Tech Takes a While

Qwest predicted Netflix, Hulu, and HBO GO in 1999.

By Micah Rosenbloom, Partner

Erin Griffith wrote a great post urging people to temper the hype surrounding the Amazon/Whole Foods acquisition, essentially arguing that Amazon will be a disruptive player in the food industry, but don’t expect them to scarf down market share immediately. It’s an important point. Food represents 13% of US GDP and many of the deal’s boosters are underplaying the sheer volume of innovation in the food category at the moment.

QSRs like SweetGreen and Chipotle have redefined fast food while supermarkets like Whole Foods and Wegmans are increasingly doubling as cafeterias/restaurants with huge takeaway selections. Recent entrants in food delivery — Seamless Grubhub, Postmates, Uber Eats, etc. dramatically changed the ease and frequency of food delivery.

We invested in Plated, Gobble, and SkipTheDishes (Acquired by JustEat), and with the exception of Uber, I’ve never seen companies grow revenues so quickly. In retrospect it makes sense, most of us eat every day, multiple times, but still, it’s astonishing that a group of companies that are less than ten years old now account for $10B in consumer spending.

As Erin notes, many armchair analysts are beginning to write off the packaged meal category, as well as grocery stores more generally, based on the current Blue Apron share price, but let’s remember that this stuff takes awhile. The force of change in food has only just started.

Change in Tech Takes Time. Lots of it.

As a reminder, this commercial which premiered in 1999 at the height of the first dot-com boom.

At the time, the promise of “every movie, ever made, in any language, day or night” was as fantastical as a flying car to most people. This was a time when Blockbuster, a store-sized version of Redbox popular in the last century, still issued fines to people who returned videotapes that weren’t rewound.

Techies and VCs of course saw the potential and started building it, but nearly 20 years later, we’re still a ways away from enjoying that enticing value proposition. Sure, I can watch Orange is the New Black on any device that has an internet connection, but if I want to stream The Americans I need to sign into my Prime account. Game of Thrones adds a third app, and yet another subscription fee to the mix. There is a cottage industry of websites explaining which shows are coming to, or leaving, a given streaming service at any given time. A great many of our cultural treasures — ER and The Fresh Prince of Bel-Air to name but two — remain in inaccessible IP vaults.

From a technologist’s point of view, video on demand is one of the simpler products to be built. It’s just packets over the internet, after all. But even in today’s world of widespread broadband, DRM, and omnipresent smart TVs and set top boxes, we’re still managing multiple services and remotes.

How to think about self-driving cars, AI, and ecommerce

Now imagine how long it may take one of today’s hot technologies, like self-driving cars, to take hold. If you read some pundits, they believe truck drivers will soon be as obsolete as fur trappers. Let’s examine this hypothesis critically.

If it’s taken us 20 years to stream Three’s Company reruns, do you really think we’ll have technical solutions to route through Boston or SF’s backward traffic patterns, unexpected detours caused by exploding water mains, falling cranes, drunk drivers, and distracted pedestrians in short order? How quickly will the business case for fleet ownership be hammered out? Will 50 states be able to coalesce on laws governing the design of autonomous vehicles and licensure for the operators or coders in a timely fashion? Do we need new infrastructure, like sensors embedded in streets, to make fully autonomous driving work? Will progress be slowed the first time an AV clips a kindly grandmother crossing the road? Will the drivers of these trucks, who often pull double duty stocking shelves at their destinations, be replaced with robots? These are not problems that are going to be solved in the next year. It’ll probably take a decade or more. That’s why smart companies, who are able to play the long game, are investing in companies like Cruise today. Self-driving will only be a reality if companies are willing to do the hard work to make them a reality.

For those panicking that AI will put most workers out of jobs, remember that Amazon’s vaunted machine learning recommendation engine is regularly a punchline in tweets because it advertises products that we’ve already purchased. Amazon has been grinding for over 20 years and ecommerce still only accounts for 10% of retail sales. Ebooks should have replaced dead trees by now, but paper has proven to be as resilient as Katniss Everdeen and their electronic counterparts have plateaued at 50% market share.

Customers are weird and unpredictable. Regulation often routes technology in suboptimal directions. Established players can be surprisingly durable and loathe to adopt new technology. Food is not videos and is not AI and each will follow its own path, probably slower than you think.

It’s fun to make bold proclamations about the future, but the perspective of history tells us we should be more circumspect. As Bill Gates said “We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next ten.”