Hard to read but very honest retrospective on Everpix, which recently shut down. I wish more companies would share like this when things don’t go right (but I get why they can’t stomach it) — we’d all learn so much.
Google Loon is just one example why they’re the most ambitious company in the world. Have loved following this project.
Over the weekend, this tweet went viral:
I don’t appreciate this guy’s unnecessary snark, and I don’t know enough about the context within Twitter’s dev team to know if it’s fair criticism or not in this particular case. But I think this went viral because it resonates with the storyline of a startup gone corporate — the cool kid that got popular and left behind what got him there. Twitter was a startup darling because of how seemingly simple and almost non-existent the app itself was. It was scrappy and felt like something that was whipped together in a weekend. So what could all these developers, designers, and PMs possibly be doing all day?
The response to this is that, when you’ve got 200M+ users across the world and you’re building a platform to support that sort of scale, things gets more complicated. Tweet below from a Twitter engineer:
Not an entirely convincing argument in response, but there’s some truth there. Scale brings edge cases (like accessibility) and other complexities. The reality is likely somewhere in between the two extremes, as usual. I find the debate interesting because it’s an adolescence every successful startup turned legitimate business is going to go through:
How do you get bigger as a product team without “getting big” — slowing down, adding bureaucracy, and losing focus? How do you widen the pipeline for growth and experimentation without stumbling into diminishing returns and inefficiency?
What worked for a product team when it has 2 engineers, one designer, and few existing customers or stakeholders is not going to work after a couple years of rocketship growth. I had to learn this through experience. Early on at Trunk Club when we had just a couple people working on technology, if we had an idea or concept we wanted to try, we could shoot first and aim later. Downside risk was limited and it was fairly easy to coordinate amongst ourselves. But our pipe was so thin — we felt like our hands were tied without more talented people on our team to help us execute against the opportunity we knew we had in front of us. So we started hiring.
We’ve been able to grow from 1-2 people up to nearly 20 engineers, designers, and PMs while maintaining our entrepreneurial, hypothesis-driven culture. Ironically, maintaining that culture requires a recognition that the team has to change and mature as it grows — scrappy only works for so long. We’re not perfect by any means — we have fits and starts in our process and coordination — but I learned a few necessities about how to pull this off relatively successfully:
- Recruit the best: More people is not a panacea for growth. In fact, hiring the wrong people can easily make the team less productive. Scaling without “getting big” requires really careful judgment about who joins the group — we want builders, not talkers; entrepreneurs, not ladder-climbers. When people join a team, they look around, evaluate the culture, and model after it — every new hire needed to see that we moved quickly as a team and didn’t tolerate politics or individualistic thinking.
- Brutal roadmap discipline: As the team gets bigger, it’s tempting to just want to work on more stuff. “We have the people, so why not do this” goes the thinking. Leadership on the team has to force focus and brutal prioritization of the roadmap. What makes a small team successful is that there’s no choice but to focus — as the company gets bigger, that choice has to be made more intentionally and decisively. It’s hard. You should only hire more people if you need them to execute on the stuff that matters, not just because you can.
- Cohesive culture from the top: When teams get bigger, it’s easier for each individual to feel more distant from the success of the team. “I can’t really move the overall needle, so I’m going to look out for myself”. It’s really tempting to start focusing on what I’ve accomplished rather than what the team has. So that has to be stomped out from the top. One of the things I’m proudest about our culture is that we recognize and promote the people who kick ass but don’t take any credit for themselves. That sets a tone all the way around the company.
It’s not easy to get bigger as a product team without losing focus, adding painful process, and slowing down. But lots of successful startups face just that challenge as they grow.
Fall in Chicago
Online retail sales, while growing rapidly, still represent only about 6% of total retail sales in the US today
US Census Bureau, August 2013
Everything you’ve heard about Amazon’s dominance and every other retailer scrambling to catch up to them in e-commerce is true. Everything you’ve heard about the stunning growth in mobile transactions is true too, probably to a greater degree than most people realize. Those trends are rocketships that everyone needs to get behind. Yet, still with all of that, only 6% of retail sales today in this country are happening online?
Why is that?
At Trunk Club, our central thesis has always been that there’s a major gap in the customer experience between physical bricks-and-mortar retail and online shopping as it’s been defined since the mid-90s, i.e., a search-driven UX that’s aimed at users who know what they want to find.
For many types of products and for many types of customers, that gap has made for a pretty crappy experience — apparel (huge market) and dudes who hate shopping (huge market), for example. Many guys don’t know what they need or want, so a “pull” oriented service like most online retail doesn’t serve their needs — they need a service that “pushes” the right products to them at the right time. In our model, that means empowering a human advisor with unprecedented amounts of data, intelligence, and tools to deliver delightful experiences that the customer didn’t even need to ask for.
As a technology team, we’re focused on helping our guys get great clothes like everyone else at the company, but our vision is more agnostic and universal — we want to build the technology platform underneath that advisor relationship, the service platform that enables a customer experience you just can’t get from physical retail or online shopping as they are today. To the customer, our software is largely the part of the iceberg underneath the water, enabling the advisor to make all the right decisions and deliver targeted recommendations without losing the final slice of judgment and human touch that we hope makes our service so delightful.
When we recruit engineers and designers, we often hear that they had no idea how much fairly sophisticated and cutting edge technology — from server-side to client-side — goes into delivering our service, given how human it feels. We take that as a big compliment (and then we tell them all about it because we think it’s cool). We think the platform to deliver a human experience at true scale is rare and undiscovered technology territory — there’s 94% of the market left to take, after all.
"Great films are never finished. They escape." - George Lucas
Building software is a constant tension between the “best” result and the fastest result. The lean startup and agile movements are dedicated to finding the right balance between those two extremes. The thesis of those schools is that entrepreneurs should focus on doing as little work as possible in order to learn something new about their product and users. Short on cash and time, the best thing to do is release something to the world as soon as possible and start getting feedback.
I’m a firm believer in that advice, but I also think it needs some judgment when applied. Releasing something and learning from it doesn’t mean it needs to be shared with every customer or user right away. It doesn’t even mean it needs to be complete — it could be a paper sketch. What’s important is that you get feedback so that you can make a judgement call about when the product is good enough to represent your brand to the world at large. First impressions matter — if your business is a payments company, a lifestyle/fashion brand, a utility (like an email client), or something similar, users just aren’t going to stick with you if your experience is incomplete or sloppy. Mailbox worked for 2 years before releasing to the public; Color tried to release right away with an experience they hadn’t tested with users. Particularly in the app ecosystem we’re all in, attention spans are short and other options are very nearby.
At Trunk Club, we’ve executed extended Betas for major functionality and experiences for months before releasing them throughout the business. The takeaway for me: release very rapidly in controlled environments, learn and iterate as fast as possible, then release broadly when the product meets your standards for the brand and customer experience. Recently, I helped “dogfood” a major change to our service platform and operations. Each day, we picked up insights and often had new features or bug fixes pushed live the same day. By the time we released to the whole business, the launch was a non-event. It just worked.
Deciding when to release a product is more art than science. Either way, it’ll never feel finished. You just have to know when to let it escape.
Many mornings, I leave the house an hour early and set up my laptop at the Starbucks around the corner from our office. I like to use that time in the early morning to clear my head and work on the stuff I wouldn’t get to once the real day starts, like reading news or writing down ideas in a notebook. It’s where I am now, actually.
I’ve probably sat on every single chair or stool in this Starbucks, many times over. In the first year or two of Trunk Club, we had no space for private meetings or conversations, so we took everything across the street to Starbucks. When I look over to the coffee bar in the back corner, I remember standing there with a piece of paper sketching out our earliest tech stack with John Tucker. When I look over to the small tables near the entrance, I remember trying to close a recruit who turned out to be our VP of Engineering and one of my close friends. The high tables near the north windows remind me of 1-on-1s with Brian.
Memories live in space. There’s a lot of history between these walls for me and our company. I’m sure there’s a lot of history for other people in here too. That might just be the best possible compliment for what community centers like coffee shops represent.
I don’t even drink coffee outside of here — I just didn’t want to get kicked out for loitering.
Enjoyed doing this interview with Kontagent about what we’re doing in mobile and how much usage we’re seeing from members in our iPhone app.
Fantastic long read by Atul Gawande on the history of healthcare innovation around the world and what it means more broadly.
I should preface this by saying I’m almost back to 100% health now and my go-forward prognosis is excellent. I was very, very fortunate and this post is about what that fortune taught me.
About a month ago, I was eating lunch at my desk on a pretty normal Tuesday afternoon when I got a call from my doctor. A few days before that, he’d felt something unusual in my abdomen during a routine physical and recommended a CT scan to check it out. On the phone, he spoke with some urgency: the scan showed a very large tumor — 8.5 x 7.3 x 4.3 inches, not much smaller than a football — in my abdomen. He told me I needed to meet a surgeon the next day to get a full diagnosis and find out what my options were.
I was stunned, but still dismissive that this would be anything but a minor annoyance. How could it be? I was in my mid-twenties with a perfect medical history and no symptoms or pain whatsoever. When I called the surgeon at Northwestern I was referred to, the receptionist on the other end said, “Thank you for calling the Comprehensive Cancer Center at Northwestern, how can I help you?”. Thump — the “C” word. The weight of what I could be dealing with finally hit home. This might take a while.
The next six or seven days were an excruciating waiting game, inching closer and closer to a definitive diagnosis of what kind of tumor this was. I tried to stay away from searching hypotheticals on Google and Wikipedia (which had more than enough information to keep me awake every night). My parents, who are doctors themselves, flew up to Chicago to help me navigate the maze of next steps. I had consultations with several oncologists and surgeons at Northwestern and UChicago, an MRI, blood tests, and a needle biopsy to pull sample cells from the tumor to analyze its malignancy.
Then, some amazing news. Preliminary biopsy results indicated I was dealing with a Desmoid tumor, a very rare (2 per million in the US) type of mass which many doctors don’t even classify as cancer because it behaves in a very benign fashion. Of the 9 or 10 things my doctors had hypothesized this could be, this was the rarest and the best.
Soon afterwards, I had successful surgery at Northwestern, led by an incredible team of doctors. I’m amazed at their skill and grace under pressure — modern medicine is approaching magic. Within two days, I was out of bed and walking the hospital hallway with limited pain. Then, even better news: post-operative analysis on the tumor indicated it was actually less aggressive than a Desmoid, an even more incredibly rare outcome that means it’s unlikely I’ll need cancer drugs or further surgery to stop this thing from coming back. Fingers crossed, once I finish recovering and put some weight back on, I’m done.
Me, about to roll into surgery
Talk about a heavy dose of perspective.
Behind every “one in a hundred” story like mine, there are ninety-nine stories that aren’t so neatly wrapped up within a month. I’m young, I work hard, I’m smart, and a lot of people care about me, but some really bad stuff almost happened anyway. I only had to look around the waiting room in the cancer wing at Northwestern to see other young, hard working, smart, and loved people who maybe weren’t as fortunate. In fact, we all probably know someone who has gone through a much tougher road than I had to and won’t have time to write a blog post about it. These days, I’m thinking about them a lot.
I didn’t need a close call like this to understand that — I like to think I had my head on straight beforehand too — but you’ll just have to take my word that it gets a lot clearer once you do.
During those six or seven days of uncertainty before we got biopsy results, I remember laying in bed and thinking that life had gotten really simple. I was only focused on a few things — my health, my family, the girl I love, and my friends. If I could emerge from this with those things intact, I’d have everything I needed. Well, here I am.