Ghosts

Fear and Loathing on a dark, rainy NYC Night

For the first time in over a year, I’m resting my elbows on the carved wooden bar at the hole in the wall, only in NYC dive bar underneath our (now former) office. 

Undercover cops shot a black security guard on the doorstep in 2000, after the security guard refused to sell them pot. The Grand Jury declined to indict them. Anthony Bourdain,  the Hunter Thompson of our time, drank here. 30,000 New Yorkers have died of COVID since I’ve last sat at this bar.

This bar has five different names — on the sign, on google, inside, etc, including one calling it a “distinguished cocktail lounge.” I recommend buying only bottled beer and paying only in cash. 

This bar is a survivor, between a LensCrafters and a Dos Toros. It’s one of three businesses on this block older than 10 years, the others being a hardware store, and a sex shop that has a sign proclaiming its “busines hours.” (sic)

This bar is a survivor, but in between the (mostly) socially distanced bar seats, are ghosts. Of Bourdain, of Patrick Dorismond, the father of two murdered by undercover, unidentified cops, of the 30,000+ New Yorkers killed by “just the flu.”

Much like scars, the ghosts never go away, but they make us who we are as a city. On dark, rainy nights like this, where we can uneasily re-experience something we took for granted for so long, I remember them.

Survival Bias in Startup Advice

It's not always a fairy tale ending

Lots of startup advice is told like a war story. Founders of (now) successful, thriving companies talk about the dark days, full of all nighters, personal debt, stalled growth, and low runway. 

These war stories generally include advice like “never quit” and “keep on pushing.” The storytellers talk about taking huge personal risks, working themselves to the bone, and approaching their mental breaking point. 

These stories all have the same ending. The efforts paid off, the company turned around, and the founders did (more than) well. They look at these dark days almost nostalgically, and talk about how they overcame them on Twitter or over drinks.

These stories all have happy endings because the companies (and their founders) survived. Very few people look to advice and stories from unsuccessful founders of failed companies. 

Companies that ultimately fail have plenty of dark days as well. Their founders, in many cases, work just as hard and take on just as much personal risk as founders of companies that work out. However, their stories end in a range of begrudging acceptance to soul-crushing sadness. It’s not pretty.

3 years ago, in April 2018, we went through one of our darkest days at Healthie. Our sales and marketing strategy was not working, and we had to lay off 20 people (85% of the company) in a single day. What followed was three months of little sleep and crunching the team, trying to rebuild our product from basically the ground up. It was an absolutely miserable time, but ended up turning our company around, and put us on a sustainable path to fast growth, profitability, and success.

We survived those dark days, but a lot of companies don’t. There were times where it looked like we wouldn’t, and I still have mixed feelings about how close myself and other employees came to burning out. Our decisions were the right ones, but I definitely made them with a great deal of naive optimism. 

If you’re struggling and thinking about the best path forward, it’s important to be mindful of survivorship bias in startup advice. Before taking on immense risk and emotional burden, get a clearer picture of your expected outcome, not just stories told through rose-tinted glasses. 

Entrepreneur is a Dirty Word

When I meet new people, and explain Healthie and what I do there, some of the most common responses are “It must be cool to be an entrepreneur” or “I want to be an entrepreneur.” I always smile, nod, and cringe internally.

I am not an entrepreneur. I build healthcare software.  No one is an entrepreneur. Despite literally millions of LinkedIn profiles with that title listed, “entrepreneur” is so generic and non-descriptive, that it ceases to be a real thing. You can be the founder of a specific business. You can even be a “serial founder” if you really want to call yourself that.  Those words describe actual actions taken.  “To Found” is a verb. “To Entrepreneur” is an absurdity.

When asked about your job/career, You would never say that you are a “business person” or a “human.” Calling yourself an “entrepreneur” is equally as generic and unhelpful. Worse, the word “entrepreneur” has been latched on to and weaponized by the Grant Cardone, Kevin Zhang, fake Lambo-owning, get-rich-quick-course-selling crowd. They talk about their glamorous lives as “entrepreneurs” and use the concept to take advantage of people.

They can only do this successfully because so many legitimate people currently call themselves “entrepreneurs.” When I searched the job title “Entrepreneur” on LinkedIn, many of the results are people who have built real businesses, who have created amazing products and teams. These are people that I respect, look up to, and want to emulate as we build Healthie. Frankly, the fact that they refer to themselves as “entrepreneurs” undersells their achievements.

We should be using words and titles that focus on specific actions taken and work completed. The more we do that, the easier it becomes to recognize and celebrate builders, while leaving scammers and charlatans behind. Let’s stop using the word entrepreneur.

Finding a Developer for your MVP

Navigating The Wild Wild West Of Online Freelancers

The most common question new founders ask me is how to hire developers to help build their MVP. These founders are normally passionate and knowledgeable, but are not technical. Their companies (like most nascent start ups) are normally tight on cash and time.

In an ideal world, these potential founders would all have a trusted technical friend who had the time, willingness, and ability to help build an MVP. In a slightly less ideal world, new founders could go to a meet up, hit it off with a person who is a great initial fit, and have them build it. Unfortunately, both these scenarios are rare. Good developers are in high demand, dev jobs are well paid and stable, and new startups are risky and painful (albeit hopefully rewarding). This leaves a tiny pool of developers that fit the above two categories.

New founders can’t afford to wait for the perfect developer to come along. To get the ball rolling quickly and affordably, these founders normally need to look to a much bigger pool of talent, namely online freelancers.

What follows are the tips and learnings I’ve gained from hiring contractors at Healthie, as well as helping a bunch of brand new companies find their first engineers. 

The below advice is for you, assuming you’re a non-technical (or not technical enough) founder building a web or mobile application with minimal tech risk. For example If the MVP of your app is based around machine learning or has intensive graphical needs like custom 3D rendering, there is likely better advice out there. This advice also assumes you’ve exhausted preferable options like finding a developer who is a great fit and you know directly (or comes highly recommended as a referral).

First, three pieces of general advice on hiring and working with your first engineer hire. 

  1. References, references, references - no matter who you hire, you should always do reference checks. No one is offended by it, and talking to references really helps you work with the hire. In the case of good references, they help you learn how the hire likes to work and interact with others. On the downside, the hire can’t provide references (or the references are lukewarm or negative) and you can avoid a huge amount of trouble. The worst candidates we hired at Healthie looked amazing on paper and from their initial interviews. We were very excited, and didn’t want to risk losing them, so we made offers without doing reference checks. In hindsight, references would have exposed the issues that we had to learn very painfully. 

  2. Own your accounts - You should be the primary account holder/admin on all services that your engineering hires are using to work with you. Most notably, this includes where your code is stored (e.g Github), where your server is hosted (e.g AWS), and the tool you use to manage the project (Trello or Clubhouse). Your first engineering hire may have (strong) opinions on the specific tools to use. Whatever tools you use, you need to be the one to create the account, and then invite the engineer as a user. This has a range of benefits, but most importantly, it protects you and your company if your relationship with the engineer ends badly. It makes you much less vulnerable to a bad first hire, which unfortunately, does happen. 

  3. If you’re non technical, avoid making technical assumptions. When it comes to development, some seemingly complex things are trivial, and some seemingly trivial things can be incredibly time-intensive to add. It is important for you to trust your developer, to respect their estimated degrees of difficulty, and to not micromanage. That said, it is worth doing the work upfront to feel very good about your engineering hire. It can also be helpful to have an outside technical resource that you can bounce basic assumptions off of (e.g does two weeks to build out a graphing dashboard make sense?)

With those pieces of advice in mind, it’s time to find the engineers to build your MVP. I normally see founders look at five main sources -- agencies, Toptal, Upwork, AngelList, and Fiver. 

 I’ll go through them one by one, but first, here’s a TL;DR summary. You should use 

  • an agency if money is of no concern, 

  • Toptal if you have time, (less) money, and few tech skills

  • Upwork if you are comfortable taking a risk, very cost-conscious, or able to thoroughly vet candidates. 

  • AngelList along with Upwork (it’s more focused on full-time employees and with a smaller talent pool)

  • Fiverr never

Agencies

Full service agencies promise the world, and sometimes, they deliver it. They normally pitch an approach where you give them the project details, and they can handle everything from design to tech specs to development to QA to release. Working with an agency is generally more opaque, hands-off, and expensive than trying to work with developer(s) directly.  That management layer needs to be paid for somehow, and they are, through the prices you pay the agency. They also generally charge high upfront project deposits that can be cost-prohibitive for new companies. 

When their work turns out good, it is high-quality, cohesive, and takes some serious weight off your shoulders. When it doesn’t, you can be left high and dry midway through a project. In my experience, you really get what you pay for with agencies. If the price sounds too good to be true, it is. 

Toptal

Compared to the free-for-all nature of the other platforms, Toptal is a guided tour to finding a freelancer. You start using their service by getting on a call with a “matcher.” You explain the project and what you are looking for, and the “matcher” introduces you to potential fits that have already been interviewed and vetted by Toptal. You then get to interview the different suggested candidates yourself, and if you want, hire one. This approach has serious pros and cons. 

On the plus side, you completely avoid a lot of the outright scams and fraud that can be prevalent amongst online freelancer sites. If you are non-technical, you can feel comfortable knowing that Toptal has done both personality and technical interviews, and the candidate is who they say they are. I’ve worked with a lot of great  freelancers from Toptal, and the average quality is the highest of all freelancer platforms I’ve used.

However, Toptal does have large drawbacks. First, you are basically at the mercy of your “matcher.” I have seen cases where the matcher doesn’t set up many introductions, or is slow to respond. Other times, the matcher doesn’t have a good understanding of your project (or the project is badly explained to them), and you’ll be introduced to candidates that are clearly not a fit. This can be frustrating, and outside of badgering your matcher, you don’t have meaningful direct control over the speed or quality of matches. Finally, this extra layer of vetting and staff need to be paid for, and that is reflected in the general higher rates that Toptal freelancers charge. From my experience, freelancers on Toptal are not meaningfully cheaper than devs in New York. 

Upwork

Decades old and formed by mergers of large freelance platforms, Upwork is the largest freelancer marketplace out there. It is a marketplace that covers every type of skill and experience level. Every job posting you make will be quickly inundated with applications. However, Upwork does close to no verification of it’s candidates. It’s basically a wild, wild west where you’ll find amazing developers for great prices, fraudulent but convincing charlatans, and everything in between. With Toptal, the difficulty is finding candidates. With Upwork, the hard part is sifting through them.

For some ”fun” anecdotes, we’ve had Upwork candidates switch out people between interview rounds, switch out people between the interview and the first day of work, and secretly take multiple full-time jobs and farm out work to lower paid devs. I have also been asked to give references on Upwork candidates who I have never heard of. It turned out there’s a whole series of fake portfolio websites out there that have added Healthie as work experience. Despite all those above stories, we still use Upwork! Why? If you can separate all the noise out, there are some fantastic freelancers on Upwork. If you are comfortable and capable of doing the vetting, and ok with some risk, Upwork has the best price to value freelancers that I have found. 

Here are some tips that have made a big difference to our usage of Upwork.

  • Have your application require specific questions to be answered. Try to make the questions specific enough to your project that candidates cannot just copy and paste answers. These questions lower the number of total applicants, show that the candidate put some initial effort in, and make it much easier to choose who you want to interview. In other situations, a longer application can deter good candidates, but that will never be an issue with Upwork. 

  • Be very selective about who you interview. Your big cut of candidates should come between the application and the first round interview. Otherwise, you will spend way too much time on video calls with clear non-fits. 

  • Require video to be turned on for all interview rounds. I like taking calls without video personally, but this is an important verification step

  • In the interview, make sure the person’s description of their experience and work timeline matches what was in the Upwork application. 

  • Take a screenshot of the person you interview, and at every round of the interview. Make sure that the picture matches anything online (e.g LinkedIn), and also that it is the same person at different interview rounds. This sounds insane, but candidate swapping (e.g a different person interviews than applied, or different people do different interview rounds) happens regularly. 

  • Ask for references, but don’t stop there. You want to do due diligence on the reference, both to ensure that a) the reference company is real b) the person you speak to actually works at the company. Candidates will provide fake work experience, and even fake references. If possible, you want to speak to a technical person at the reference company. 

  • If you move forward with the hire, hold them to a high initial standard of a) joining a daily check-in on time, and b) joining with video on and giving their update. This helps cut down on the farming-out-work problem, and you can be a lot more lenient once you get comfortable with the dev. 

AngelList

AngelList is similar to Upwork. Anyone can join AngelList, and they do not vet candidates for you. Compared to Upwork, AngelList is more full-time focused, and is more U.S-centric. In general, it takes minimal extra time to post the same job on Upwork and AngelList, and there is no harm in doing so. 

Fiverr

Fiverr can work well for odd-jobs (e.g freshening up the design of a pitch deck), but your MVP is not an odd-job. Personally, I’ve had bad results with even trivial non-programming projects I’ve tried to use it for. It’s not a fit for this. 

Predictions for 2021

2020 is (thankfully) over. With the new year starting, now is the time to review my 2020 predictions, and add some new ones. This year, I’m going to avoid specific stock prices (since I can’t tie back underlying rationales to exact prices), and political races (since I’m burnt out on politics at this point).

Stopping reckless securities trading will become a cause célèbre, leading to self-regulation.

With a booming stock market, and an incredibly low barrier to entry, general excitement around securities trading (stocks, options, etc) has reached a fever pitch. Robinhood has been leading the charge here, offering a commission-free, user-friendly, and gamified experience.  Robinhood makes it easy to trade stocks, but also to trade options, and trade on margin. Many of Robinhood’s features are good (e.g fractional shares), but they also put Robinhood (and the industry as a whole) on the edge of PR disasters caused by users who are dangerously in over their head. These PR crises are magnified by issues with the apps themselves, leading to ultimate tragedies like this one. Over the next year, enough of these issues will accumulate, leading to greater self-regulation from the apps.

Prediction: In the face of consistent, widespread public outrage, Robinhood (and similar apps) will self-regulate, increasing their requirements to trade options and trade on margin.

Miami, as a tech capital, will be a short-lived Twitter fad. There won’t be a new tech capital.

If Twitter trends are to be believed, San Francisco is falling to pieces, and Miami is rising to take its place. VCs and founders are moving there, and throwing their weight behind it. The mayor of Miami has become a cheerleader (and meme), pledging support and cafecitos to all.  COVID (and issues with SF) will surely lead to Miami having a larger tech scene than before. However, the increase (especially in raw numbers of jobs) will be a rounding error, and not more than other mid-sized American cities like Austin, Seattle, Philadelphia, etc. In this new remote-first world, it is easy for individual people (like the VCs and founders who hype Miami on Twitter) to move. It is not easy, nor beneficial, for employers to ask all employees to. Large, established tech behemoths will remain in their current cities, Stanford is not going anywhere, and the Miami weather will sound a lot worse come summer. This all leads to a whimper, not a bang.

Prediction: By August, we’ll be talking about Miami, the same way we talk about Seattle or Philly. Miami will not see a meaningful increase in tech jobs, nor will there be a large increase in VC investments into Miami-founded companies.

It will become common for startups to incubate in-person, and grow remotely.

Remote work sounds great, and, in most cases, is great. We are big fans of it at Healthie. It has allowed us to scale our team much more easily, and gives our employees (and us as founders) a great deal of flexibility. However, I still think it is very hard to get a company off the ground in a fully remote setting. In those nascent stages, iterations need to happen faster, situations change rapidly, and targets can be less straight-forward and measurable. Effective remote working requires a great deal of written communication, and process overhead. This additional structure is great for established companies with clear next steps, but is just extra baggage for companies in the pre-product and pre-revenue stage.

Prediction: A new model will emerge with founders and early employees working in-person together for the first three to six months of the business. After that period, companies will transition to being remote first.

Traditional IPOs will get less and less popular, as more companies opt for direct listings and SPACs. 

This has been a very interesting year for taking companies public. We’ve seen three major trends this year. First, traditional IPOs have continued to have large first-day “pops”, leading to companies leaving a lot of money on the table when they go public. This is not new. However, in 2020, new alternatives have finally begun to gain real traction, which brings us to the two other trends. SPACs hit the mainstream this year. These “reverse-mergers” allow companies to go public with the least amount of moving parts. Instead of a drawn-out traditional road show, private companies can work with a single SPAC. This simplified process has benefits for both the SPAC and the private company, and we will continue to see it used in cases where the company wants to go public early (Hims, OpenDoor) and/or is in a “sexy” industry (Nikola, DraftKings, or Virgin Galactic). On the other end, recent SEC approvals , have expanded the utility of direct listings, making it a better option for more companies.

Prediction: Q3 and Q4 will see well-known tech companies opt for alternatives to a traditional IPO. If Stripe goes public this year, it will be via a direct listing.

The U.S will get their vaccination act together

The U.S vaccination effort is off to rough start. Despite months to prepare, the country has been caught flat-footed, and progress is slow. It is jealousy inducing to watch Israel, a country with a population the size of New Jersey, be on track to reach herd immunity in just a couple of months. The U.S predicament is pretty close to cat herding, with fifty different states, a federal government in transition, and a president with his mind on other matters. However, despite our slow start, the situation will stabilize, allowing America’s resources to be put to good use.

Prediction: After a couple of months of non-stop media coverage of vaccination delays, the system will get into gear, and we will reach herd immunity by the fall.

Loading more posts…