Why this internet boom isn't a bubble

The Internet bubble of 1999-2000 saw the rapid rise of many internet ideas. At the time, it was all about gathering as many active users as possible, with revenue and certainly profit a secondary or tertiary concern. During the rapid descend (aka "flame out") of the same ideas, the notion of eyeballs got a bad taste. I believe the companies were fundamentally right, but lacked the critical tools to be successful. Today we have those tools and find that the current boom is nowhere near bursting.  

It was all about Eyeballs

As far as Silicon Valley goes, 1999 were the "good old days" - anyone with a business plan that included the word "web" seemed to be able to get sufficient funding to throw major parties. The particularly lucky ones received a mention in the Red Herring magazine, which guaranteed even more parties. At the time, we were also reminded that Pet's can't drive by the infamous pets.com that spent most of its revenue (there were no profits) on advertising. I still have my sock puppet.

As new companies popped up left and right looking to get funded, valuations took center stage. However, how do you value a company that has no profit and sometimes even no revenue? The solution at the time was to value a company by how many "eyeballs", i.e. active users it could attract (whether the numbers were appropriately divided by 2 is a well-kept secret of the VC community of the day). If I recall correctly, valuations ran in the order of magnitude of about $40 per user. Ironically, compared to today's internet giants' valuations this figure might appear outright cheap - there aren't enough human eye balls on the planet to multiply this into the three-quarter-trillion Dollar valuations we see today.

Alas, as high as these companies flew, they all came crashing down pretty hard by the end of 2000. The Y2K doomsdayers' predictions of a major IT shake-up in the year 2000 thus came true, but it didn't have anything to do with the way dates are encoded. The flaming fireball also dragged the notion of valuation by eyeballs down with it: the approach was quickly put aside as a crazy thing of the past whose lack of foundation somehow went unnoticed by the whole investment community, at least on the West Coast. It must have been all those parties.

It's still about eyeballs

In hindsight, though, I am convinced that those companies generally had the right idea. Not by spending all their revenue on advertising, but by focusing on eyeballs. The power of "owning" the customer, or more precisely the interaction with the customer has been proven by some of the current internet heavyweights. Take Netflix for example, who was able to move up the value chain from distributing content into content production - an area that used to be the well-guarded domain of media giants. Or remember how Apple rocked the music industry with the iTunes Music Store, also by owning the customer's eyeballs, or, well, ears.

Eyeballs fuel the digital economy

The digital world thrives on experimentation and rapid feedback. The closer you are to your customer, the deeper and more up-to-date feedback you can get. That's what makes customer proximity a critical ingredient as it enables the rapid feedback cycles that drive disruptive innovation. So, owning eyeballs helps drive innovation. Customer proximity is also a revenue generator in a world that makes the word "information overload" feel like an understatement. Customers engage with companies who understand their preferences and make matching recommendations - think Amazon and Netflix.

I believe two main reasons the companies in the first bubble, despite having the right idea, didn't succeed were the lack of cloud platforms and machine learning. Parties are OK!

Cloud Platforms

The start-ups of the first bubble spent considerable amounts of time and money building up infrastructure, racking and wiring tons of Sun E10K servers at up to $ 1 million a piece. Next, they spent time building and installing frameworks to run their applications at scale. This approach made Sun happy, but not the start-up investors as large portions of the funds and energy were spent on infrastructure instead of innovative customer features.

Cloud platforms have changed this completely. As fascinating as proprietary hardware like the E10K were, they were quickly superseded by horizontally scalable software running on mostly commodity Intel hardware. Next, cloud platform providers managed to automate virtual machine deployment, relieving start-ups from the need to build and fund their own infrastructure. The rest is history, so to speak: elastic pricing and instant provisioning allow you to start a company with an IT infrastructure budget of less than a hundred Dollars - the cornerstone of innovation democratization. Modern companies can build their solutions on top of managed Kubernetes clusters or even as cloud functions without any need for infrastructure deployment or management.

Machine Learning

To make eyeballs count, one needs to understand and predict customer behavior. This is largely performed through analytics and machine learning. You'll captivate customers by showing them other items or movies that they will like or by making their engagement natural through voice and vision interfaces (think Amazon Echo or Google Home).

Progress in this area has been enormous in the last couple of years. Detecting cats in photos has long made way for complex object detection in video and sentiment detection in voice data. Facebook can predict break ups and Google AdWords can target people who recently graduated or got married. With these capabilities widely available as cloud services, start-ups can now turn their eyeballs into profits.

Raising the bar

Having cloud capabilities for compute and analytics allows start-ups to innovate with a minimal invest in IT infrastructure. That's one reason the current wave of start-ups aren't in a bubble but are here to stay. This also means that to compete with this breed of company, incumbent players must embrace a similar approach. Ordering and setting up infrastructure in a proprietary data center slows you down and distracts you from providing customer value, even if your business has the funds to do it. It's no longer about size and financial might, but about speed and customer focus.

Tell me more

If you enjoy reading my posts, please re-share them with your network.

Read more about what one chief architect knows about IT transformation at https://architectelevator.com.