies – small teams that haven’t taken a formal round of financing yet. These companies are the riskiest investments since the team usually hasn’t even started building their idea. True focuses not just on giving them money to give it a go, but reducing the overall risk by attempting to eliminate as many of the factors that tank a company in the first year.

Specifically, True puts a tremendous amount of effort into the community of founders with whom they work. They host frequent events, bring in experienced and inspiratiional speakers, and help the entrepreneurs they’ve funded connect with each other. We knew we could pick up the phone and talk not only to the partners, but to any of the other founders and we’d find someone who’d been through whatever we were facing that day.

The most interesting dinner party ever

I may have lots of experience developing products, but navigating the ambiguity of a startup is a very different set of skills. I learned while working on Typekit that both the product and the business benefit from a user-centered philosophy. When you’re stuck with a problem and don’t know what to do next, talk to people. Ask them what they think. Collect a bunch of perspectives. Filter what they say through your own intuition and experience. Then ask them who else you should ask. Keep going and never stop. To put it simply: Great products are built on solid relationships.

So we took Tony’s advice and surrounded ourselves with people we respected. And it turns out that the best way to get someone’s continued attention is to take their money. We set aside a portion of our first round equity for a group of angel investors who, in Bryan’s words, “Would make the most interesting dinner party ever.” And, as soon as we’d got everyone signed on, we sat down for dinner with Evan Williams, Caterina Fake, Matt Mullenweg, Chris Sacca, Josh Felzer, David Samuel, and the legendary Ron Conway. Some of them I’d known for years, like Ev and Caterina who had both been clients at Adaptive Path. Matt became a friend back while I was working on getting Measure Map integrated with WordPress. I worked with Chris at Google and he helped me to think about how things can get really, really big. But Josh, David, and Ron were new relationships we developed and I’m grateful we did. Everyone’s combined experience – and their willingness to share it – was both a tremendous head start and a reassuring saftey net for us.

My only regret in this whole funding process was that we weren’t able to work with twice as many people as did. We had so many great meetings and offers for support. In the end, we had to find a balance between how much of the company we were willing to part with and the number of investors who would end up with a reasonable stake. It was amazing that so many people believed in us, but it was difficult to make the final decision.

If I were an advice-giving guy, I’d tell entrepreneurs this: Raising money for your startup is not about the money. It’s about finding people to work with that you like and trust. Someone you wouldn’t hesitate to ask anything, and won’t make you feel stupid when you do. Use this test: Imagine your phone ringing; does the name on the screen make you feel eager to answer? Not nervous. Not dreading the call. Excited.

If you do go down the VC path, choose wisely. Your investors are one of the most important hiring decisions you’ll make.

Video: Designing for Big Data

2009-04-21T00:00:00+00:00

</param></param></param></embed>

This is a 20-minute talk I gave at the Web2.0 Expo in San Francisco a couple weeks ago. In it, I describe two trends: how we’re shifting as a culture from consumers to participants, and how technology has enabled massive amounts of data to be recorded, stored, and analyzed. Putting those things together has resulted in some fascinating innovations that echo data visualization work that’s been happening for centuries.

I’ve given this talk a few times now, but this particular delivery really went well. Only having 20 minutes forced me to really stay focus, and the large audience was very engaged. I’ll be giving an extended version of this talk in June at the UX London conference, with a deeper look at how we integrated design and research while I was at Google.

Wikirank: Tracking what's popular on Wikipedia

2009-03-26T00:00:00+00:00

A few months ago, the four of us in Small Batch Inc were kicking around ideas for what we should build next. We had just launched the Twitter Election site – a fun real-time data visualization project – and wanted to keep building things that help people make sense of the enormous amount of information that bombards us each day. Late one night soon after, I followed a link to a repository of Wikipedia server logs. There were gigabytes of data sitting there just begging to be visualized. We got to work.

The result is Wikirank, a tool for exploring what’s popular on Wikipedia, discovering comparisons between topics, and sharing them with the world. We launched it yesterday, happily coinciding with the 14th anniversary with Ward Cunningham’s invention of the wiki.

There are a bunch of reasons why I think Wikirank is cool, but my favorite is how it helps people find stories in the data. One of the great things about the web is how measuring tiny behaviors reveals patterns that tell stories. The data we get from Wikipedia is no different; as we started playing around with the numbers, we saw loads of interesting shapes emerge in the charts.

For example, big news stories show up as dramatic spikes where there there was no data before. When the astounding story broke that an airliner had made an emergency landing in the Hudson, a page was created on Wikipedia within minutes. Over the next two days, that page was one of the most popular on the site.

Comparisons are where Wikirank really shines, however. The weekly viewing habits of television watchers comes into clear view – as do the day on which the shows air – when we compare Heroes to Lost. (It’s also my guess that the occasionally perplexing plotlines of both of those shows leads a fair number of people to Wikipedia to find out what the heck just happened.)

I’m really pleased with how this project came out. I’ll write more about the technology behind the project in a followup post, but the reality is this launch represents a pretty big shift in how we build web apps. With only a couple developers and a rack of rented machines in the cloud, we pulled this all together in just a few weeks. That simply wasn’t possible the last time we did this.

The team behind all of this includes my long-time friends and business partners Bryan Mason, Ryan Carver, and Greg Veen. We were also extremely fortunate to be able to work with Dan Cederholm of Simplebits, who helped us with visual design and identity. Dan rocks.

Update: On March 10, 2010, Wikirank was acquired by Thing Labs, Inc. The were, in turn, acquired by AOL later that year. Priorities change when little companies are subsumed by big companies and, sadly, Wikirank has been retired.

Conference hack: Embracing the backchannel at Start

2008-07-25T00:00:00+00:00

I had a very interesting experience a few months ago while participating in a panel discussion. Once again, I realized that the content on stage is merely the spark of a broader conversation, and that the backchannel is rapidly becoming the whole point. So we’ve decided to try an experiment at the Start Conference in a couple weeks to see how we might hack traditional presentations.

But let me back up a bit first. I was on a panel at this year’s South by Southwest talking about the role of analytics in design. With me were two veterans of the advertising industry who’s work included some of the biggest ad campaigns of the past few years - some really amazing stuff. I started the conversation by saying how the remarkable amount of audience data available to us gives designers tremendous power to affect user experiences. My collegues suggested my approach sucked the creativity out of design. I countered that they were mistaking preferential research from behavioral. The argument heated up.

While this was happening, my phone was buzzing non-stop. I slipped it out of my pocket to discretely turn it off, but noticed a stream of Twitters going by - many from audience members in the room. So I set the phone down on the table in front of me and kept an eye on it. I’m so glad I did.

As the conversation on stage continued, the stream of questions and comments from the audience intensified. I changed my tactics based on what I saw. I asked questions the audience was asking, and I immediately felt the tenor of the room shift towards my favor. It felt a bit like cheating on an exam.

I guess it really wasn’t cheating, but it does illustrate one of the frustrations I’ve had at conferences lately. Most of the events I attend have a rich conversation happening in the room, yet the only people not able to participate are those on stage. A couple times, I’ve seen organizers project a live IRC channel, but that usually bring out the worst in people (“First!!!111”) - and is terribly distracting. So I’ve been wondering for a while if there was something smart we could do at our conference.

Apparently, Bryan had the same idea. As we were planning Start, he said, “We should have someone onstage the whole time to represent the audience. Like an ombudsman does for a newspaper.”

So we decided to put a desk on stage and have our friend George Oates fill that roll. She’ll be on Twitter, IM, and email listening to what people are talking about. (We’ll also have volunteers collecting index cards for those not wanting to be online during the sessions.) And she’ll synthesize questions, interrupt us if we get boring, and call bullshit if something sounds like it. George has been the designer at Flickr since it started; her personality has always shown through there, and will be a great fit for what we’re trying to do.

What do you think? I’ve never really seen a conference do something like this before. Have you?

Who is John Snow?

2008-07-07T00:00:00+00:00

Last month, while attending the fantastic @Media conference in London, we paid a quick touristy visit to a street corner in the heart of Soho. There, we found a small monument - an old hand-drawn water pump on a stone pedestal, its handle notably missing. We climbed up on it, had our photo taken, then ducked across the street for a pint in an old pub named The John Snow.

The pub’s name was interesting to me for a couple reasons. I had just finished the Steven Johnson book, “The Ghost Map,” a riveting account of the neighborhood’s 1854 cholera outbreak. Johnson tells the story of Dr. John Snow, who had lived near where we had our picture taken, and how he followed a hunch that the disease could be spread by water, rather than being airborne. After days of investigation, Snow proved to London’s health commission that this Broad Street well was the source of the outbreak, convincing them to remove the handle. It was an early instance of science and evidence winning against superstition and prejudice.

But even more interesting to me was one of the tools he used in his work - a data visualization showing cholera deaths marked on a map correlated to the location of the pump. In the speech I gave that week, I pointed out how his graphic techniques are still a model for good design over 150 years later. Contrary to the popular myth, Snow didn’t create that famous map; he was working from an existing visualization first drawn by sanitation engineer Edmund Cooper months earlier. Snow did, however, redesign the work by carefully eliminating elements and emphasizing the casualty data until the narrative of what had happened was perfectly clear. You can’t look at that map without thinking, “Yikes! Stay away from that pump!” Snow’s early geo-mashup can still inspire.

So back to the eponymous pub. As we stood near the crowded bar, I decided to do a little experiment. Even though there was a plaque on the wall detailing the establishment’s history and namesake, I started asking the folks around me, “Hey, why is this place called ‘The John Snow’?” hoping to find fans of the historic story - or maybe even someone interested in the visualization. Instead, here’s what I got:

The tall Irish guy with a ponytail: “No bloody idea, brother.”

The group of women laughing in the corner: “Oh, he’s some wanker on the telly who bought this place years ago.” (They were thinking of this Jon Snow, who is actually quite respected.)

The bartender, who came close, but really should know better: “Before this was a pub, there was a hospital here and he was the doctor in charge of it.”

How quickly history fades - especially after a few drinks.

Stories we'll hear at the Start Conference

2008-06-24T00:00:00+00:00

One of the nice things about organizing your own conference is getting to choose who you want on stage. That’s been particularly fun these last couple weeks as Bryan and I have been thinking about who we’d like to talk to and what we’d like them to tell us.

We started with some good friends whom we’ve worked with and respect. Ev Williams, for example, was an easy choice. We’ve known him forever, and worked on a redesign of Blogger together a few years ago. We’re talking to him about how his new thing, Twitter, is growing and evolving, what the challenges and opportunities are, and the lessons he’s learned so far. I’ve had similar conversations recently with both Matt Mullengweg of WordPress and Mena Trott of 6apart. Their stories have interesting parallels, but even better divergence.

So, yeah, a lot of our friends are in the blogging world. But we do know a few people outside of that world. Julie Davidson and Narendra Rocherolle, for example, are joining us to talk about their unique partnership, both in business and in life. They’ve built Webshots, one of the first photo sharing apps, as well as 30boxes, a fantastic calendar app that stands as a fantastic example of how to do the Right Thing when building on the web.

On the other end of the entrepreneurial spectrum is our pal Merlin Mann. He’s built a solid brand out of his productivity site 43Folders and we’re looking forward to talking to him about how he did that. But also, I’m particularly interested in his ideas for how small teams can find the best ways to work together and communicate with each other. Also, he is a bit, um, energetic on stage. We should have a good time.

We’ve got a bunch more people joining us on stage - people who follow the industry, comment on it, invest in it, and help companies get started. I’ll post more on them soon. I hope you can join us.

Introducing the Start Conference

2008-06-11T00:00:00+00:00

When I recently decided to leave my job, I began a little personal project I called “100 Lunches.” I wanted to sit down with a bunch of people over a meal and just listen. What was interesting to them these days? What where they building? Were they nervous about the state of the industry? Excited by all the new opportunity?

As I started the project, I quickly realized that the stories I was hearing needed a wider audience. My friends and colleagues told tales of late nights, harrowing server crashes, exhilarating growth, and touching emails from their users. Not one of them knew what they were getting into, nor would any of them change their paths.

Wouldn’t it be great, I thought, if everyone could hear these amazing start up stories?

So I got together with my long-time friend and business partner Bryan Mason and started planning. We decided to get a great group of entrepreneurs, investors, and other people associated with new web companies together for a day of discussion and advice. And that’s how the Start Conference came to be.

So here’s the deal: On August 7, we’ll spend the day at the beautiful Fort Mason Center talking about starting companies. We’ve got a lot of fun things planned, including a fantastic party on the Bay afterwards. And we’re making it cheap - just $200. We didn’t want to turn this into a conference filled with investors hunting for the next big thing. We wanted something for people who make web apps and dream of starting their own thing. (Also, we understand that you probably won’t want to submit an expense report for a conference that encourages you to quit your job.)

If you’ve ever thought of staking out on your own, we hope you’ll join us. It’s going to be great fun.

_All the details are on the Start Conference web site. If you have any questions, let us know in the comments here.

Charles Joseph Minard's visual stories

2008-05-14T00:00:00+00:00

You may not be familiar with the name Charles Joseph Minard, but it’s likely you’ve seen his work. He served as a civil engineer in 19th century France and developed an interest in cartography later in life. In particular, he was intrigued with showing variable data on maps – how quantities of shipped goods moved along waterways, for example, and later troop movements in military maneuvers.

Minard started drawing in his mid 60s and didn’t create his most famous work until he was 80 years old. In it, he shows the progression of Napoleon and his army to Moscow and back in the campaign of 1812. This chart was renown as a masterpiece of economy and insight, and imortalized in Edward Tufte’s “The Visual Display of Quantitative Information” – a virtual bible for information design since its publication 26 years ago. Tufte even sells the chart in poster form from his web site, making it a staple in designers’ cubes at nearly every creative office I’ve ever visited.

So I was researching Minard a bit more deeply for a presentation I’ve been giving and came across the following quote. In it, he talks about his “carte figuratives” – a phrase referring to his particular style of data visualization mixed with geography:

The aim of my carte figurative is … to convey promptly to the eye the relation not given quickly by numbers requiring mental calculation.”

I really like how this idea of making something visually apparent while reducing intellectual work – a sort of cognitive ergonomics. Minard was suggesting that stories and meaning can be found in any collection of data. It’s up to us to uncover those stories and tell them clearly and accurately.

It’s a strong lesson for much of the work we do today. Much like the patterns in data visualization, designers seek out stories and meaning when crafting interfaces. Well designed sites “convey promptly to the eye” what’s possible, while doing so intuitively as to avoid “requiring mental calculation.” That’s not to say we should treat people as stupid. Rather, we should help them focus on what they’re trying to do, rather than struggle with the means for achieving it.

Or, another way, don’t make me think.

Leaving Google

2008-05-02T00:00:00+00:00

Today marks my last day at Google and an end to a truly remarkable period in my life. If feels like just yesterday we were signing the papers, telling our friends, and announcing to the world that our team had been acquired.

We were nervous, of course. Acquisitions can be tricky - different cultures, different values, different technology platforms can all conspire against successfully merging companies. But moving to Google couldn’t have been a better fit. We immediately jumped into the Analytics team and started working with them on a redesign of their product. Their openness to rethink every aspect of the app still amazes me. It’s one of the most meaningful professional collaborations I’ve experienced in my career.

I also had the opportunity to work with Google’s User Experience Team - a shockingly talented and effective group of designers and researchers. I wish I could talk about all the amazing things they’re working on; all the ways they’re changing how we’ll connect to the world’s information and to each other.

The decision to leave was a tough one. Google clearly is an amazing company to work for. After consulting with many companies during my time at Adaptive Path, it’s clear that Google is like no other: they move fast, think clearly, and push strategic decisions out to the people closest to their users. But in my career, I’ve always swung between the big and the small and it’s time for another shift.

So what’s next for me? I’ve got a couple of small projects in the works, but mostly I’m going to take a little break, travel a bit, and catch up on some serious miles on my bike. It’s been a crazy couple years … I could use a nap.

An introduction to 'Mental Models'

2008-02-15T00:00:00+00:00

Congrats to Indi Young on the publication of her new book, “Mental Models: Aligning Design Strategy with Human Behavior.” Indi and I were partners at Adaptive Path and worked together for a bunch of years. I’m thrilled to see her insights into research and design process captured in such an accessible and informative book. And I’m honored she asked me to write the forward, which I’ve included below.

“You’re researching all the creativity out of this project!”

I can’t tell you how many times I’ve heard designers, developers, and even business owners say this. It usually comes just after a project has begun, as I’m preparing for interviews with users. Designers just want to start designing, developers want to start writing code, managers want the thing to ship - so why are we spending all this time talking? And all this stuff just seems so obvious. Do we really need users to tell us what we already know?

I try to be diplomatic. “Maybe a few interviews now will save us lots of grief later,” I tell them. “Think of this as insurance: Let’s make sure we’ve got the basics right before we’ve designed everything and written all the code.”

But no matter what I say to convince a team to do research early in their project, I never let them know my dirty little secret: I used to be just as skeptical as them.

I’ve always believed in a user-centered design methodology. Even early in my career, when I was journalist, we always started with the mantra “know your audience.” Later in my career, I’d go to conferences and watch presentations with process diagrams - boxes representing users needs with arrows pointing to boxes representing product requirements. Intellectually, I agreed. But when I started a new project, in that intoxicating first stage when anything is possible, I’d jump straight to solutions. “Let’s use Flash for this part! And over here, we’ll design some awesome icons for navigation…” Our users were still important, but they were there to bear witness to how cool our designs were.

Then I met Indi Young. Indi and I were among the founding partners of Adaptive Path, a user experience consulting company that focused on research-driven design. We founded the company in the dark days of the Web industry. It was 2001. “Dot com” was a dirty word, companies were cutting their Web budgets, and projects were drying up everywhere.

It was then that “research-driven” started having real meaning to me. As Indi introduced her methodology and resulting visualizations, it became clear that she wasn’t just trying to make designs better in some abstract way. Rather, her process was simple enough to resonate with anyone on a Web team. And perhaps more importantly, it would help connect Web teams to other core parts of their organizations who were skeptical of spending even another cent on their web sites.

In the end, using Indi’s process, we were able to convince teams that we weren’t researching all the creativity out of their projects. We were researching the risk out. And no matter how the industry is faring, that’s a story people want to hear.

This book is an excellent guide to a research method firmly grounded in common sense. But don’t let the simplicity of the process detract from the power of the change it can enable. Talking to users in a structured way, analyzing in a collaborative way, and diagramming with clarity can transform the way you approach the Web.

And it might just ignite your creativity.

The record and the bet

2007-08-08T00:00:00+00:00

It was nearly six years ago, in a bar in the Mission district of San Francisco. I was having a drink with friends and we were talking about baseball - home runs, in fact. It felt like just about anyone could put one over the fence - Sosa, McGuire, and Bonds were competing for single-season records and someone even suggested that Hank Aaron’s all-time mark could fall.

Mike Monteiro, one of the founders of Mule Design, was there that night. “Bonds will do it,” he said. “I have no doubt in my mind.”

“Bonds?” I said. “You’ve got to be kidding. He’s old, injured, and taking so much junk - he won’t even last the next season.”

“You have no faith.”

“You’re blinded by hero worship.”

And on it went. Eventually, we found ourselves at an impasse. And, as guys do when they argue about sports, we decided to make it interesting.

“Alright. A hundred dollars says he does it,” Mike said.

“Don’t be so minor league. Five hundred,” I countered.

“Oh yeah? Loser buys a luxury box at a game.”

I paused. Clearly, this wasn’t about money. It was about pride. I had to go after his. “We have to make this count. This has to be something really meaningful.”

“What do you have in mind?” he asked.

“Loser wears a dress to a game.”

We shook hands.

Last night, Bonds put one into the centerfield bleachers. There were fireworks, speeches, and tributes. My IM client popped up a few minutes later with a message from Mike:

http://www.chadwicks.com/Department.aspx?DeptId=8295

See you at the game…

I am a man of my word. Later this season, I will uphold my end of the wager. There is already talk of reserving the centerfield bleachers and getting a little something up on the Jumbotron. I’m considering my wardrobe; elegant and tasteful while still fun and flattering. Your suggestions would be most appreciated in the comments.

My only consolation? I’ll be a 6’6” guy with a goatee wearing a dress in San Francisco. Nobody will blink an eye…

Fixing the Web

2007-08-07T00:00:00+00:00

The editors of xhtml.com recently invited me to participate in a series they’re running titled Fixing the Web. In particular, they asked, “In your opinion, what parts of the Web need to be improved or fixed in order for the Web of today to evolve into the Web of the future?” Here’s how I replied:

I wish every device that was capable of talking to the network could send its geolocation. I’d like this to be fundamental – let’s send longitude and latitude in the HTTP header of every request. Let’s make it as ubiquitous and accessible as the time stamp, user agent, and referring URL.

When location is assumed, we can go beyond the obvious applications of a geo-aware Web. Yes, the map application will be able to center on you when you launch it. But what else becomes possible when email knows where you were when you sent it, or a social app knows the proximity of your friends? We can’t even begin to imagine what will emerge.

But what about the security and privacy implications? We’ve been dealing with them since the first Web sites launched. Issues with cookies, cross site scripting, and search query logs have given us precedent for enhancing rather than exploiting personal data. Giving users control of their location can learn from the same paths.

There are hurdles, to be sure. Hardware manufactures still believe access to location data should be a business opportunity, much like software companies once believed proprietary file formats would protect their bottom line.

That’s a shame. The Web grew to prominence providing the “what” and knowing the “who.” Let’s add the “where” and see what happens next.

Raiders of the Lost Ark and the mystery of inspiration

2007-07-12T00:00:00+00:00

At a conference recently, I heard Dan Cederholm from SimpleBits talk about inspiration. He showed a bunch of different techniques he uses, including howw he uses Photoshop’s mosaic filter on an image to blow up giant pixels representing the basic colors in the picture. He uses those as pallets for the design work he does. Very cool.

Inspiration can come from process, but it can also come from the most unexpected places. For example, when we were designing the data-over-time visualizations for Google Analytics, we were totally stuck with bar graphs. We’d iterated dozens of times, scoured the web for examples to steal, and had tried just about everything. The result felt muddy and chartjunked; the data didn’t feel clear and was weighing down the whole page. Finally, I told the team to forget about that problem for a while - we had some time, we could come back to them in a couple weeks. So we did.

The next month, I got the box set of Indiana Jones movies and watched Raiders of The Lost Ark. I still love that movie. But that night I had a dream inspired by the travel montage - the one where Indy is flying to Nepal. The plane’s route is traced as a line that bounces from city to city across a map, leaving a big dot where they landed. In my dream, I could see the airplane flying over our charts, trailing a line behind it, leaving a dot at each data point. The next morning I woke up, grabbed my laptop and drew a line-with-dots chart in OmniGraffle while still in bed. It’s what we ended up using in the new version of Analytics (with a tremendous amount of polish from the team, of course).

Moral of the story: sleep with your laptop.

What I've been working on: The New Google Analytics

2007-05-08T00:00:00+00:00

It’s been well over a year since Google bought Measure Map and I left Adaptive Path. And wow, have we been busy. The problem is, though, that whenever someone would ask me what we’ve been up to, all I could do is smile and say, “Oh, just wait.”

Well, the waiting is finally over - we can talk about what we’ve been working on all this time. Today, a completely redesigned version of Google Analytics is launching, bringing a lot of the simplicity and data visualization techniques we learned building Measure Map to a whole new scale. This was some of the toughest design work I’ve ever tackled, and I’m immensely impressed with well our team worked together, especially Greg Veen, Ryan Carver, and Douglas Vander Molen.

I wrote more about the redesign over on the Google Blog, and my colleague Avinash Kaushik has written a fantastic and comprehensive overview on his blog.

On a personal note, I’ve got a much different job now that the design work on Analytics has wrapped up. I now lead a team of over 30 designers and researchers responsible for the user experience of Google’s web applications. We’re working on Gmail, Calendar, the Office-like tools, Blogger, Orkut, Picasa, Talk and a bunch more. And that means I get to dig into some of what I consider the Web’s most interesting challenges and work with some of the smartest people in the industry.

Can’t ask for much more than that, really.

Remembering Zork

2007-01-11T00:00:00+00:00

Did you ever play text adventure games years ago? They were the sort of command-line games where you would type simple directions and work your way through a story. I played these for hours and hours as a kid at my friend’s house on his dad’s teletype. I remember feeding the green and white paper back through as many times as we could to use all the available space. Later I moved on to the Zork series and other Infocomm titles, staying up way to late at my Commodore 64.

I don’t know what made me remember Zork recently, but it was a pleasure to find that they’re playable online. Here’s an excerpt from one of my favorite puzzles:

>look
Tiny Room
This is a tiny room carved out of the wall of 
the ravine. There is an exit down a precarious 
climb. On the north side of the room is a massive 
wooden door, which has a small window barred with
iron. A formidable bolt lock is set within the 
door frame. A keyhole covered by a thin metal lid
lies within the lock.

>open door
The door is locked.

>inventory
You are carrying:
 A sword
 A lamp
 A place mat
 A letter opener
 A newspaper

>examine door
The door is made of oak and is closed.

>look in keyhole
No light can be seen through the keyhole.

>open lid
The lid is now open.

>put newspaper under door
The newspaper crumples up and won't go under
the door.

>put placemat under door
The place mat fits easily under the door.

>put letter opener in keyhole
There is a faint noise from behind the door 
and a small cloud of dust rises from beneath it.

>get placemat
As the placemat is moved, a rusty iron key falls
from it onto the floor.

>get key
Taken.

>unlock door
What do you want to unlock the door with?

>unlock door with key
The keyhole is blocked.

>get letter opener
Taken.

>unlock door with key
The door is now unlocked.

>open door
The door is now open.

Success! I can distinctly remember the feeling of first pulling that place mat out and seeing the key on it. Such a sense of accomplishment for a 12 year old.

Today, people still write and explore interactive fiction, though I pretty much gave it up when I hit high school. These days, I solve tricky problems by typing into an email client rather than a green-screen game. The skills, however, are pretty much the same...