💡JCNews #60
Amazon's data centers
Dear friends,
Welcome to the 16 people who have joined us since last Tuesday! If you're reading this and haven't subscribed yet, please join the 3118 people who have already joined the move. It's right here!
In JCNews, I share the articles, documentaries and books that I enjoyed the most in the last week, including a must-read.
Let’s talk about this together on LinkedIn or on Twitter. Enjoy!
💡The weekly must-read
Each week, I share with you a must-read which you chose.
👉How Amazon replaced its data centers and how it led to AWS (Twitter Dan Rose)
I was at Amzn in 2000 when the internet bubble popped. Capital markets dried up & we were burning $1B/yr. Our biggest expense was datacenter -> expensive Sun servers. We spent a year ripping out Sun & replacing with HP/Linux, which formed the foundation for AWS. The backstory:
My first week at Amzn in '99 I saw McNealy in the elevator on his way to Bezos' office. Sun Microsystems was one of the most valuable companies in the world at that time (peak market cap >$300B). In those days, buying Sun was like buying IBM: "nobody ever got fired for it"
Our motto was "get big fast." Site stability was critical - every second of downtime was lost sales - so we spent big $$ to keep the site up. Sun servers were the most reliable so all internet co's used them back then, even though Sun's proprietary stack was expensive & sticky.
In 2000, brand new Sun servers started appearing on eBay for 10 cents on the dollar as VC-backed start-ups went out of business (this was pre-AWS when you had to roll your own datacenter). Amzn could have negotiated a better deal with Sun, but Jeff chose a more radical approach.
Amazon's CTO was Rick Dalzell - ex-Walmart, hard-charging operator. He pivoted the entire eng org to replace Sun with HP/Linux. Linux kernel was released in '94, same year Jeff started Amzn. 6 years later we were betting the company on it, a novel and risky approach at the time.
Product development ground to a halt during the transition, we froze all new features for over a year. We had a huge backlog but nothing could ship until we completed the shift to Linux. I remember an all-hands where one of our eng VPs flashed an image of a snake swallowing a rat
We should always ask ourselves the question of the radical changes a company can do to avoid its fall. What fundamental hypotheses can we question?
This coincided with - and further contributed to - deceleration in revenue growth as we also had to raise prices to slow burn. It was a vicious cycle, and we were running out of time as we ran out of money. Amzn came within a few quarters of going bankrupt around this time.
Even today’s biggest companies had moments when they got very close to going out of business.
But once we started the transition to Linux, there was no going back. All hands on deck refactoring our code base, replacing servers, preparing for the cutover. If it worked, infra costs would go down by 80%+. If it failed, the website would fall over and the company would die.
We finally completed the transition, just in time and without a hitch. It was a huge accomplishment for the entire engineering team. The site chugged on with no disruption. Capex was massively reduced overnight. And we suddenly had an infinitely scalable infrastructure.
Then something even more interesting happened. As a retailer we had always faced huge seasonality, with traffic and revenue surging every Nov/Dec. Jeff started to think - we have all this excess server capacity for 46 weeks/year, why not rent it out to other companies?
Around this same time, Jeff was also interested in decoupling internal dependencies so teams could build without being gated by other teams. The architectural changes required to enable this loosely coupled model became the API primitives for AWS.
We also used the work “primitives” at Alan for elementary layers of our health services.
These were foundational insights for AWS. I remember Jeff presenting at an all-hands, he framed the idea in the context of the electric grid. In 1900, a business had to build its own generator to open a shop. Why should a business in 2000 have to build its own datacenter?
Cloud infrastructure would have emerged eventually even w/out AWS (like electric vehicles w/out Tesla), but how much later and at what opportunity cost ? After the cost of starting a company was reduced dramatically by AWS, innovation exploded and the modern VC ecosystem was born
PS: Amzn recently spent years ripping out Oracle, something few have attempted. It takes muscle to do hard things, and muscle gets built by doing hard things. The best companies look at every challenge as an opportunity and engrave that mindset into their culture.
I am completely aligned with the two last sentences: if we want to create differentiation, we should do the hard work that nobody dares doing.
🏯Building a company
In addition to selected articles, I share one of Alan's leadership principles each week - the same one I share internally and with our investors every Wednesday.
👉 Alaners ask questions (Healthy Business)
It is not a failure not to know something, even if the question is dumb. You will learn from it. If people don’t have the time to help, they will let you know.
Learning new skills, a new industry, a new corporate culture is hard, and it is normal. We should encourage people who want to grow and are seeking more knowledge and information. Let’s welcome all those questions, publicly!
👉How to have a successful employee onboarding from Shopify (and how to manage partnerships) (Alex Danco)
Onboarding:
So far it’s going basically on schedule: as I was told, “Your first couple months you’re going to have zero idea what’s going on. Then around month three you’ll come up for air and think, ok, I got this; and then you’ll try to start doing stuff. Then you’ll really struggle, because you won’t be in that happy new float-around-and-learn-people’s-names mode, you’ll be in oh-shit-can-I-really-do-this mode. It’s actually a little scary. But then around month five or six, you start to actually figure some things out for real. And then it starts to feel fun.”
“In your first 6 months here, here is your number one job. Familiarize yourself with the dozen senior people at Shopify who have the final call on really important decisions, from Tobi and Harley on down. You need to familiarize yourself with their operating philosophy around business and around how Shopify works. Go consume every written memo and every podcast episode (we have a great internal podcast called Context) they’ve ever done, get inside their heads, learn their perspectives and their preferences, and learn what gets them to say Yes to things.
Partnerships:
“if you’re a startup, try not to waste your time talking to big companies unless you’re doing it deliberately. They have an absolutely endless capacity to consume your time with meetings.”
The art of getting these partnerships right is really solution architecture: making sure from the very beginning that someone who understands every piece involved can get them lined up right, and interacting the right way, from the very outset of the project.
Unless you really fix things quickly, the interface between your two companies (and the mechanics on both sides) just compound with problems on both sides, and never get really fixed – just patched over and over.
👉What is the cost of an IPO? (Crunchbase)
The dollar value of an IPO is what will determine the bank fees: around 6.5 percent to 7 percent for a $100 million IPO. That percentage will get lower as the deal size increases, according to PwC, with deal sizes at $1 billion or larger having an average 3.5 percent underwriting fee.
Nowadays, however, the initial IPO conversations start about 18 months before a company’s public debut. [...] Board composition requirements–such as California’s gender and diversity requirements–have increased the lead time to an IPO.
Law firm fees are somewhere around $1.7 million to $2 million for this type of transaction, and auditor fees are also around $2 million, Ben-Tzur added. Banks on a typical tech IPO charge a standard 7 percent commission on the proceeds raised.
🗞In the news
📱Technologies
👉 Apple’s Arizona Challenge (The Information)
A committee in Arizona’s state House of Representatives voted by a slim margin to advance a bill preventing Apple from forcing app developers to use its in-app payment system, which takes a 30% cut of transactions. The bill now has to go to the vote of the full House.
The bill would also affect Google’s app store for Android devices. The vote was a win for Epic Games, which has been complaining to antitrust regulators about the app store owners’ practices. Epic had lobbied for a similar bill that was defeated last week in North Dakota, although that bill was broader, and also would have required Apple to allow third-party app stores on its phones.
🏥 Healthcare
👉 Diving into Oscar Health’s S-1 in the context of their IPO (In Silico) (Kevin O’Leary)
Oscar Health has raised $1.6 billion to date.
As of Dec 31, 2020, Oscar had a total of ~400k members.
Membership does continue to grow quite nicely in the individual and small group products - growing from 230k members in 2019 to 400k in 2020 (and they’re currently at 529k total members as of 1/31/2021)
In 2020, Oscar collected $1.7 billion in premiums
Interestingly, New York (where Oscar got its start) saw a large decline in members this year from 51k to 39k.
Oscar has accumulated a $1.4 billion deficit over the years, and it lost $261 million in 2019 and $407 million in 2020.
Some interesting financial metrics: the volume of loss so far is an important data-point, and the fact that it keeps increasing as well.
With Oscar, there are 2 main ways in which technology can be used to rethink health insurance:
1. By dramatically improving the member experience
2. By dramatically reducing costs
This combination is similar to the one of what where the pillars of Amazon.
Lastly, they get matched with a personalised Care Team, who can help answer questions about providers, prescription, and benefits throughout the member’s journey.
More than 88% of their members have interacted with their Care Team, almost half are monthly active users, and members give Oscar an NPS score of 30 (compared to 3 for the average payer).
I really like how they have organised their “care team”.
Engagement rates are comparable between members aged 55+ and 18+, and NPS is actually highest in members aged 65-74.
Super interesting data-point on how different age populations can have high engagement rates.
Oscar has less to show on reducing costs? Care Team triaging resulted in a median member savings of 7% for those who followed recommendations, and proactive Care Team intervention in urgent situations reduced the number of ER visits by 13%.
For Oscar to show that they are not just a tech-savvy insurer with better design, they need to find a way to leverage technology to fundamentally shift the cost equation of their business.
7% starts to be an interesting impact, and still is only the beginning.
Oscar has established with other organizations, including both payers (Cigna) and providers (Cleveland Clinic, Montefiore Health System, Holy Cross Health). The language used to describe these partnerships seems to imply an important strategic purpose, but it is unclear at the moment what these agreements actually entail.
It appears that Health First is the first fee-based service arrangement Oscar has in place.
Oscar may see themselves providing more of a tech infrastructure service for payers and providers over time
It seems they have decided to sell their platform as a service to other providers.
Oscar: “[Our tech] platform enables us to run nearly 80 automated population health programs, or campaigns, as of December 31, 2020, that allow us to collect data and trigger real-time interventions at scale.”
Oscar’s virtual care services are provided through three independent, physician-owned organizations known collectively as Oscar Medical Group
The combination of automated programs with access to doctors is a powerful combo.
💚 Alan
👉 [French] Podcast Happy Work by Gaël Chatelain-Berry (Happy Work podcast). Pleased to discuss last week with Gaël occupational health and innovative HR initiatives we try to lead at Alan.
It’s already over! Please share JCNews with your friends👇
Let’s talk about this together on LinkedIn or on Twitter. #JCNews are also available on Alan's Blog. Have a good week!