Amazon Web Services is the ‘Juicero’ of Web Hosting, so Why is it Winning?
Growth Study: How did Amazon win despite having an overcomplicated and higher priced solution?
Juicero has become the poster child for silicon valley excess. They raised and spent $120M developing a $700 juicer which is essentially useless. If we take away all the hype, they took a simple industry (juice) and created a technology solution that is overly complex and more expensive. In the business world, this is a disaster.
Amazon Web Services (AWS) is Amazon’s cloud hosting business that boasts millions of customers and a $13B run rate for 2017. If those numbers aren’t enough, consider this: the AWS cloud is 10x bigger than it’s next 14 competitors combined. In the business world, this is what we call #WINNING
Here’s the issue… AWS’ success makes NO sense.
If someone had pitched this business to me 10 years ago, I would have predicted almost zero chance of success. The hosting industry is a global, price sensitive, commoditized market and AWS has two big problems:
- It can be up to 4x more expensive than existing options
- It’s a complex offering that requires proprietary education
More expensive & overly complex — this sounds a lot like Juicero to me. So why have they been able to succeed? Let’s explore each problem and why they won.
Problem #1 — AWS Costs 4X More Than Existing Solutions
Because server hardware, network reliability, and services levels are virtually the same across all major providers, the hosting industry is incredibly price driven. AWS storage, compute, and bandwidth bills often come in at 50% more than other cloud providers and up to 4x premium over what you’d pay for bare metal resources. By creating an incredibly complicated pricing model, it’s nearly impossible for companies to estimate how much they’ll be paying.
So how did Amazon overcome this? The Drug dealer pricing model.
AWS Gives Customers the First Taste Free and Then Locks Them In
We already know that pricing strategy and customer acquisition are some of Amazon’s primary strengths. With AWS they played to those strengths beautifully. Believe it or not — businesses can use Amazon Cloud absolutely free. They provide 750 hours per month of compute time and database usage and 5GB of standard storage for free. World-class infrastructure at zero cost? Where do we sign up? (BTW, you can sign up here, but keep reading…)
Once your business grows out of the free level, the bills start to arrive. Sometimes the usage isn’t even your fault, but when it is, and you’re just growing as normal businesses do, you’ve already spent months (or years) building your app on their proprietary platform. Packing up and moving somewhere else is not an easy task, no matter how much you’re paying.
Problem #2 — AWS is a Complex Offering that Requires Proprietary Education
The web has been singing the praises of open source and moving away from proprietary tech for years. Tech-types are specifically adamant about moving to open source. Despite this, Amazon chose to build a proprietary system that requires developers to learn a unique new set of skills.
How did Amazon overcome this? A Free Global Sales Team.
AWS Created Value for their Primary Advocates & Influencers
Having a killer sales team goes a long way in building a successful business. Rather than just build an in house sales team, Amazon developed a system of proprietary certifications — AWS Certified program — which essentially turned IT professionals into a huge, free, global sales force.
You see, In the IT world, certifications mean higher average salaries so this program gave every IT professional a strong motivation to ‘sell’ AWS to their employers. Convincing their boss to choose AWS means company-sponsored education and certifications that will get them higher pay and a distinct competitive advantage.
The genius here is that business owners are often at the complete discretion of their IT employees and consultants when it comes to what solutions they use. Of course, this leads to a scenario where IT executives and staff have a lot to gain by pushing AWS, regardless of whether or not the solution is best for the business.
In a 2015 Forbes article, Tom Gillis openly acknowledges the fact that AWS is more expensive but still advocates for it: “Where does the truth lie? In my opinion, it doesn’t matter. The driver for using the public cloud is not a 10 percent or even a 90 percent cost improvement. It’s about something more important.” He goes on to talk about CISCO’s IT struggles with scaling — but that isn’t a problem that most businesses will ever need to deal with. So while this rationalization only applies to an elite set of enterprises, many SMBs read articles like this and end up choosing a service that isn’t really right for them.
It seems like Amazon made all the right moves, despite having what appeared to be a less than ideal product. Is Jeff Bezos & team some sort of real life King Midas? Maybe, but there might be a better answer.
Amazon Started Serving Customers Early and Built a Product Around Them
Even before EC2 (Amazon’s Compute Product) exited beta in October 2008, Amazon had already secured deals with Dropbox and Netflix, two of it’s largest customers. In fact, AWS tech had been in private beta with several customers since 2005.
They identified a core market — Enterprises and heavily funded startups that need rapidly scaling resources and aren’t expressly concerned about the increasing costs — and they worked with them from the start to create a product that solved their needs.
In fact, they took 3 years to get their product right before making it publicly available. As Ben Einstein correctly pointed out, Juicero should have “…focus[ed] on getting a product to market to test the core assumptions on actual target customers, and then iterate. Instead, Juicero spent $120M over two years to build a complex supply chain and perfectly engineered product that is too expensive for their target demographic.”
By getting to know their customer and creating a product that worked for them, the final product was able to reach far beyond their initial market and dominate an industry to which they are a relative newcomer.
So Now What?
I don’t know where Juicero is going next, but Amazon proved that a complex, expensive solution can still win big. However, even Amazon isn’t above organizations clueing in to the huge cost differences and moving back to bare metal IaaS solutions. They announced reduced pricing just a few months ago (Nov 2016) in an attempt to bridge the gap and I’m sure more changes and concessions will come in the next few years.
Updated to answer questions and comments brought up in responses:
First of all, to those of you that read and responded — I truly appreciate you. I love being able to have an intelligent discourse on a topic that is important. IT Infrastructure is often a fixed cost of doing business and that means choosing wisely can make an impact on your bottom line.
“It’s not true that AWS is more expensive than all of their competitors. It’s also not clear who you’re comparing them to. AWS, Azure and Google Cloud all have different offerings and one is not always cheaper than the others.” ~Laurent Perrier
I beg to differ on this. There are a litany of detailed posts that discuss the elevated costs associated with AWS — but it is important to note that this article is not comparing AWS to Microsoft Azure or Google Cloud.
When I reference AWS competitors or alternatives, I’m primarily referring to bare metal, infrastructure as a service products. This includes options like dedicated servers, private clouds, and possibly cloud-based VPS options if your needs are small.
“AWS is not more expensive but it does add tremendous value unlike Juicero.” ~Bob Warfield
With regard to costs, see my paragraph on AWS Pricing above. In terms of value, what are you referring to? Amazon may have a lock on simplifying rapid, large-scale scalability, but in terms of performance, ease of use, lifetime cost of ownership, etc., I just don’t see any clear added value.
“You claim Amazon is too expensive for SMB’s, but my bill is a couple of hundred bucks. I [can’t] touch a bare machine architecture for that. This has been my experience at every SaaS company I’ve worked for.” ~Bob Warfield
If your bill at Amazon is only a ‘couple hundred bucks’, you aren’t using very much in terms of resources. You could probably get a dedicated server for a fraction of what you’re paying now and it wouldn’t take any specialized knowledge to run it. If you’ve never worked at a SaaS that used more than $200 worth of resources at Amazon, you must be working with very small or early stage companies. Particularly in the SaaS world, I come across companies spending many thousands per month on AWS and a cost comparison to Bare Metal usually reveals fairly dramatic improvements in both cost and performance.
“What are you selling, Pedro.” ~Eric Martin
This question of what I’m selling can be understood in two ways. If you’re asking what alternatives I am comparing to AWS, please see the paragraph above on AWS Competitors. If you’re asking what is my motivation for writing this article, keep reading.
Why did I write this article?
I am a consultant so I sell advice. This article is not about whether you should use AWS or not. You’ll notice the article subtitle defines this as a “Growth Study” on how AWS overcame some fairly important business issues to get to where it is now. This topic is important to me because I consult with SMBs and startups on growth strategy & market viability (some other areas too, but they aren’t related to the topic of this article).
I have a heavy focus on digital so IT questions like ‘where to host?’ and ‘what platform to use?’ come up. I currently have accounts at Rackspace (cloud servers & email), IBM Softlayer (dedicated servers), ServerPronto (dedicated servers), and LiquidWeb (cloud sites). Full disclosure: I do not own stock in any of the companies above, but I am a consultant for ServerPronto and my company, WebLift, resells some Rackspace and Liquidweb services to our clients. I have not been paid to write this article.
“Have you ever built a datacenter for your business? I have, it’s an expensive time consuming, resource hog that requires long-term contracts, investment in hardware and infrastructure, and expensive specialized people to run it.” ~Eric Martin
Why would you compare using AWS to building a data center? I have not built a data center and the vast majority of companies will never need to. Building data centers is best left to large enterprises with massive budgets. There other 97% of us should just utilize Infrastucture as a Service (IaaS) or colocation. Neither of these options suffers from the pitfalls you list.
Pedro Sostre has always loved business. When he was 13, he begged for a copy machine for Christmas so he could start a comic book company. The company failed (what did you expect, he was 13?), but his passion for creating businesses has never waned.
He is a Consultant, Author (Web Analytics for Dummies) and Serial Entrepreneur. He currently serves as the CEO of WebLift, Co-Founder of Black Helmet, and holds advisory roles with several organizations. In the course of his career, he’s worked with companies like CBS Sports, BMW Motorcycles, Reebok and Motorola, among others.