Origins of a Billion Dollar Business: Amazon Prime Video
Early product decisions, implications, and how a small team of 9 kicked off Prime Video (Chapter 1)
Hello, it’s Ethan & Jason. Welcome to Level Up: Your source for executive insights, high performance habits, and specific career growth actions.
New course announcement ‘Billion Dollar Product Innovation’. I built 5 different billion dollar products and experienced 10 more. If you want to learn the horizontal cross-cutting product lessons and the product decision principles of what to do and not do, sign up to get on the waitlist (this will be a fun course).
FYI, we have 3 upcoming events—join us live!
(March 12) Take BIG Bold Career Risks. Ethan & Molly Graham will discuss how people can manage their fear to make the leap. RSVP here.
(March 18) Guide to High Profile Events (Davos, TED). Ethan & Alisa Cohn will discuss how to get into elite events, what to do, and how to follow-up. RSVP here.
(March 28) Leadership Lessons from Startup to IPO. Jason & Brian Distelburger will discuss leading a 2,000+ person company, what Brian looks for when hiring executives, and his next big thing. RSVP here.
Chapter 1: Conceptualizing Prime Video
I joined Amazon on April 11, 2005, and I had no idea what I would be doing.
Amazon had recruited me for a secret project and told me nothing about it.
Even so, the decision to join them was easy since I was at a failing startup, and they offered me more money. When I took the job, I was worried that I was joining the IT department of a bookstore because back in 2005, Amazon was known as little else than an online bookstore. AWS, Kindle, and other innovations would not become public until later.
On my first day at “the bookstore,” I learned that I was joining a group called “Digital Media.” A few days later, my Vice President, Neil Roseman, told me there were two possible jobs for me: I could build an online video product for Amazon, or I could run the infrastructure team that would build a unified ingestion pipeline for ebooks, music, and video.
I knew that I preferred customer-facing projects to infrastructure, so I picked the video project. This caused a bit of a problem because another manager who started the next week felt that he had been promised the video effort and instead got the infrastructure team.
This was perhaps my first lucky break.
In the year before my arrival, in 2004, Jeff Bezos had decided that the future of media was digital.
One of Jeff’s many important divinations.
However, at that time, almost all of Amazon’s sales were coming from a division called BMVD, which stood for Books, Music, VHS, DVD. Jeff’s essential thesis was that if Amazon wasn’t ready when books, music, and video went online, they were in trouble.
In order to focus on winning in this transition, Jeff recruited one of his top executives, Steve Kessel. Steve and his team started brainstorming products with Jeff and hiring a staff.
I was part of that wave of hires.
By the time I arrived in 2005, a few things had happened:
Jeff had realized that Amazon operated in the middle of the value cycle for physical goods, focusing on distribution and logistics. However, he believed that the value in a digital world would accumulate at the ends of the cycle, either with the content creators (authors, musicians, and filmmakers) or near the consumers (control of reading, listening, and viewing devices). In other words, owning the logistics was only valuable in a world of physical products. In a digital world, we either needed to own the content or the means of consumption.
A decision had been made to lead the digital transformation with ebooks and Kindle, and a veteran manager had been hired back to Amazon to lead the music effort. Video, my product, was the third priority.
While a product leader had been hired, the company had not yet settled on a product plan.
So, when I started at Amazon, I had no team.
I was given a total headcount of six and I learned that three of them were to be filled by new college hires who would show up in May. I would also get one intern. In a stroke of good luck, this intern, Brian, had been at Amazon the year before and already knew the internal systems. As a result, Brian would end up teaching me and other members of my team how things worked.
When I finished building out the team, the final composition was three new college graduates, one intern, one internal team veteran with a year of experience, one hire with two years of experience, and one internal transfer with four years of experience.
Thus, my team of seven had a total of seven years of experience between them.
While we built the team, the managers and leadership debated products. We considered things that seem crazy in hindsight, like competing with the Netflix DVD service by asking customers what movies and TV shows they wanted and then burning them on custom DVDs and mailing them.
However, after a long process and plenty of regrettable ideas, we decided to build a basic digital movie and TV service that offered content for both rental and purchase.
This was somewhat of a “me too” product, though we tried hard for small differentiations.
With this broader decision made, we needed to make a series of decisions to further develop the product.
These included:
Product Decision 1: Online movie and TV rental and sale.
Product Implications: Others could sign the same deals, making it hard for us to differentiate this product.
However, fast broadband penetration was low in 2005, so customers would have to download the content before they could watch it. Streaming as we know it today did not exist widely. This led to product decision 2.
Product Decision 2: Download content because few customers could stream it.
Product Implications: There would be a long delay (hours) between wanting to watch something and being able to do so. Note that at this time, competing options were to drive to a DVD rental store like Blockbuster or to choose something on Netflix and get it in the mail in a couple of days. So, a few hours of passive preparation to watch something was a competitive option. Today it would not be.
Another challenging implication of allowing for downloads was that the content owners (movie and TV studios) were very worried about piracy. They saw the black market that Napster created for MP3 music and wanted to avoid anything like that for video. As a result, we had to have strong Digital Rights Management (DRM) in place.
The studios had a short list of approved DRM systems, and one of them was baked into Microsoft’s Windows Media Player. Basically, Microsoft had spent the money and time to convince the studios that their solution was safe. This guided our hand to product decision 3.
Product Decision 3: Our main format would be Windows Media DRM, downloading content to PCs.
Product Implications: Our content would go to your PC. Macs were not supported. In 2005, smart TVs were just coming out and had tiny penetration, so there were few better options. This also meant that we needed to build a PC-based application as a company that only knew how to create web services.
With these decisions made, we started building. The team was six engineers, one manager (me), one product owner, one business developer, and one product manager.
Today, Prime Video has thousands of employees, but we began with a team of 9.
Despite our small size, our team gelled quickly and was moving fast.
But, all of the Digital Media teams were incredibly lean and Kindle, music, and video were all sharing legacy Amazon resources. At the time, the most critical of those resources for us were:
Windows machines to build our PC application.
The product catalog system.
One relatively young UX designer named Chip.
The challenge with finding Windows machines was that, at the time, Amazon was a pure Linux shop. So, one of our first technical hurdles was where to put Windows Media Servers. The data center teams had no expertise or infrastructure and did not want to patch or maintain a new OS.
The challenge with the catalog system was this:
Behind each item you can buy from Amazon is a database called the catalog.
Even in 2005, this catalog contained millions of items with many different attributes. It had grown organically from books to CDs to VHS tapes to many different kinds of products like toys and clothing.
But, Amazon was already 9 years old and the catalog was a fragile legacy system. We spoke to the owner about getting digital videos included in the catalog, but he explained that he had a backlog of requests to extend the catalog to new types of items and that he couldn’t get to us for at least a year.
The final challenge was the UX. Chip, the single UX designer for all of Kindle, music, and video had been briefed on the product priorities– Kindle, then music, then video.
So, when I went to my VP and asked for Chip’s time, I was told I would have to wait.
This was the first time the prioritization of our products was explained to me, with video being third behind Kindle and music.
The result of these resource constraints was that my team had to build and launch our product without dedicated server space, without the ability to be listed properly in the catalog, and without a dedicated, professionally designed user interface.
We dealt with these hurdles by creating and employing a number of hacks, both technical and organizational, to get the product off the ground without waiting the year that it was estimated we would have to wait.
I am going to go through each of these hacks in detail in next week’s newsletter, so make sure you are subscribed to read how my team ‘MacGyvered’ the original version of Prime Video and got it to launch, only to realize that it had a few major problems… (update: read chapter 2 here).
Readers- I want to invite people who worked on Prime Video to share their view of what happened in the comments. If you worked at Prime Video, tell your version of the story in the comments.
Connect with Ethan & Jason
Follow Ethan on LinkedIn.
Get Ethan’s career advice on YouTube.
Connect with Jason (Ethan’s COO) on LinkedIn.
Learn more about Ethan’s live online courses and on-demand courses.
Contact us for corporate training, speaking, podcast appearances, and more.
In-person course enrollment open (10 seat limit): The next cohort of my course ‘Breaking Through to Executive’ will have 10 seats for an in-person experience that includes small group coaching, structured networking, and more. Read here if you are interested.
Many subscribers expense this newsletter to their Learning & Development budget, use this email template to send to your manager.
+1 me too, i like to know how the two role differs at that point of time.
Enjoy reading it. Can't wait for next one. Most funny part was Infra team saying that they don't want to manage or patch new OS lol This is still the challenge and problem in different organizations.