Shopping for the Web

It's the summer of 1995. And computer scientist John Wainwright just received a book that he ordered online. It's a pretty technical read–but it arrived with relative ease. A couple months earlier, he was given a beta invite to a brand new startup, a company that was trying to sell books on the internet. And now he opens his package, pulls out the receipt, and reads these words:

Thanks for shopping at Amazon.

John Wainwright was their very first customer. Today, Amazon's headquarters has a John Wainwright building. And every year around the world, we spend more than $3.5 trillion U.S. dollars making online purchases of books and groceries and clothing and... well, everything. And not just on Amazon. An entire ecosystem of online retail has taken hold in the 25 years since that fateful book delivery. The COVID pandemic only pushed us further toward e-commerce. More and more people adopted digital buying habits. Billions of us now shop from our sofas.

I'm Saron Yitbarek and this is Command Line Heroes, an original podcast from Red Hat. This season, we've been unpacking the longest year in tech history–1995. The year our future began. So much started in that time. But few changes were more consequential than the advent of online shopping. In just a couple of decades, it's grown from a handful of startups to a team of Goliaths–and we couldn't live without them. But as we'll see, the dominance of certain online shops is only partly due to brilliant business maneuvers.

Now, like I said, Amazon is not the entirety of e-commerce. But it is the big fish in the pond. So let's begin there. Why, in 1995, did Amazon start delivering packages? What made it possible, and what set founder Jeff Bezos on a path towards shipping billions of packages every year?

He immediately understood the whole idea of transferring currency and making trades with this new use of this relatively new technology, the internet.

Robert Spector wrote the first book on Amazon. It's called Amazon.com: Get Big Fast. And that should tell you something about the Amazon origin story. It's a story about capitalizing on a particular moment. Seizing the day. In the early 90s, Bezos lived in Manhattan. He was making more than a million bucks a year working at a hedge fund that relied on algorithms. Cutting-edge stuff at the time. And it speaks to the sort of business Bezos believed in. A data-driven business. An automated business. Meanwhile, the hedge fund's founder, David Shaw, asked Bezos and a few others to look into this newfangled internet thing. See what kind of opportunities were there.

So Jeff and a few other people looked at various product categories and of all things, books rose to the top as the most obvious product category. Why books?

For starters, the biggest bookstores at the time—Barnes and Noble and Borders–only had 25% of the market. So, you had a lot of smaller players in the field. And one other thing: there was already something called "Books In Print," a digital catalog of, you guessed it, every book in print. Those 2 pieces of information didn't get David Shaw super excited. But Jeff Bezos saw an opening.

Jeff decided that he was going to leave D. E. Shaw, and he was going to start a company that sold books online.

And so, a legend was born.

And this is the early days of the internet when people are starting to get used to the notion of something called a dot-com. He always referred to it as Amazon.com, never as just plain Amazon. That was part of his branding.

If you missed it, check out the first episode of the season to learn about the creation of the domain name system, and how it fed into the dot-com mania in 1995. Okay: so Bezos saw his opening and he's setting up Amazon. He headquarters the company in Seattle because it's close to the biggest book distributor, but what really made it different from all the other dot-com hustles out there?

If you look at how Amazon succeeded, even in those early days, it's that Jeff Bezos understood the technology. He understood what could be done and what couldn't be done at that time. He understood the financials. He had worked in banking, so lots of zeros and commas did not intimidate him. He was a visionary, and finally he understood the value of public relations. So he made himself the face of internet commerce.

In other words, while some book suppliers were planning to use the internet, Jeff Bezos was a creature of the internet, ready to use book suppliers. He had flipped the focus.

Amazon officially launched July 16th, 1995. And by early 1996, they were on a course for an annual revenue of $5 million. So they took hold very, very quickly.

Kleiner Perkins, the venture capital firm in Silicon Valley, invested heavily. And on May 16, 1996, the Wall Street Journal's front page shouted, "Wall Street Whiz Finds Niche Selling Books on the Internet." It was front page news that someone could tap a major profit there. And when people came to their site, they didn't just find books there. They found a boisterous marketplace.

They allowed reviews by regular people and people could comment or agree or disagree on the reviews, so the users were providing Amazon with content.

It all added up to an enormous opportunity, much more than that simple niche that the Wall Street Journal mentioned. And Bezos knew he was capitalizing on a new tech reality that others could capitalize on too, which meant he had to...

Get big fast. That became the company mantra, and that was very much of a Silicon Valley approach. You want to get bigger than everybody else so you can extinguish the competition.

Depending on who you talk to, that mantra reflects Amazon's greatest asset, and its greatest flaw. In the fall of 1995, just a few months after Jeff Bezos got his startup going, a Silicon Valley engineer named Pierre Omidyar was exploring the possibilities of an online marketplace in his own way. Omidyar had built an auction website. And he built it off the side of his desk while working at another company. Nobody really seemed to get why he cared about the project. But, shortly after Amazon shipped its first book, Omidyar listed his first item for auction. A broken laser pointer. Asking price? $1. When he called up the buyer just to, you know, make sure the guy knew the laser pointer was broken, the guy said, "Yeah, I collect broken laser pointers."

eBay had a community orientation that was completely different from all other e-commerce players.

That's Sandeep Krishnamurthy, the dean of the business school at the University of Washington's Bothell Campus. And he points out that eBay was empowering a totally different market from Amazon.

When other e-commerce players were saying, "Let us sell, insert product category here, to a group of customers," what eBay was doing was paying attention to both buyers and sellers. So they were talking about, "How do I create a marketplace very much like what we would see in a farmer's market in your local city?"

If you want to sell a broken laser pointer–more power to you. Someone else will sell first-edition Spideman comics. Vintage furniture. An old coffee mug. It gave the community the power to discover—through millions of trials and errors—markets that other businesses were going to miss. Remember that in 1995, nobody knew what the web was going to be, so inviting the whole world to help define the marketplace wasn't such a bad idea.

Pierre was a techno idealist. He was not really trying to do something very conventional. And so for him, the community was the biggest thing. And that was something that I think to this day—I think what distinguishes eBay from the plethora of companies that have entered—had to do with this community idea.

Pierre Omidyar understood that one person's junk is another person's treasure, and that to him was the genius of the web. It allowed users to discover miniature niches as easily as they could discover a global marketplace. It was a more radical vision than it might seem. Omidyar was inspired by the 1980s distributed discussion forum called Usenet. Like Usenet, his eBay was a democratic vision.

eBay had this ethos of democratizing capitalism itself, that you don't have to always go to some well-established seller who claims to know everything.

Twenty-five years later, the success of that democratization is undeniable. Millions buy and sell their wares on eBay. From Beanie Babies to Game of Thrones, every consumer trend has found a home there. And every tiny niche, too.

I remember even in the early days in the comparison between Amazon and eBay, there was a metric that we used to use called stickiness, which refers to how long somebody stays on a website. And eBay's stickiness was off the charts in comparison to Amazon. There's simply no comparison, because Amazon was a place where you quickly bought something and then you left, while eBay really demanded your attention and engagement.

So, eBay and Amazon settled into their two distinct roles as online shoppers arrived. Bezos and Omidyar were celebrated in all the business magazines. But you know, the advent of e-commerce was not just the result of two dudes with bright ideas. In 1995, a whole landscape of new retail possibilities came into view, and it relied on two crucial new categories of technology. From eBay's broken laser pointer to Amazon's most recent delivery, none of it would have been possible without breakthroughs in security and storage.

Let's start with security. Figuring out how to make secure payments online wasn't always so obvious, and if customers don't trust that it's safe to share their credit card information, well, good luck with that whole Amazon idea.

E-commerce is enabled by the trust of the consumer, and that trust depends on the security of the encryption.

Angela Robinson is a research mathematician at the National Institute of Standards and Technology. She's been fascinated with privacy and encryption ever since she was young, when she used to encrypt her diaries. We got her to explain how something called public key encryption changed the game for online retail. For centuries, encryption worked like this: I lock my message in a box with a key, send you that locked box, and you open it up with a copy of the same key. That's called symmetric cryptography. Senders and recipients have the same key. And it turns out in the digital age, that's a pretty bad idea.

How can you ensure that the sender and the receiver can agree upon the same key? Maybe they are on two different continents. How can we ensure that if a group of people want to share the same key that they all are able to share that key and not risk losing the key or someone mishandling the key? How can you update the keys?

She could go on. But what's the solution? The answer arrived in the 1970s. And it was called public key encryption.

In public key encryption, every participant has their own key pair. So if I have a key pair, I have a secret key and a public key. I can publish my public key anywhere, and all those interested in communicating with me just need to encrypt their messages under my public key.

But here's the crucial part. Everyone can use that public key, but you are the only one in possession of your secret, private key.

Therefore I should be the only one who can decrypt that message. So this totally revolutionized encryption as we know it and brought us into the realm of modern cryptography.

Public key encryption unlocked the possibility of sending your credit card info safely across the web. And in 1994, the inevitable happened. A man named Phil Brandenberger sat at his workstation in Philadelphia. He used his Visa to buy a Sting CD for $12.48. A venture called Net Market Company had produced a powerful new form of data encryption that made possible that first totally secure commercial transaction over the internet. Their 21-year-old CEO boasted that even the NSA couldn't have captured that credit card number. And so, just months before the 1995 launch of eBay and Amazon, the groundwork for transactions was laid. Newspapers cheered about the birth of "a shopping mall in cyberspace." And, well... you know what happens next. Billions of purchases, every day, from phones, from laptops–from all of us. Today, public key encryption lets all that happen by safely sharing bank information and credit card info online. The entire field of e-commerce relies on that technology, and it will rely on it even more so in the future.

Once quantum computing technology is full scale, the algorithms that we're using now to encrypt our data even on these e-commerce platforms will be rendered insecure. So the next phase of public key encryption is post-quantum encryption, and this new standard is in development right now.

In 5 years or so, retailers will rely on a new standard of encryption technology, an enormous aid provided by the public sector. By the National Institute of Standards and Technology, in fact. And remember, all this help with encryption is only one of the major legs up that e-commerce relies on.

So, that other big piece of tech that made e-commerce the Goliath that it is today? Data storage. Before the internet, data storage in retail didn't look like much. You'd have some transaction information that shops would feed back to their headquarters. Unless a customer signed up for a loyalty program, shops just didn't have much info on the people who shop there.

All that changed when shopping went online. Think of all the data you could suddenly collect, and if you could find a way to harness that data, you could target your ads like never before. But a data deluge means you suddenly have to decide where to store it all.

As far as challenges on storage, it was a huge challenge.

Kartik Shastri is a chief technology officer in the retail industry. He describes the early days of e-commerce when data about customers started flooding in. But storage for that precious data was nowhere to be found. Remember, in the mid-90s, cutting-edge storage meant Zip disks and DVDs.

Typically the storage disks back in those days, one, were not very big. And two, they were very expensive. And especially when you got into server storage, it was very, very expensive.

As data collection increased exponentially, retailers were adding servers non-stop. They also had to maintain storage across multiple systems. And so, with storage systems in disarray, leveraging that data into ad decisions felt dicey at best. It would take a full decade before the storage problem was solved. But at last, in 2006, Amazon Web Services arrived. And since then, Microsoft and Google launched their own cloud services, too. Most companies today use a cloud service for data storage. We did an episode on all things cloud, if you're interested. It's season 1, episode 6. But back to e-commerce. The emergence of cloud storage means shops can finally use all that data they've gathered for some nano-targeted advertising. Shastri says the problem now is sifting through all that data.

The speed at which the information comes in is just super fast. So people are still struggling with trying to get the data from all these different advertising systems, marketing systems, analytic systems, put it all together and figure out what the systems are telling them.

In the same way that public key encryption evolves alongside retail, storage tech has been growing to meet demands, too. But here's the thing about all those improvements: together, encryption and storage tech have been giving e-commerce an enormous advantage over older, brick-and-mortar shops. And some are asking whether the advantage has gone too far. At a congressional hearing on October 29, 2020, U.S. Senator Roger Wicker noted that the tech industry's enormous success was largely due to...

A light-touch regulatory framework.

In other words, those online businesses had been left alone. Government had stayed out of the way while they grew. But when companies like Amazon grew to unheard-of sizes, that sparked questions. The U.S. House Judiciary Committee's antitrust subcommittee released a report about competition in digital markets, and they found that companies like Amazon and Facebook simply had too much power. The digital economy had been dominated by monopolies. The confusing part: these new monopolies didn't behave the way railroad monopolies or oil monopolies had in the past. We asked Katie Wilson, a columnist at the Crosscut news site, to explain the difference.

In the late 1970s under the influence of the Chicago School of economic thought, antitrust really narrowed to a laser-like focus on consumer welfare, and in particular prices. And in general, the Supreme Court stopped upholding antitrust actions unless it could be demonstrated that what was at stake was higher prices for consumers.

So, being a huge online retailer wasn't necessarily enough to inspire government regulation. They'd only step in if your customers were being forced to pay higher prices.

How could you pin a charge on a company like Facebook or Google that offers most of their services to individuals for free? Or a company like Amazon that built an empire by offering the very lowest prices? So these big tech companies were able to grow with very little antitrust scrutiny because their model is so different from the idea of a traditional business that sells widgets or whatever.

Congress moves a lot slower than Silicon Valley. And so, thanks largely to public key encryption and advances in storage, online retailers had in a quarter century made the world of shopping over in their own image. Wilson reminds us that when Congress looked at monopolies in the past, they worried about more than just cheap prices.

People saw the social impacts of eliminating small and mid-sized businesses, right? As we saw happen with Standard Oil where all of these smaller oil refineries were just gobbled up. What does that do to the people who worked for those companies? To the people who owned those companies? To the communities that they were in?

The free-market approach, where cheap prices were all a monopoly had to provide, may have allowed companies like Amazon to avoid scrutiny up till now. But some argue that we can demand more from these companies than just cheap prices.

What we're seeing now is that perhaps the pendulum may be swinging back the other way to a broader conception of the goals of antitrust.

We've reached a tipping point where those amazing tools–encryption tech and data storage–have allowed e-commerce businesses to become Goliaths. Tens of thousands of brick-and-mortar stores are expected to close this decade while online shopping continues to grow. The question is, can we take the benefits that our tech provides to companies like Amazon and eBay, and find ways to get them supporting every business, whether they're on the web or just down the block?

In 1995, the earliest days of online shops, the cost of starting up might be hundreds of thousands of dollars. But infrastructure costs have dropped, and I could start an online boutique for 30 bucks this weekend if I felt like it. Those encryption and storage technologies have trickled down so that everybody, even a new indie store, can use them to superpower their business. We know the marketplace is going to continue to rely on advances in encryption, in data storage, and so much more. But the question is: how are we going to use those technologies to make room for everybody?

For more details on the rise of e-commerce, or any part of this season's journey into 1995, you can visit RedHat.com/CommandLineHeroes. We've loaded up our site with fascinating bonus material. Next time, it's a search for the origins of search engines. We're rediscovering that pioneering navigation tool: AltaVista. I'm Saron Yitbarek, and this is Command Line Heroes, an original podcast from Red Hat. ‘Til next time: Keep on coding!