Read more about how we’re thinking and where we’re investing here.

Recent Posts

The importance of checking in

Last week was a rough week. I can be a pretty open book when it comes to my personal and professional life (though I’m not going to delve into the details here). As I shared what I went through, the kindness and encouragement from family, friends, and colleagues helped tremendously. I proactively sought support from those closest to me. But every other connection spun out from an unrelated call or conversation. Many of you reached out for some other reason, and I felt compelled to share what I was going through before we dove into discussing the intended topic. After all, I wanted to give some context as to what was shaping my current perspective. And the outcome of sharing was helpful: by being honest, I felt supported and “safe” to do work without judgment. Though what I've come to realize is that while I’m open, not everyone is. It’s natural to feel embarrassed and many don’t feel free to bring their personal lives into the workplace because we don’t want to burden others or feel like it’s not “productive.” However, this is far from the case. In my experience, sharing my story with colleagues brought a deeper connection. I was open about what was going on and we were able to move forward within that context. This got me thinking about an instance where the board of directors of one of our portfolio companies was disappointed that the CEO was not prioritizing the right things. What they didn’t realize, and what I found out, was that the CEO was overwhelmed… going through a break up and family illness at the same time. Had there been a “check in”, we could have had productive discussions on how to support the CEO. Over the past few years, I try to start every conversation with a personal check-in before discussing business. The organizations that we are building and are a part of, are only as strong as the people who are in it. Perhaps this is a no-brainer given that our business is so relationship driven, but it might serve as a great reminder to all of us. In fact, as Boris wrote earlier this year while inspired by Bill Campbell, <a href="https://www.versionone.vc/bill-campbell-the-trillion-dollar-coach/">great leaders show empathy</a>. So, really ask yourself: when’s the last time that you asked your investor, your portfolio company, your colleague, your partner, your manager, your employees how are they actually doing? It’s not prying… it’s human touch which can really lift spirits and make someone’s day better (more than you’ll ever know), and inspire productivity. The more we share (regardless of power dynamic: investor-founder, manager-employee, etc.), the more context and perspective we have around the business and better understand why things are moving in a certain direction. And if this exercise seems awkward, you can try personal check-ins with a concept like “<a href="https://www.fullcontact.com/bart/red-yellow-green/">Red, Yellow, Green</a>” as indicators of mood (i.e. red = hypersensitive; yellow = a bit on edge; green = calm). So think about checking in with others, and being open to share your own experiences. And thanks to everyone who put work on hold for a few minutes last week - I really appreciate it.

Angela

July 10, 2019

Version One

The importance of checking in

Angela

July 10, 2019

Q2 2019 Summary: Portfolio News and Activities

Happy Canada Day to our friends and colleagues north of the border, and Happy Fourth of July to those of you south of the border. We hope everyone is having a great start to summer. Here’s a quick recap of the key activities and announcements from this past quarter. Welcome to the V1 family! We welcomed <a href="https://dolly.com/">Dolly</a> (same-day retail delivery service) to our portfolio, after participating in a <a href="https://t.co/MnGMmFPi5q">$7.5m Series A round</a>, alongside Unlock Ventures Partners and Maveron. Dolly is focused on delivering big and bulky items. Their fastest growing use case is furniture and other large item deliveries from big brand retailers like Lowes, Costco, and Big Lots!. We <a href="https://buff.ly/2GQpjkS">announced our investment</a> in<a href="https://kobalt.io/"> Kobalt</a>, a security as a service company. Kobalt is an exciting investment for us – it’s out first security investment (after looking for the right opportunity for some time). It’s also the first company we’ve helped incubate! We also invested in <a href="https://www.guesser.io/">Guesser</a> whose mission is to make <a href="https://medium.com/guesser/guesser-expands-product-and-announces-seed-round-to-make-prediction-markets-more-accessible-to-7f5c4e1e997d">prediction markets more accessible</a> to mainstream users. They’ve created a super simple interface to interact with Augur, in the same way that Coinbase created the easiest and most beautiful experience to buy and sell crypto currencies. And last but not least, we announced our investment in <a href="https://celo.org/">Celo</a>, an open-sourced crypto-powered global payment platform. Their mission is to create conditions of prosperity for all, providing financial applications (like cash transfer programs, P2P lending, collaborative small-scale insurance, digital assets and wallets) to the large populations on this globe who are unbanked or underbanked. Other funding announcements and exits! <a href="https://www.outreach.io">Outreach</a> raised a $114m Series E round at a $1.1b valuation - the first Version One unicorn! We couldn’t be more proud of this team. You can read <a href="https://www.geekwire.com/2019/raising-big-round-outreach-ceo-talks-post-unicorn-life-ipo-plans-creating-culture/">some thoughts from CEO Manny Medina</a> on post-unicorn life, creating culture and more. <a href="https://blockstack.org/">Blockstack</a> announced that it intends to raise <a href="http://ow.ly/Q0Kx50pRGxC">$50 million in a token sale</a> that would leverage the SEC’s Regulation A+ crowdfunding exemption. And, <a href="https://www.twenty20.com/">Twenty20</a> is joining the <a href="https://envato.com/blog/twenty20-joins-the-envato-family/">Envato ecosystem</a>. Product launches and partnerships <a href="https://www.ada.support/">Ada</a>, a market leader in automated customer experience, <a href="http://share.ada.support/p7znk">launched Ada Glass</a>. This provides a seamless customer handoff from Ada’s AI-powered chatbot to live chat platform partners including Zendesk, SalesForce, and Nuance. <a href="https://goshippo.com/">Shippo</a><a href="https://hubs.ly/H0jnl0T0"> announced their UPS Ready Program</a> and Fedex Compatible partnerships. <a href="https://www.abstract.com/">Abstract</a><a href="https://www.abstract.com/blog/sdk/"> released their SDK</a>, so that you can start building your dream integrations. <a href="https://www.manifold.co/">Manifold</a> launched their “<a href="https://tcrn.ch/2ZbrpBY">Marketplace as a Service,”</a> making it incredibly easy for developers to add apps to any platform. <a href="https://www.lolli.com/">Lolli</a> launched its <a href="https://www.producthunt.com/upcoming/lolli-for-firefox">Firefox extension</a> (to add to the Chrome extension). <a href="https://www.coinbase.com/">Coinbase</a> expanded its cryptocurrency <a href="http://ow.ly/p25U50uBXNi">Visa Debit Cards across Europe</a> and also <a href="https://blog.coinbase.com/expanding-crypto-to-crypto-support-to-more-countries-around-the-world-42905b733a30">welcomed 11 </a><a href="https://blog.coinbase.com/expanding-crypto-to-crypto-support-to-more-countries-around-the-world-42905b733a30">countries</a> in Latin America and Asia. Coinbase also now supports <a href="https://blog.coinbase.com/30-assets-now-supported-on-coinbase-custody-ae9816ec2bf9">30+ assets on custody</a>. <a href="https://blockstack.org/">Blockstack</a> released their <a href="https://blog.blockstack.org/announcing-the-blockstack-whitepaper-2-0/">Blockstack whitepaper 2.0</a>, which builds on two years of R&D work, peer-reviewed research, and presents a more detailed specification of their technology. <a href="https://www.dapperlabs.com/">Dapper Labs</a> (who brought us <a href="https://www.cryptokitties.co/">CryptoKitties</a>) released <a href="https://www.cheezewizards.com/">Cheeze Wizards</a>…the world’s first blockchain battle royale with cheese. In Cheeze Wizards, players summon Wizards to fight in magical duels and compete to become the Big Cheeze. And finally, we loved these books and sent them to our founders… it you find yourself with some extra time this summer, we recommend reading them! <ul> <li><a href="https://www.versionone.vc/bill-campbell-the-trillion-dollar-coach/">Trillion Dollar Coach</a>, from Eric Schmidt, Jonathan Rosenberg and Alan Eagle </li> <li><a href="https://www.versionone.vc/finding-and-turning-your-flywheel/">Turning the Flywheel</a>, Jim Collins’ monograph to accompany his classic business book, Good to Great</li> </ul> Happy Summer to everyone!

Angela

July 2, 2019

Version One

Q2 2019 Summary: Portfolio News and Activities

Happy Canada Day to our friends and colleagues north of the border, and Happy Fourth of July to those of you south of the border. We hope everyone is having a great start to summer.

Angela

July 2, 2019

Thinking about a managed marketplace for healthcare

A few years ago, I explored the idea of a <a href="https://www.versionone.vc/healthcare-update/">healthcare marketplace model</a>. At the time, I imagined it would be pretty difficult to build a true marketplace for several reasons: low frequency and high unpredictability of use (because thankfully, we’re not always sick), the monogamy issue (people prefer to stick to the same doctor once they find one), and most transactions are handled offline and won’t be captured by the platform. These factors are just as valid today as they were back then. In fact, it seems that the common destiny for true healthcare marketplaces we’ve seen has been a directory of providers. However, Boris’s recent post on <a href="https://www.versionone.vc/what-exactly-is-a-managed-marketplace/">what exactly is a managed marketplace</a>, as well as some recent conversations with entrepreneurs, have made me reflect on what might be possible. Could a managed marketplace model work for healthcare? Maybe it starts off by serving basic health needs, then expands into more complex ones. In particular, I’ve been thinking about three key elements: <ol> <li>Aggregate and curate supply</li> </ol> A managed marketplace can aggregate and curate a pool of providers. I could see the marketplace showing which physicians have last-minute openings – i.e. what types of doctors can patients see this afternoon? And this “available inventory” can be updated on a daily (or more frequent) basis. I can see this system working particularly well for non-complex cases from colds and flus to UTIs, and teeth cleaning to botox injections. <ol start="2"> <li> Manage the pricing for services on the platform</li> </ol> In most cases, patients don’t necessarily want to haggle on pricing when it comes to healthcare. With that said, a “Priceline for healthcare” might not be a great UX for patients who likely don’t know how much a certain service or procedure should cost. In this sense, I think of the UX of Uber and Lyft, and its resulting success (i.e. both companies set the price for rides vs. the more Wild West feel of Sidecar where each driver set his or her own price). Perhaps a healthcare marketplace can manage and standardize the prices on its platform – and even advertise discounts when “inventory” is expiring (i.e. a provider needs to fill a few more slots for the day)? <ol start="3"> <li> Connect the patient to the physician after the service is booked</li> </ol> When a marketplace standardizes the price of a service or procedure, patients won’t think to correlate it with quality if every physician is charging the same. As such, the marketplace can assign patients to available physicians, just as Uber/Lyft assigns riders to drivers. When there is no choice for specific supply, this can be advantageous for several reasons. First, it ensures that communication and transactions between patient and doctor remain on the network. This, in turn, might address the interoperability issue and poor data sharing – a big problem in healthcare. In addition, this model can reduce the chance of leakage since patients are more likely to associate the service with the marketplace brand, rather than an individual provider. And patients can always find a cheaper price going through the marketplace since supply is based on inventory that is expiring. From here, once you have built a high quality network of physicians and won the mindshare and trust of patients, you can imagine the potential expansion into more complex cases and advance appointments. And perhaps, with this underlying network, you can even become a new type of insurance! Now, a managed marketplace might not overcome all the challenges of scale, but it seems promising, and I think it has a strong potential to challenge the high cost of healthcare. Is anyone else thinking about this business model? If so, we’d love to talk to you!

Angela

June 27, 2019

Version One

Thinking about a managed marketplace for healthcare

A few years ago, I explored the idea of a healthcare marketplace model. At the time, I imagined it would be pretty difficult to build a true marketplace for several reasons: low frequency and high unpredictability of use (because thankfully, we’re not always sick), the monogamy issue (people prefer to stick to the same doctor once they find one), and most transactions are handled offline and won’t be captured by the platform.

Angela

June 27, 2019

Version One

Balancing the upside and downside of data sharing

This year we have seen privacy become top of mind for the general public, the media and “big tech.” The business model of Google and Facebook (collect as much data as possible about users for highly targeted advertising) is under attack. Europe’s legislators were addressing some of these privacy concerns when they put into effect GDPR in 2016, the primary law regulating how companies protect EU citizens’ personal data.

Boris

June 18, 2019

Version One

Finding and turning your flywheel

Good to Great is one of those iconic business books that every founder, CEO and business leader should read at some point. In the book Jim Collins identifies eight common traits that “good to great” companies share.

Boris

June 13, 2019

How close are we to a smart toilet?

Imagine a future, maybe 20-25 years from now, when there’s a smart toilet in every home. Picture it: the smart toilet, a device that collects our fecal matter and provides information about our health via our microbiome. The human gut microbiome is a complex ecosystem (consisting of trillions of bacteria, viruses, fungi) that is receiving increasing attention for the role it plays in our health. Gut microbes help regulate our metabolism, ward off infections, break down fiber, and may be biomarkers of health and disease. <a href="https://gut.bmj.com/content/67/9/1716">Numerous studies</a> have shown changes in gut microbes during obesity, diabetes, liver disease, cancer and neurodegenerative disease. Yet despite this, we are still far away from a smart toilet. Yes, some microbiome tests exist today. And some companies claim to improve your health by prescribing probiotics and suggesting diet changes based on your fecal sample. But so far, these “solutions” have little backing from the medical world. Few physicians order these tests for patients and few payers reimburse. But with that said, what’s important is that we are starting to collect information about our microbiome. Today, there’s sound science for sequencing fecal samples to give you a count of specific bacteria. However, there’s very little scientific evidence correlating bacteria to your health. That’s because gut health is all relative: what is healthy for you is different from what is healthy for me. As such, you need a frequent and long-term sampling of your gut microbiome in order to track how any changes correlate to any symptoms you are experiencing. But herein lies the problem: just collecting one fecal sample is full of friction (who likes touching their feces?) – so imagine how unpleasant the UX is for collecting longitudinal data. Working backwards, we can think about how to get to a smart toilet… If the goal of the smart toilet is to tell us about our gut health so that we can make changes to our lifestyle and be healthier, then what we need are: <ul> <li>Implanted assays in the toilet for immediate analysis/diagnosis, which require:</li> <li>FDA approval for these assays developed as a result of proving a correlation between the microbiome and certain diseases, which is possible when there is:</li> <li>A frictionless sample collection device that can collect longitudinal data for researchers to conduct their microbiome studies.</li> </ul> And what might a smart toilet business look like? Is it vertically integrated? Could the toilet be like a smartphone (hardware) with a software platform where people can build “apps” (assays) to test for particular diseases? Should the company be B2C so that it can completely own the data? That seems logical, but how does one launch a B2C if there’s no immediate value (i.e. correlation to health) to provide the customer on Day 1? What about a B2B model? Does this make more sense at first? Perhaps it starts with empowering researchers to find a correlation between the microbiome and disease with an easy-for-patients-to-use tool. On this point, companies like <a href="https://www.biomesenseinc.com/">BiomeSense</a> are building a collection device that attaches to the toilet so patients don’t have to swab themselves or defecate in a cup. The question then becomes how one goes beyond being just a collection company (which would likely become a commodity)? What value-add can be provided? Perhaps analysis? Clearly, this is all a simplified look and there are many more steps involved, questions to answer, and challenges to consider. And the entire model is precipitated on the fact that we believe in the potential of microbiome to tell us about our metabolism and health. I, for one, am convinced that this hypothesis is true. Is anyone thinking about this? I’d love to speak with any entrepreneurs, investors, scientists, and doctors, who have a similar vision.

Angela

June 5, 2019

Version One

How close are we to a smart toilet?

Angela

June 5, 2019

What we can learn from high performing athletes

I have been on cloud nine since Saturday night. As a long-time Toronto Raptors fan, my hometeam is headed to the NBA Finals for the first time in franchise history. Throughout this playoff run, I have been thinking a lot about high performing athletes, and what we can all learn from a champion’s psychology. One moment in particular stuck out for me. Two weeks ago during Game 7 of the Eastern Conference Semifinals against the Philadelphia 76ers, Toronto Raptors superstar Kawhi Leonard hit a legendary game-winning shot. Sixers fans may not appreciate the video replay, but <a href="https://www.youtube.com/watch?v=ChT3ewZXTfM">here it is</a>. Kawhi made this shot after missing a similar one just a couple of possessions before. In his <a href="https://www.youtube.com/watch?v=YUvSaaOEKGE">postgame interview</a>, Kawhi was asked by a reporter to walk through that final shot. His response: his team had run a similar play during the previous playoff series against the Orlando Magic. He remembered that moment, knew there was a little more time on the clock this time around, and was optimistic about his chances (at least optimistic for Kawhi, who is known for expressing little emotion). Turns out, visualization is a very powerful tool in sports psychology, used by elite athletes from LeBron James to Lindsey Vonn. It’s a way of conditioning the brain for successful outcomes by creating a mental image of what you want to happen and feel. Without realizing it, people are programming their subconscious brain all the time with negative: they replay past mistakes or critiques and visualize all the negative scenarios that might happen. The process of visualization and guided imagery consciously controls the script running in your head…in these scenarios, you execute flawlessly, dominate the competition, reach the summit, etc. In short, you’re programming your subconscious for success. How can we apply visualization in the tech world? When I see Kawhi hit clutch shots and read stories about LeBron watching <a href="http://fortune.com/2015/05/12/lebron-james-attitude-success/">a highlight reel of himself making threes every night</a>, I can’t help but wonder what we in the tech world can take away from these athletes. We pour so much of our self worth into work and focus on what we want to become. But too often, we don’t pause to reflect on how far we have come. How often do you hear about someone suffering from burnout, imposter syndrome or some other non-productive and harmful metal state? Entrepreneurs are famous for their drive and hunger. These are great qualities, but not when the focus is always on your shortcomings. Reflecting on our journey and successes can be the best fuel to perform even better. To this end, I’ve been thinking about the tech world’s equivalent of guided visualization. Recently, my father suggested that I write down my achievements, big or small, every quarter. This list can be a good reminder of one’s track record of tenacity and hard work. You can do this as an individual, as a team, as a company. Another great recommendation comes from one of V1’s CEOs: when you are journaling, take a moment to flag the positives. Then go back to these excerpts from time to time. Each individual achievement, like every shot that Kawhi has made until his game-winner, might not stand out on their own, but they can give us (and him) the confidence to reach for the next milestone when it matters. What do you do to prepare yourself mentally to perform even better than you already do? I would love to learn what other suggestions our readers have. P.S. Go Raptors! P.P.S. If you haven’t caught this story on what makes Toronto and Canada special, I recommend learning about <a href="https://twitter.com/MuhammadLila/status/1132705598632538113">Raptors super-fan Nav Bhatia. </a>

Angela

May 28, 2019

Version One

What we can learn from high performing athletes

I have been on cloud nine since Saturday night. As a long-time Toronto Raptors fan, my hometeam is headed to the NBA Finals for the first time in franchise history.

Angela

May 28, 2019

What does a Head of People do? Learning from Ada’s Chelsea MacDonald

Over the past few months, we have been running a series of posts where we profile one of the many incredible operators in our network. These operators are generous enough to share their expertise and wisdom, and we are lucky to do deep dives into unique topics. We’ve already learned about <a href="https://www.versionone.vc/3qw-bala/">marketplaces</a> from <a href="https://www.linkedin.com/in/balachandergs/">Bala Subramaniam</a>, Head of Fulfillment at Instacart. And <a href="https://www.linkedin.com/in/sinmeitsai/">Sin-Mei Tsai</a>, VP of Engineering at Shippo, discussed <a href="https://www.versionone.vc/sin-mei-code-quality/">code quality</a>. In this installment, I’m thrilled to introduce <a href="https://www.linkedin.com/in/chelseamacdonald/">Chelsea MacDonald</a>, Head of People Operations at <a href="https://www.ada.support/">Ada Support</a>. Chelsea started her first business when she was 12, and her first non-profit when she was 13. She was a data analyst and has been in People and leadership roles since 2011. She has grown teams from 140 to 550 and from 20 to 60 (twice). And she has just about completed 60 to 140 (<a href="https://boards.greenhouse.io/ada">Ada is hiring</a>!). She loves sneakers, travel (50+ countries) and running very, very fast. Unlike “traditional” departments such as Engineering, Product, Sales, Marketing, BD, Customer Support/Success, People is a relatively new leadership function. That’s why we’re so excited to learn about it from one of the best! What does a Head or VP of People do? What are your responsibilities and priorities? How is it different from HR? Chelsea: A Head of People sets an organization up to do their best work in a way that scales. It’s probably the role in an organization that requires the broadest and most diverse skill set, because you’re looking for skills in: marketing & PR (employee branding, communications, events), BDR and sales (recruiting), product (employee experience, diversity, data), customer success (employee performance), and legal (HR law, terminations). Throw in a bit of wellness coach, and you have yourself a role for a renaissance person (or alternatively, someone who knows how to hire for their weaknesses). The difference between HR and People is really the difference in priorities. The question that keeps me up at night is: “Is everyone at Ada working on the most important things, at the edge of their abilities?”. I think traditionally with HR, what I’m supposed to worry about is if we’re going to get sued :) I still worry about that, but the cost of a lawsuit is drastically less than the cost of our entire team working on the wrong things, or not working to the best of their abilities, so it’s not my first priority. As a snapshot, here’s what one of my most recent days looked like: <ul> <li style="font-weight: 400;">Ran reporting on OKR updating and progression</li> <li style="font-weight: 400;">Spent 45 minutes thinking through how to make it easier for people to update their OKRs and incorporate it into 1 on 1s</li> <li style="font-weight: 400;">Interviewed a candidate for Head of Finance</li> <li style="font-weight: 400;">Planned 2020 future salaries projections</li> <li style="font-weight: 400;">Reviewed performance of BDRs and Sales</li> <li style="font-weight: 400;">Reviewed data on bot conversion, had a discussion about % of future that’s managed services vs product</li> <li style="font-weight: 400;">Spent an hour talking to people about an org change (3 separate conversations)</li> <li style="font-weight: 400;">Did a 45 minute review of our IT infrastructure</li> <li style="font-weight: 400;">Ran reporting on recruiting pipeline</li> <li style="font-weight: 400;">Re-hashed salary bands for 2020</li> <li style="font-weight: 400;">Posted new coaching pilot to our leadership team for feedback</li> <li style="font-weight: 400;">Posted communications plan for marketing and product features launches for feedback</li> <li style="font-weight: 400;">Talked to a lipstick company about Ada-branded lipsticks for recruiting events</li> <li style="font-weight: 400;">Shared my thoughts with you! </li> </ul> What does a Head or VP of People not do? Chelsea: Here are three things that a Head/VP of People shouldn’t do: 1) Don’t just blindly follow HR best practices. I remember very clearly my first moment of imposter syndrome around HR things. I was a data analyst masquerading as a people person, and so I went to our local HR association to learn what I didn’t know. When I asked about the evidence behind best practices, they looked at me blankly. Thankfully, since then, HBR, Google and others have led the charge on measuring outcomes of many best practices. Even so, always ask yourself how this would be unique inside your organization, and how you’ll measure the success of anything you try. 2) Don’t expect people to do something because they’ve been told to do it. We all know we should eat well, exercise and sleep more. Most of us don’t do it. Telling people how to be a better manager doesn’t work if you don’t figure out how to support the behaviour change. You have to understand what are someone’s competing priorities, and how does the change you’re asking for fit into their day-to-day. Then you make it as easy as possible for them to make that change. 3) Don’t do it yourself. When things are moving really quickly, the People team can sometimes try and force the function by just doing it themselves. To scale, you need to be giving people the tools to drive their own performance and the performance of their teams. While you might get a momentary lift from doing something for someone, you won’t succeed long-term unless you make it easy for everyone to do it for themselves. When and how do you measure success? Chelsea: Often and with a grain of salt. For people initiatives, the data sets are often small, and the link between correlation and causation is often murky. The data tells you where to start asking questions, but it’s rarely a penultimate answer until you’re at scale (and even then, only if you’ve built the tools to measure across the organization). Examples: <ul> <li style="font-weight: 400;">I measure long-term performance of hires based on source. It actually broke all of the recruiting norms. The highest performing people at 12 months were people who applied for jobs (though they were the poorest performers at 3 months). Employee referrals were the lowest performing at 12 months, the highest at 3 months. So basically, the bias of employee referrals was super strong, and the key to accelerating performance is actually relationships. The dumbest questions answered the quickest! I have no idea if this replicates itself across multiple organizations (though I’m measuring now!)</li> <li style="font-weight: 400;">Measure time to deliver, and not time to hire. So how soon after we decide we need someone is someone excelling in that role?</li> <li style="font-weight: 400;">I measure everything against work outcomes. The point of management training should be the teams that receive the training perform better. If there’s no pick up in OKR outcomes, then there’s likely no point in management training aside from making people feel good (which could be enough reason to still do it)</li> </ul> When should an organization bring on a Head of People? Chelsea: It depends on how quickly you’re growing. It’s the point in time where you’re no longer confident that your team is actually working on the most important thing. Usually it’s around 60 people for fast growing organizations, but you should plan on it taking about 6 months or more to find the right person, so you might want to start early. What are the qualities/skills/experiences that an organization should consider for a candidate in this position? Where in the organization does this role sit? Chelsea: You want to find someone that’s business-focused and data-focused. If you ask this person about their impact on the business, they should be able to answer it well! They’ll also probably tell you about all of the things they tried they *thought* would impact the business that didn’t work out. Also, you need a builder. A lot of HR work is going to be setting up repeatable systems. If they aren’t operationally minded, you’re going to end up with a ton of bloat in your organization. I’ve always reported to the CEO. It means I can talk to other leaders as a peer. What is the most challenging part about your work and how do you overcome it? Chelsea: With all of the new focus on culture, this can often get cheapened to employee happiness. Is everyone happy? Are they playing enough foosball? Did they show up at the employee event? Do you want miserable employees? No. But you want uncomfortable employees. Challenged employees. Employees who feel like they’re on the edge of what they’re capable of. Managers who will ask for a bit more than what someone thinks they can do. Some work days will be scary. Sometimes you’ll disagree with your colleagues, your manager or yourself. It’s easy-ish to make someone happy; it’s very hard to make them fulfilled. And it’s very difficult to figure out where that edge is, and how to build the trust and relationships that make that a mostly enjoyable experience. My tip is to be transparent with people about what you’re trying to do, and listen when they feel like you’ve pushed too far. Any tips on diversity/inclusion? Chelsea: Raze it all. :) Really. You can’t get to diversity and inclusion through nudges. There are systems in place that have continually shown us to produce biased results. For Q2 this year, we wanted to include Diversity and Inclusion as a focus, but it wasn’t going to be P1, and so we’re waiting for Q3 when it can be P1. We’re going to redesign a hiring process from scratch that’s built for diverse candidates. I don’t actually know if that’s going to work, but I’ll let you know!   *** Many thanks to Chelsea for taking the time to share and deepen our knowledge on this emerging role. As always, please feel free to add comments or ask additional questions in the Comments section below!

Angela

May 23, 2019

Version One

What does a Head of People do? Learning from Ada’s Chelsea MacDonald

Over the past few months, we have been running a series of posts where we profile one of the many incredible operators in our network. These operators are generous enough to share their expertise and wisdom, and we are lucky to do deep dives into unique topics. We’ve already learned about marketplaces from Bala Subramaniam, Head of Fulfillment at Instacart. And Sin-Mei Tsai, VP of Engineering at Shippo, discussed code quality.

Angela

May 23, 2019

Version One

Announcing our investment in Dolly, a peer-to-peer marketplace for on-demand moving and retail delivery

We are excited to announce our investment in Dolly, where we participated in a $7.5M funding round alongside Unlock Venture Partners and Maveron.

Boris

May 15, 2019

Version One

Bill Campbell, the trillion dollar coach

Bill Campbell has long had a legendary reputation as one of the best coaches of all time, but until now, only the people he worked with got to understand and experience the power of his approach.

Boris

May 7, 2019

Announcing our investment in Kobalt, a security as a service company

We are very excited to announce our investment in <a href="https://www.kobalt.io/">Kobalt</a>, a new security as a service company. Security is increasingly on everybody’s mind, with several high profile, large scale hacks over the past few years. Here at Version One, we had our<a href="https://www.versionone.vc/learn-from-our-nightmare-tips-to-prevent-getting-hacked/"> own unfortunate incident</a> (and learning experience) only a few weeks ago. Yet while security is a real concern for all, the reality is that many small and medium sized organizations don’t have the resources to have a full-time security expert on staff to monitor security incidents. This is where Kobalt comes in: Kobalt levels up your security efforts by continuously monitoring for security incidents so that companies can respond quickly and effectively to incidents before they become security nightmares. There are two important aspects about this investment for our firm. One, this is our first security investment. We have been looking at this space for a while, but never found the right entry point. I am excited that we finally are active in the security vertical and hope to invest in more security start-ups down the road. Second, Kobalt is the first company we have helped incubate. Incubation requires much more hands-on involvement than simply investing in a company. I don’t anticipate that we’ll do many incubations. But, this was a unique chance to work with one of Vancouver’s best entrepreneurs. <a href="https://www.linkedin.com/in/pankajoptimus/">Pankaj Agarwal</a> has built several very successful services businesses in the tech space, including <a href="https://www.optimusinfo.com/">Optimus Information</a> and<a href="https://www.trinimbus.com/"> Trinimbus</a>. And, last but not least, we got incredibly lucky to find a superbly talented team to start this company: <a href="https://twitter.com/michaelargast">Michael Argast</a> and <a href="https://twitter.com/jefflaporte">Jeff La Porte</a>. Michael has spent almost all of his career in the security space, working for Sophos and Telus, among others. Jeff La Porte is one of the best start-up CTO’s I had the privilege to work with in the past. Kobalt launched their service only a few weeks ago and initial market reaction has been very positive, especially from enterprise SaaS companies that are getting increased scrutiny from their clients about their security standards. If you run a SaaS company (or any other company that needs to level up its security in a cost-efficient way), please email Michael at michael [at] kobalt [dot] io or <a href="https://landing.kobalt.io/launch-version-one">sign up for a 1:1 meeting here</a> and we can get you more information about Kobalt and how we might be able to help you. As the <a href="http://kobalt.io/why-kobalt/">Kobalt team says</a>, every leader deserves a good night’s sleep and every organization deserves great security.

Boris

May 1, 2019

Announcing our investment in Guesser, the easiest way to engage with prediction markets

We are extremely excited to announce our investment in <a href="http://www.guesser.io">Guesser</a>, the easiest way to engage with <a href="https://www.augur.net/">Augur</a> prediction markets. Among all the crypto opportunities, we are most excited about<a href="https://www.versionone.vc/crypto-native-applications/"> crypto native opportunities</a> – and prediction markets are clearly native. We spent some time with Augur right after it launched (you can read about that <a href="https://www.versionone.vc/a-few-take-aways-from-the-augur-prediction-market/">initial experience here)</a>. The potential to create millions of markets around any potential outcome in the world – from sports to elections, weather and stock prices – is pretty amazing. When I first engaged with Augur eight months ago, two key challenges stood out: 1) the UI / UX was complicated and 2) there were questions about regulation. Guesser has an answer to both challenges. It has created a super simple interface to interact with Augur, in the same way that Coinbase create the easiest and most beautiful experience to buy and sell crypto currencies. Guesser can make investing in prediction markets as attractive and frictionless as investing in coins and stocks. They began with just a few markets – but it was enough to see the potential. And they’ve just expanded to four dedicated sections for weekly events on Crypto (BTC, ETC, Maker...), Finance (IPOs, derivatives...), Politics and General News. Secondly, the Guesser team is based in Madrid and European regulators are much more welcoming to crypto start-ups and online betting activity (which some of the activity on Guesser could be characterized). We led this round of investment (a $1.1m seed round) together with our friends at Compound with participation from Boost VC. It is very early days for decentralized prediction markets, but we are excited to see the Guesser team shaping some of these opportunities going forward.

Boris

April 24, 2019

Version One

Announcing our investment in Guesser, the easiest way to engage with prediction markets

We are extremely excited to announce our investment in Guesser, the easiest way to engage with Augur prediction markets.

Boris

April 24, 2019

How Sin-Mei Tsai, Shippo’s VP of Engineering, Defines Code Quality

Six weeks ago, we kicked off a new series where we profile an awesome operator in our network to learn about his/her work and contributions to his/her organization and community at large. Our inaugural guest was <a href="https://www.linkedin.com/in/balachandergs/">Bala Subramaniam</a> and we asked him <a href="https://www.versionone.vc/3qw-bala/">three questions on marketplaces</a>. This time, I am very excited to introduce <a href="https://www.linkedin.com/in/sinmeitsai/">Sin-Mei Tsai</a>, VP of Engineering at <a href="https://goshippo.com/">Shippo</a>, a Version One portfolio company. Over the years, I have had the pleasure of getting to know Sin-Mei over many lunches. She is a Silicon Valley veteran - having been the Senior Director of Engineering at Yodlee, then holding several CTO, Chief Architect, and other engineering leadership roles at a few startups before landing at Shippo in 2016. Without saying, we were thrilled to have her join the team because of her experience. And she has not only been instrumental to Shippo, but also an extremely valuable resource to other engineering leaders in the Version One family. In fact, every time there is an engineering-related question, the first person I go to is Sin-Mei. With that said, I thought it would be great to share Sin-Mei’s insight and wisdom with our blog community. Instead of asking three questions as we did with Bala, we’re going to take a very deep dive into a critical topic: code quality. What follows below are Sin-Mei’s words on how she (and Shippo) define code quality. As you’ll see, she is an incredible resource – so if you have any questions or thoughts on this topic, please leave a comment below! ********** How do you define code quality? At Shippo, we give a lot of thought to our interview questions. When we interview managers, one of our standard questions is “How do you think about code quality?” or “What does code quality mean to you?” Recently, I was pleasantly surprised to be asked this question by a candidate. Typically I answer these questions spontaneously, reflecting my stream of consciousness, and I liked what came to mind. At Shippo, we think of code quality from 3 different points of view. One of them is the orthodox definition of code quality, but we have expanded on code quality with a couple of points of views of our own. Functional quality The first and foremost category of quality for engineers is the mainstream definition of quality: “Did the code work as designed?” Most quality assurance practices address functional quality by identifying defects in code. Defects are cheapest to address when found early. For that reason, unit testing is very important. At Shippo, we focus our unit testing on functional testing. Instead of testing whether the code behaves as coded, we focus on whether the code behaves as the user would expect. We try to use mock objects only when we must, such as in stubbing out an external dependency. In our legacy codebase in Python, an interpreted language, the engineers rightly wanted to focus on code coverage. For an interpreted language, code coverage can help ensure that each line of code has been exercised and does not contain syntax errors. With a compiled language like Go, which is the language we are using in our new platform, we have the advantage that syntax errors are caught by the compiler. This allows us to focus on functional testing, which is a hybrid of unit testing in that the test focuses on a functional unit, a function, and black box testing where we exercise the function as the caller without mocking any internals. To help us perform functional testing without mock objects, we maintain a seed database that reflects the database schema and contains the data definitions that we would expect in production. This seed database provides the initial data set for any local and integration testing environments. Besides unit testing, engineers also perform manual developer end-to-end testing. This end-to-end testing is performed to ensure that when we integrate different pieces of code, the code behaves as expected. It is in manual integration testing that we test some tricky combinations. Sometimes we find that the new code does not play well with old state, or with old code, which is a transient problem that happens during deployment. To make doubly-sure that code behaves as expected, we also employ independent testing during the quality assurance phase. Even though developers are required to thoroughly test their code and regression test any impacted functionality, developers tend to be myopic in their testing. Developers will test from their own point of view, from having written the code. An independent tester will test cases that are unexpected. One such example that we ran into in the past was a new function to add addresses to an address book. An independent tester thought of the case to remove the last remaining address, whereas the developer assumed that it couldn’t happen because the requirements stated that it shouldn’t happen. At Shippo, the engineers take on the role of testers and test the code developed by their fellow engineers. So long as the engineer didn’t contribute to the code being tested, they are eligible to be the independent tester. For API changes, we perform parity testing where we record API requests and responses from production, and then playback the recorded API requests against the new code and compare the responses from the system under test against actual responses recorded from production. Because the API is stateful (a POST or PUT changes the state), we have developed heuristics for performing the comparisons in a parity test. This is especially important because we support many versions of the API, including some with unexpected side effects or object representations, such as in the case of null vs. empty object. Technical design quality At Shippo, we also consider tech debt in our definition of code quality. Architectural and software design integrity is very important to us. The question that we ask is: “In adding or changing a piece of code, did we create any intentional tech debt?” Not all code needs to be scalable. Sometimes, we intend to create a rapid prototype to gauge user interest in a new feature. Once the prototype has served its purpose, it can be thrown away. Code needs to be written considering its longevity in mind. Code that is intended to last longer than a prototype should be written in such a way that future developers can easily change or extend it. Therefore, readability and maintainability are important. When the code we write is meant to last more than the next year or two, we rely on technical designs to ensure the code ages well. Technical design is one of the most rewarding parts of code development. Nothing feels as satisfying as an elegant design. While elegance may be subjective, a formal peer design review process can help steer a technical design in the right direction. In a design review, we evaluate the proposed solution against the problem that it is meant to solve. In reviewing a design, we start with the end first. What will it be like to maintain the solution? If the existing infrastructure can already address the new problem, we want to understand whether the existing infrastructure or patterns should be reused. Creating unnecessary new infrastructure or patterns will create additional maintenance overhead for the team and reduce our ability to create value. If a new solution provides enough additional benefit over the old solution, then we might evaluate moving the old code to the new infrastructure. In a peer design review, the designer shares with the team the benefits and drawbacks of the proposed solution, as well as any other solutions that were evaluated. Typically, the solution will be presented in graphical form, such as a system architecture with flow diagrams. And the reviewers will provide input with regards to scalability, fault tolerance, and maintainability. Components that need to handle large volume need to be designed for scale. Post development, we make use of performance (stress) testing to ensure that the new component can withstand multiples of the expected volume. The engineer is responsible for understanding the load, for example, the mixture of API calls that the new system will receive once in production. The performance test entails comparing the performance characteristics of the old system and the new system. Sometimes, we will evaluate two designs using performance testing to verify that one design scales better than the other. Value to customers Above all, the purpose of product development is to create value for customers. Therefore one of the most important ways to evaluate code is whether it provides value to customers. The way we ensure that code will provide value for customers is by relying on customer-centric practices in creating our product development plans. Before even a line of code is written, our product team conducts customer research to get and vet product ideas. The engineer’s job is first to understand the customer problem that we are trying to solve, and then to construct the solution for the problem. When building solutions, we are most interested in knowing whether the solution works for the customer. Whenever possible, we favor shipping an early iteration of the solution to get customer feedback. Once the first iteration is in production, we observe customers using our product and collect customer feedback to gauge whether the solution is useful or desirable, whether it is easy to discover and use, and whether the user wants more of it. If the actual usage of the functionality meets the expected usage and engagement, then we know the code has met the most important quality metric of usefulness. Our automated emails are a good example of a useful feature -- we are constantly releasing variations and improvements to these emails. As we confirm whether the solution works, we have more confidence in shipping follow-on iterations of the solution. For us, shipping follow-on iterations is a sign of a successful feature. Conversely, sometimes a new solution can create unexpected outcomes. New functionality can get in the way of something users need to do. Even worse, sometimes new functionality can introduce unintended side effects and detract value or erode trust. We consider these to be cases of bad code quality. To limit negative effects that arise from releasing new functionality, we roll out code in small increments. For critical functionality, we may start with 5% of the user population and slowly step up to 10%. We rely on instrumentation to provide us visibility as to how new functionality works for users, and as soon as we detect issues, we either fix the issue quickly or dial back the % to 0 if a fix cannot be released quickly. People-first At Shippo, we are known to be people-first. Our priorities are no different when it comes to code quality. In considering value to customers and technical design quality, we are keeping the customer and the future developer in mind when it comes to code quality. In a recent conversation with our VP of Product, Yotam Troim, we talked about conducting post mortems on projects that run later than expected. It occurred to me that late code is a form of bad code. Perhaps this can be a topic for future conversation on code quality.

Angela

April 11, 2019

Read more about how we’re thinking and where we’re investing here.

Recent Posts

The importance of checking in

The importance of checking in

Q2 2019 Summary: Portfolio News and Activities

Q2 2019 Summary: Portfolio News and Activities

Thinking about a managed marketplace for healthcare

Thinking about a managed marketplace for healthcare

Balancing the upside and downside of data sharing

Finding and turning your flywheel

How close are we to a smart toilet?

How close are we to a smart toilet?

What we can learn from high performing athletes

What we can learn from high performing athletes

What does a Head of People do? Learning from Ada’s Chelsea MacDonald

What does a Head of People do? Learning from Ada’s Chelsea MacDonald

Announcing our investment in Dolly, a peer-to-peer marketplace for on-demand moving and retail delivery

Bill Campbell, the trillion dollar coach

Announcing our investment in Kobalt, a security as a service company

Announcing our investment in Guesser, the easiest way to engage with prediction markets

Announcing our investment in Guesser, the easiest way to engage with prediction markets

How Sin-Mei Tsai, Shippo’s VP of Engineering, Defines Code Quality

LinkedIn Posts

Subscribe to Blog via Email