Amazon Prime Instant Video: Genius move to grow revenues or chasing skewed metrics in vain?

The question pondered in this post is, whether the metrics used by Amazon to calculate the revenue-boosting effect of Amazon Prime do stand up to this task.

Amazon’s stock price (AMZN) has risen about 20% during the days after their earnings report in January 2015. Amazon showed some profit, beating estimates. While the profits were important to show that Amazon can increase its margins, the main reason to buy Amazon is the anticipated revenue growth in coming years. The high valuation considering P/E reflects this.

The main driver for this growth is Amazon Prime. And they definitely go all-in on this idea. Amazon is spending big time on goodies that come on top of the 2-days free shipping. In 2014 they invested 1.4 Billion $ in Prime Instant Video.

Fuel for growth: To get more people to join Amazon Prime — and buy more goods per year — Amazon has morphed into a streaming-media and mobile-device company.
In 2011 the e-tailer began offering Prime members access to instant streaming movies and television shows; the retailer now produces its own TV programs as well. To ­sweeten Prime, Amazon recently added a streaming-music service and free online photo backups.


“Prime is a one-of-a-kind, all-you-can-eat, physical-digital hybrid (…)In 2014 alone we paid billions of dollars for Prime shipping and invested $1.3 billion in Prime Instant Video. We’ll continue to work hard for our Prime members.”
Jeff Bezos on the Q4/2014 Earnings Call

So the strategy consist of two major parts:

  1. Amazon Prime causes members to spend significantly more money shopping on Amazon and therefore growing the Amazon Prime Membership base is the best way to increase revenues.
  2. The investment in Amazon Instant Video and the investments in original programming are used to boost the Amazon Prime, by increasing the sign-up and renewal-rate.

But does the numbers available support the thesis? Let’s have a look at (1) first. On a first glance it just looks like it:

Bundling so many services into Amazon Prime seems to be working: According to RBC data, 40% of Prime members have spent more than $200 on the website in the last 90 days, compared to only 13% of non-Prime members.


Amazon Prime members make twice as many purchases as nonmembers, and they spend 40% more per transaction, reports ComScore


Analysts say Prime members increase their purchases on the site by about 150 percent after they join

But you have to take a second look on these, because the numbers show a correlation and cannot prove causation. Several alternative causal chains are possible. Let’s have a look at 3 examples:

1a) Amazon Prime is not the reason why people buy more, it’s just the other way round: People, who buy more an Amazon are more likely to become Amazon Prime members.

1b) Once a person has an Amazon Prime Account, he places orders for close friends, family and business too. So the overall revenue does not increase, orders are just reallocated from one amazon account to another.

1c) A person is planning to purchase more in the future, maybe because he got his first good-paying job or his spouse moves in, is more likely to become a Amazon Prime member. This causal chain is very similar to (a), with the difference, that in (c) the same person purchases more in time, not one person purchasing more than another person like in (a).

So the role of “Amazon Prime” in the great revenue growth of Amazon might not have been as important as a quick look at the numbers suggests. Even if it is difficult to analyse the impact of Amazon Prime’s “free 2-day shipping” on the revenue growth, it certainly is a success. All the other available data supports the thesis, that “free shipping” is very important to online shoppers. In fact, by far the most important factor:

source:Comscore White paper “State of the U.S. Online Retail Economy in Q1 2014; "; slide 17

But what about all those more indirect investments in content for Amazon Instant Video and other goodies?

“What we see is customers that come in who come in through our Prime pipeline for video for free trial, those customers are converting at higher rates than other channels. We see that customers that are video streamers, even though we have high renewal rates, they are renewing at even higher rates than others.”
Tom Szkutak — Amazon Chief Financial Officer, Senior Vice President Amazon Q4/2104 Earnings Call

There might be other “non-causal” explanations for this behaviour, like:

2a) People, who are more into streaming are the tech-literate, online-living kind. These people flock to online shopping and are more likely to stick to their Amazon Prime account.

2b) If you are planning to cancel Prime or not to use it after a free test period, you are less likely to start watching a series on Instant Video and less likely to setup your TV or install a streaming app on your mobile device.

So does this prove, that Amazon is chasing skewed metrics and throwing billions of dollars at Prime Instant Video in vain? No it does not. It just shows, that the numbers Amazon presents to the shareholders to show that the investment in streaming content is worth it, are no prove. It might be worth the money, or it might not. It really depends on how you look at the data, just like Ben Thompson from stratechery put it:

According to a just-released study from Consumer Intelligence Research Partners, Prime members spend $1,500 per year on Amazon, nearly triple the $625 spent by non-members. The study further estimated there were 40 million Prime members, which means that 53% growth equates to nearly 13.8 million new members. Those new members, on an annual basis, ought to be worth $12 billion in addition sales to Amazon (presuming they increase their spend by the average $875). Suddenly, with a bit of handwaving and some generous assumptions about just how many members fit in Szkutak’s characterization, the $1.3 billion that Amazon spent in 2014 on Prime Instant Video seems a little more reasonable.
source: ($)

So is Amazon aware of this and is it using it’s enormous data trove to critical monitor the success of Amazon Prime? Of course Amazon has both the capability and the data, but do they have the right internal incentives and culture to scrutinize Amazon Prime and especially the “goodies on top” part?

Jeff Bezos is a really genius, whom I admire a lot. Alas, nobody can be right all the times and sometimes the downside of being a genius is, that only few people dare to give you critical feedback. Being the CEO and the largest shareholder of the company besides being a genius adds to this dilemma.

Amazon Prime has always been Jeff’s Baby, he committed to the program early on and pushed it ever since(“The program is a “big idea,” Bezos told the group that day in the boathouse”). And it has been a great success for quite some time now. But every good idea can be taken too far, if you get carried away. Lately there has been a new Baby of Jeff’s, the “Fire Phone”, we all know, how this one turned out.

“We poured surreal amounts of money into it (the Fire Phone feature “Dynamic Perspective), yet we all thought it had no value for the customer, which was the biggest irony. Whenever anyone asked why we were doing this, the answer was, ‘Because Jeff wants it.’ (…)
Even so, this top engineer adds, “Jeff is immensely successful; he’s taken huge gambles in the past, so he has shitloads of respect from the team members.” (…) When Bezos encouraged a free-shipping initiative, executives pushed back, nervous about its impact on earnings. But Bezos prevailed and later bundled free two-day shipping into Prime’s immensely popular subscription service, which now boasts tens of millions of customers. (…) So when it came to the Fire Phone, says one former product lead, “Yes, there was heated debate about whether it was heading in the right direction. But at a certain point, you just think, ‘Well, this guy has been right so many times before.’”

I really like Amazon Prime and it’s core: The free 2 days-shipping. But I doubt, that all the extra resources invested into Amazon Prime Video, the Lending Library and Gadgets to support the “Amazon universe” are worth the money and the companies attention. Maybe Amazon would be better off concentrating more on logistics and the core shopping experience, especially the design and UX of the various Apps.

Disclaimer: At the point of writing I am short on Amazon.

Show your support

Clapping shows how much you appreciated Frank Drews’s story.