The future is built, not discovered

TL;DR — I joined Atomic earlier this year to work at the intersection of starting and investing in companies. It’s already been everything I was hoping for and more; here’s why. 

The traditional framework of a job search always felt wrong. High-achieving individuals don’t really look for jobs. (Fun fact: the word “job” comes from jobbe or “piece of work”, which is a synonym for “task” and seems anachronistic!) They look for opportunities to work with amazing people on big problems that have enduring impact.

I’m lucky to have had a wide variety of experiences in my career thus far: 

  1. Finance: I started as an Investment Banking Analyst at Barclays in New York City
  2. Venture: Moved to San Francisco to join the investment team at Institutional Venture Partners (IVP)
  3. Startup Studio: Joined the team at Expa to build a cloud kitchen startup (acquired by UberEats) and a digital currency network
  4. Crypto: Helped Polychain Capital and the DFINITY Foundation launch an ecosystem fund

Looking back on the jobs I’ve had, my predominant memories are less about the specific work I did (though that part was thrilling!), and more about the people I worked with along the way. So I embarked on a people-first search: prioritizing who I would work with above industry, role, responsibilities, title, compensation, and all else.


Now, “people-first” isn’t a setting on LinkedIn or Indeed, so I had to improvise. I started by asking my friends to help me think of:

  1. The most talented people they knew; the brilliant visionaries who make up the very fabric of Silicon Valley
  2. The friends who were always overwhelmingly happy in their jobs — you know, the ones who are always saying “I love my job!” with a half-guilty, half-satisfied grin
  3. The mentors who helped them grow the most and who were incredibly invested in the success of their team

Time and time again the names of the Atomic partners were suggested to me from across industry peers, mentors, former colleagues, and also other VCs. This was exactly the type of signal I was hoping for and it was equal to a resounding “Yes! I would work with ______ in a heartbeat.”

Crowdsourcing this “best people to work with” list helped me clarify my own priorities when it came to finding my tribe. For me, the people factor breaks down in two dimensions: 

  • Talent: finding mentors who I can hold as role models and continuously learn from for decades to come
  • Values: finding alignment in values (who we are and how we interact with each other); those that pass the airport test: “Would I want to be stuck in an airport with this person?”

Over the last few months of working with the team and partners at Atomic, I learned first-hand that I could not have found a group of people that fits this intersection of both talent and values better. Each of the Atomic partners have co-founded a 1B+ company and as such, they’ve set a high bar for the talent they’ve been able to attract. Founding and funding companies is hard work, but we also make sure we enjoy the journey as much as the destination. After all, life is ephemeral; what’s it worth if you’re not having fun?


I am not a specialist (sorry mom, no Dr. Kong in the family yet!…), and I’m not really even T-shaped (sorry #VCTwitter). Silicon Valley is chock full of visionary founders who can dream the future. My superpower is taking those dreams and doing the hard work to make them reality. The challenging task of executing and operating a business, the magic of getting shit done, is my passion. 

Getting shit done is an art and a science. The art is embedded in personal values — some call it grit, perseverance, hustle… whatever term you prefer, you know someone who can “get shit done” when you meet them. The science, on the other hand, is purely process. Process is a double-edged sword that can either improve or hamper success depending on how it is implemented and evolves. Process done right is the Mario Kart Speed Boost for your startup.


In my last few months at Atomic, here are some ways I’ve learned to think about our process:

  • Decentralize small teams for quick decision-making
  • Ruthless prioritization with a bias towards action
  • Err on the side of learning over the importance of being right
  • Get out of your own way when things start to work


Even with the right team and right processes, no organization is complete without a common goal — a raison d’etre, a reason to get out of bed every morning. As our founder Jack puts it, 

As entrepreneurs and company creators, it’s our responsibility to figure out ways to apply today’s technology to traditional industries and everyday problems with the goal of driving down costs and increasing the quality of life for all.

Day in and day out, our collective goal at Atomic is to create the next big thing. We believe that the future is built, not discovered. Come build it with me.

This post was originally published on on November 22, 2019.

The End of Disruption

or, how startups can continue to innovate and thrive.

A recent Vox article by Timothy Lee is titled: “The end of the internet startup”. In his piece, Lee goes on to explain how today’s startups face a tougher competition from technology incumbents due to two primary factors.

First, technology giants acquire early and often.

This has become easier lately especially with the help of their large balance sheet (cash) and high enterprise value (equity). There are many successful examples of this that we can point to in recent history: Google purchased YouTube and Android, Facebook acquired Instagram and WhatsApp, Amazon bought Zappos and Quidsi.

At the same time, CEO’s of public technology companies have enjoyed unprecedented power in governance, either with the introduction of multi-class shares (à la Facebook and Google) or with the support of shareholders who are willing to look beyond short-term returns (à la Amazon and Tesla). This kind of support enables executives retain their founder mentality and invest for the future; which is why we see incumbents proactively going after “FrontierTech” spaces, such as Artificial Intelligence, Virtual Reality, and Natural Language Processing.

These represent the same areas that were previously so risky — from both a technical and market perspective — that only startups dared to venture.

Secondly, scaling a tech startup has become more expensive over time.

This is a perspective echoed by Andrew Chen, in his piece Startups are cheaper to build, but more expensive to grow. Andrew Chen argues that startups have been raising bigger rounds and thereby investing earlier in paid marketing. He also points out that a subset of startups have achieved high LTVs early in their lifecycles, which drove up CAC for the rest of the industry.

I would add that online distribution channels have become more concentrated and more expensive over time as the internet maturesThis shouldn’t come as a surprise to American entrepreneurs and investors, as we’ve already seen this play out in China. Apps like Tencent’s WeChat have become the portal that consumers use to interact with the rest of the internet. Today, the US consumer likely spends his/her time on only a few internet platforms without even realizing it. Axios reports that:

Smartphone users spend 96% of their app time within their Top 10 most used apps and Smartphone users spend half their time on their #1 most used app, according to comScore. Google and Facebook own the overwhelming majority of traffic to these apps and are the #1 app for every age group.

In addition, eMarketer estimates that Facebook and Google are expected to take half of all internet advertising revenue worldwide, and more than 60% in the United States. In the U.S. market, no other digital ad platform has market share above 5%. Given the growth in revenue concentration by these internet giants, the distribution of ads and content on these platforms has become more fiercely competitive and more expensive in parallel.

Thirdly, I would add that building a startup has also become more difficult because today’s tech incumbents are keen students of “disruption theory”.

Clayton Christensen, author of The Innovator’s Dilemma, first introduced the concept of disruptive innovation in 1997. According to the Harvard Business Review:

Disruption” describes a process whereby a smaller company with fewer resources is able to successfully challenge established incumbent businesses. Specifically, as incumbents focus on improving their products and services for their most demanding (and usually most profitable) customers, they exceed the needs of some segments and ignore the needs of others. Entrants that prove disruptive begin by successfully targeting those overlooked segments, gaining a foothold by delivering more-suitable functionality — frequently at a lower price.Incumbents, chasing higher profitability in more-demanding segments, tend not to respond vigorously. Entrants then move upmarket, delivering the performance that incumbents’ mainstream customers require, while preserving the advantages that drove their early success. When mainstream customers start adopting the entrants’ offerings in volume, disruption has occurred.

The current narrative between Snap and Facebook is a case study for when disruptive innovation may not successfully bring down an incumbent. When Snapchat found success and consumer traction with their mobile app in 2011, Facebook recognized the threat to its core business early on and moved to quickly defend itself.

  • In 2012, Facebook launched Poke, an iOS app for sending expiring text, photos, and videos to directly compete with Snapchat.
  • In early 2013, Facebook started to allow users people to send Instagrams with its Messenger product.
  • In late 2013, Facebook offered $3B to acquire Snapchat, which the latter declined.
  • In 2014, Facebook retired Poke from the app store and launched Slingshot, which also focused on sending ephemeral photos to friends.
  • In late 2016, Facebook launched Stories, disappearing photos, and live video on Instagram. Views on Instagram Stories gain traction and quickly overtake Snapchat by mid-year 2017.

Over the past five years, Facebook has employed an iterative fast-follow strategy to not only defend its core business against the rise of Snap, but to also gain ground in turn. It is perhaps one of the few large company to have executed on such a plan, but this example nevertheless demonstrates thatlarge incumbents who have embraced agile development, full product verticalization, and retention of top talent have come to realize that their true defensibility is in their internal processes. Even at their larger scale, incumbents are often out-executing startups. What you ship has become less important than how quickly and how well you can ship it. The nature of innovation has changed: there’s no shame in copying if you can take something and build it better.

So, despite increasing competition from fast-moving incumbents, what can startups do?

There are several offensive tactics that startup founders can consider:

  • Software eating the world: continue to look outside of Silicon Valley for opportunities to bring the same type of digital innovation to an otherwise laggard industry, such as education, healthcare, agriculture, logistics, etc.Most of the conditions I’ve listed above apply primarily to consumer tech incumbents, so there may still green space in the enterprise tech world.
  • Work on new (and harder) problemsfind opportunities to re-imagine products and experiences outside of the realm of the large internet companies (ie. Airbnb, Planet Labs, Opendoor, Color Genomics, Coinbase, etc.)
  • Focus on your core: the best startups already knew that focus is key to success. Today, focus has also become necessary for survival. The more your company can focus on building the core product or value proposition, the better chance you have of getting there first.
  • Invest in the brandif you are a consumer startup, your brand is now one of your biggest assets against the juggernauts. This is equally relevant for a retail company (ie. Glossier) as it is for a software company (ie. Slack).

On the defense, startup founders should also consider:

  • Keep your enemies close: stay abreast of your competitor’s priorities.This is a good opportunity to leverage your board of directors’ network and also grab coffee with a VC associate to get the lay of the land.
  • Build bridges early: develop corporate development relationships and don’t don’t blow off that email from the corp dev associate. Take the meeting and foster the relationship with the team, either to hatch a partnership together or to lay the groundwork for a future acquisition.

What other tips would you give startup founders who are building a startup in the consumer tech space? Let me know your thoughts!

This post was previously published on Medium on Sep 7, 2017.

Why Your Startup Also Needs a BizOps Team

The title of “Business Operations” (or “BizOps”) has become one of the more buzz-y Silicon Valley terms as of late. At IVP, we saw an increasing number of growth-stage startups (many of which are in our own portfolio) recruit for and build out a BizOps team — Slack, Dropbox, CloudFlare, Stripe, Counsyl, NerdWallet, and ZipRecruiter, just to name a few. While this may seem like a new trend, large tech incumbents have long ago pioneered the role of Business Operations; companies like Yahoo, Google, and LinkedIn currently have upwards of hundreds of people on their BizOps team.

What is BizOps?

BizOps teams are generally an internal-facing group and broadly aim to create more value for the company and to improve the profitability of the business.Tech startups are already familiar with the role of Product Managers — people who drive product development by setting feature priorities and collaboratingwith a team of engineers and designers to ship these features. By the same token, shouldn’t other areas of your business have designated “managers” as well?

BizOps is PM for your Business Model

BizOps teams have a “get shit done” mentality and are often tasked with translating business goals (strategy) into tactical operations (execution). For example:

  • We are going international. What are the first markets we should tackle and how will we penetrate those markets?
  • We need to further monetize our user base. What are the most effective and sustainable ways to increase monetization? What are the different levers we can control and what is the expected impact of those changes?
  • We’re looking to accelerate growth. What are the key factors in our growth rate and how can we move the needle?

While BizOps responsibilities are similar across companies, BizOps teams can be structured in different ways. For some, the BizOps team is like Seal Team 6 — internal consultants deployed for specific and shorter-term projects. Because they act as a neutral third party, BizOps teams in this structure don’t interact closely with other business units regularly.

More commonly, BizOps teams are flat structures that are horizontally tied to other functional groups, such as:

  • Sales, where a business operations associate helps the sales team figure out a go-to-market strategy, perfect their conversion funnel, and track their sales efficiency
  • Product, where a business operations associate helps drive product strategy decisions, figure out a launch strategy, and track important user stats as the product launches
  • Finance, where a business operations associate works closely with the FP&A team to determine the key drivers of the financial model
  • People, where a business operations associate works closely with recruiting and HR to accelerate hiring or increase retention

BizOps Evolves as Your Company Scales

For early-stage startups, BizOps allows for a flexible, dynamic, scrappy team to work on a multitude of different types of business problems that otherwise may not fall anywhere else in the company (eg. starting a new office, doing sales hiring, owning the company financial model, etc.) As the company grows, BizOps can prioritize key issues (e.g., growth, monetization) and help balance strategy with execution. And for larger and more mature startups, BizOps provides a cross-functional layer of communication, structure, and organization across the greater company.

Depending on the stage of your startup, here are examples of a few inflection points where a BizOps team may make sense:

  • Accelerating growth: Once a startup has achieved product/market fit, you’ll probably want a dedicated BizOps team to help accelerate your early growth.
  • Starting geographic expansion: As startups go abroad, each new location you enter can behave similarly or differently from your existing geographic footprint — a BizOps team can help figure out your international strategy.
  • Avoiding functional silos: As startups and team sizes grow, it becomes more difficult to get transparency across the organization. A BizOps team can act as a cross-functional layer that keeps your startup’s different departments on the same page.

There are many approaches to BizOps, and it’s really up to founders and CEOs to decide how to utilize BizOps most effectively for their company. Ultimately, the key question is not be whether your startup needs a BizOps team, but when.

This article was originally published on Business Daily.

An Exercise in Foresight

I really love this (excerpt from a Techcrunch article by Roelof Botha):

Imagine it’s five years from now and your company is a massive success. you’ve got a fantastic product, you’ve achieved market leadership, and everyone else is trying to catch up. What decisions helped you get there?

Now imagine the same five years have passed, but this time your company is struggling. You failed to live up to your original vision. Your product has stalled. You’re backed into a corner. What choices led to that moment?

These two questions are the opening exercise in a strategic planning technique that Larry Summers — the economist and former Treasury Secretary — led the team at Square through recently. The process sparked important discussions about choice, risk and foresight.

Likely a useful exercise to do, whether you’re a startup, a larger company, or a VC, even. Probably useful for personal life, too.

The Bear Case on Fitbit

It came as a surprise to most when Fitbit finally unveiled their financial numbers in their S-1. The company did over $700M in revenue last year (that’s over 10M devices sold!) and maintained an EBITDA margin of close to 26%. It was even more surprising how well FIT was received by public market investors. Fitbit’s  stock surged almost 60% on its opening day and has continued to outperform until its first earnings call. On August 5th, the company lowered revenue guidance and indicated that margins will decline in the next quarter, sending the stock down ~16% in after-hours trading.

Since I’ve been fairly bearish on FIT ever since its IPO, this bleak financial forecast did not come as a surprise. It seemed as if investors were so impressed by headline financial figures that they didn’t look further into Fitbit’s business metrics. FIT’s cumulative device sales track closely with registered users, which indicates that users rarely buy more than one device (despite Fitbit’s wide array of offerings). 

And, although 1 in 10 Americans owns a fitness tracker, research show that ~50% of them wind up in a drawer somewhere within six months. This kind of high user drop-off and low purchase repeat rate is a leading indicator of slowed growth down the line.

I’ve also always been skeptical of the product itself. Already, we have seen several startups, like Misfit and Moov, approach the wearables from the low end. In fact, you can buy a pretty decent activity tracker on (by Xiaomi) for $15! With Fitbit also victim of product recalls (the latest for the Fitbit Force in October 2014), it becomes clear that FIT will be subject to margin compression.

While Apple has indeed created general lift for wearables with the release of the Apple Watch, it has also set a price ceiling for these devices. For $350, you can get the lowest-end Apple Watch that does activity tracking on top of all its watch-specific apps… so why would you pay more than $300 or even $250 for a Fitbit?

Given these dynamics, Fitbit really runs the risk of becoming the Samsung of Wearables. While Samsung was able to grab a large share of the Android phone market a few years ago, the company has been slowly losing its footing to the likes of Apple and Xiaomi. Formerly the top mobile device maker in China and India, Samsung lost its footing in both crucial markets in the second quarter of 2014. How can it avoid this fate? Here are some suggestions:

  • Go Niche: Sometimes the right answer is a counter-intuitive one. Samsung failed in some cases because it flooded the market with a slew of low-end models – none of which appealed to the average consumer. Instead of trying to please everyone, Fitbit should take some of SoulCycle’s ethos and target a small and passionate niche. This may mean creating a Fitbit targeting fashion-forward women (a la Ringly), or creating a Fitbit for hardcore athletes (a la Athos). The wearables market will be large, but not necessarily homogeneous
  • Application and Service Integration: Samsung failed to build usable software and useful services on top of its hardware component. Samsung’s proprietary OS, Tizen, also never took off. I know Fitbit is incredibly focused on their consumer mobile app, which is performing well in the iOS and Google Play stores. The company has already made headway in this area by integrating more social and smartwatch-like functions to their fitness trackers, and should building the software and services layers.
  • Put Health First: Fitbit has proven its ability to make an activity tracker with a few bells and whistles. With the acquisition of FitStar, the challenge becomes making the app and wearable integration increasingly valuable to the user from a health-first perspective. In order to increase its unique value proposition and user retention, the Fitbit app should not only track activity, but find ways to increase and promote well-being.

Kitchen 2.0

My friend Star Simpson and I co-authored this piece a couple of months back, and I love that I’m still getting feedback on it! Love that there are so many entrepreneurs tackling this space.

Introductory excerpt below:

We are collectively obsessed with food these days. Since the launch of the Food Network in 1993, TV shows such as Iron Chef and Anthony Bourdain’s No Reservation have become immensely popular. Modernist Cuisine, a 52-pound masterpiece published in 2011, was deemed “the cookbook to end all cookbooks” and the term “foodie” entered the mainstream lexicon. We spend hours staring into the kitchens of others, and at HD renderings of exquisite meals. But, while Americans have become more familiar with the newest celebrity chefs and the hottest food trends (e.g. Cronuts! Liquid nitrogen ice cream!), we are also collectively spending less and less time in our own kitchens.

3 Ways Thumbtack Can Improve User Experience

Ever since I’ve had to find movers (SoMA to Jack London Square), I’ve been a big fan of Thumbtack. My friends and I have used Thumbtack to find local service providers such as contractors, golf instructors, housekeepers, and even a face painter.

I love that Thumbtack does the work for me. It essentially reverses the workflow of search for local services — instead of me taking the time to pore over pages of search results on Google or Yelp, Thumbtack takes some information on what I’m looking for, then returns up to 5 quotes from local professionals who may be a good fit for my request. While the user experience is pretty good, here are some ways I think they can be better.

1. Better communication

Below is a screenshot of the messaging platform on Thumbtack.


Each professional sent me a moving quote with a price per hour, as well as a short message introducing themselves and their business. However, I often found a long lag-time between message response, which led me to wonder if the professional was checking Thumbtack regularly. In fact, one professional I hired on the platform stopped answering my messages all together! One way to improve communication could be read receipts — a check mark for whether the other party has seen my message, or a time stamp showing their last login.

I don’t think it’s clear to the user that Thumbtack purposely caps the number of quotes to five. I can easily imagine a new user in limbo, waiting for more quotes to arrive before s/he makes a decision. I’ve also made a couple of requests that have only received one or two quotes — in this situation, Thumbtack can maybe let me know that I have tapped out the market or prompt me to widen the request criteria. On the supply-side, for requests that do not get five replies, Thumbtack can also waive posting credits to encourage service providers to submit a quote.

2. Easier transactions

When it came time to pay my movers, it was definitely awkward. I had to pull up the Thumbtack messages to recall the rate we had discussed and then run to the ATM to withdraw cash. For jobs that had slight delays or additional complications, I sometimes had to re-negotiate the transaction amount.

If Thumbtack can own the payment flow of the experience, it could remove some of the annoyances of transacting with a local service provider. Owning the entire experience will also give Thumbtack the data to close the loop, from a request to a transaction.

Providing a way for customers and service providers to transact on Thumbtack will also help the Company increase / diversify their revenue. For now, Thumbtack sells “credits” to service providers, who redeems them in order to send customers a quote. In addition to charging for qualified leads, Thumbtack could also take a cut of the transaction from the service provider who was hired for the request. (This is something Thumbtack had tried before, but was unable to close the loop on the transaction.)

3. Using data to improve the consumer experience

Given the number of quotes Thumbtack facilitates, they should be able to provide a certain level of transparency for the consumer to get more value out of the platform. One example would be to dynamically show how many professionals fit my request — this will help set expectations on how many responses I may receive and how much supply I can expect in this market.

I’m also surprised at how little customer data Thumbtack has on me. Thumbtack is one of the few online services that doesn’t have a Facebook sign-on – which could actually tell the Company a lot about my local service needs. This type of customer data would be incredibly helpful in re-targeting me for additional services. For example, if I like Crossfit, maybe I also need a sports massage, or if I like Sephora, maybe I’m interested in beauty services. Given that I just hired movers to move from 94105 to 94607, Thumbtack could probably also estimate some basic demographic data, like my age and income bracket.

Given the market they’re facing, it’s still really early for Thumbtack and they’ve got multiple directions in which they can grow their business. With the increased competition from large incumbents such as Amazon and Google, I’m rooting for the underdog and can’t wait to see how their product will develop in the next couple of years.

The Emergence of “Pre-Seed” Capital

Notation Capital, founded by former Betaworks folks, is a new pre-seed investment fund launched a couple of months ago. 

Notation Capital is attempting to institutionalize the round of funding that’s usually done by “Friends and Family”. With just an idea and a co-founder, one could actually raise up to $500K in exchange for 5% to 10% of equity.

Really, pre-seed investing isn’t anything new. A few years ago, these rounds were simply known as seed or angel rounds, lead by early stage investors like SV Angel, Lerer Ventures, or Thrive Capital. Startup accelerator programs like TechStars and Y Combinator also use a similar model.

The thesis here is clear – firms are entering both from below and above the traditional VC-backed period of a startup lifecycle in hopes of capturing more of the returns generated by high growth companies. This is why we continue to see more large and institutionalized seed rounds (by pre-seed capital funds), as well as more large and highly-valued growth rounds (by public and/or PE funds). 

From a risk/reward perspective, I think funds with “bookend” strategies (either preceding or following the traditional VC funds) likely have lower return thresholds. What would be interesting is to see a partnership between very early and pre-IPO strategies – if one could be used to identify the winners and the other used for putting more capital behind those winners. 

Six Things Technology Has Made Insanely Cheap

I believe that technology and democratization goes hand-in-hand. If you examine the now-commoditized products listed in this article (PC, software, TV, trading commissions, camera, cell phone plan), you can also follow how such technologies became much more widespread and accessible to the masses as prices dropped.

The subtitle of the article boldly proclaims, “behold the power of American progress”! And it’s interesting to me that the author (Aki Ito) states:

For anyone bearish on the progress made by the U.S. economy, consider this: Computers are now one-1,100th of their price 35 years ago.

On the contrary, I believe that much of this price deflation actually comes from international manufacturers (read: China, India, etc.) who are able to produce virtually the same item at a fraction of the cost. 

With those two factors in mind, tech advancements and cost-efficient copycats, here are few things that I believe will face the same deflationary pressures over the next decade:

  • Mobile phones: This is a no-brainer and has already happened with the likes of cheaper Android handsets, courtesy of Xiaomi.
  • Automobiles and trucks: Asian manufacturers, such as Hyundai, are innovating quickly and will be able to rival Western brands soon in terms of quality. Furthermore, if Uber’s expansion continues world wide, demand for cheaper and more efficient cars will rise as drivers proliferate and riders opt to forgo car ownership. 
  • Education: With the current status of rising student debt, something’s gotta give. Disruptive Education Technology startups, such as General Assembly, Codecademy, and Coursera will begin to offer non-accredited alternatives to higher education. For profit education companies, like Minerva Project, will offer degrees at a fraction of what it costs today.
  • Food: I have high hopes for companies like Beyond Meat, who are looking to product petri dish-grown meat in a more cost-effective and environmentally sustainable way. Before that becomes mainstream, however, farmers will continue to lobby for government subsidies which hopefully will be passed through to consumers. 

On the other hand, there are a couple of things I wish would drop in prices, but I think unfortunately will continue to rise:

  • Healthcare:Healthcare is notoriously a laggard vertical when it comes to tech adoption, and the burden of outdated IT/infrastructure is eventually passed through to the consumer. An aging population, the impending shortage of doctors/nurses, and America’s sedentary lifestyle will all pose to be challenging to the current healthcare system. Without the right incentives for health systems and individual consumers to change their behavior, healthcare looks like it will only increase in the years to come. 
  • Housing: While this is a particularly stressful topic for those of us living in the Bay Area, I think it’s a pain point that all young adults will face sooner or later. Given high student debt and low employment, young adults will find it much more challenging to become home owners than the generation before did. 

What do you think will become cheaper or more expensive over the next decade?

Six Things Technology Has Made Insanely Cheap