The Containerization of Everything

One of the major forces in computing right now is the deployment of software via ‘containers.’ From Docker: “A container…is a lightweight, stand-alone, executable package of a piece of software that includes everything needed to run it: code, runtime, system tools, system libraries, settings.” Software is deployed through a container to make sure that it runs the same no matter what system it is on.

I think that this trend is exactly the same as what is happening in the consumer durable and certain sectors of the CPG markets. Before I unpack that statement, let’s discuss why software containerization has become so popular.

The idea behind containers is quite simple: software is running on a wide variety of devices, and the developer wants to ensure that the software runs the same everywhere. TechCrunch has a great 101-level piece on containers if you’d like to read more.

There are many forces at work behind the development and proliferation of containerization, but I’d like to call out two in particular.

Varied operating environments: Computing was first done by a large mainframe, then individual desktops via a client-server relationship, and now via mobile-cloud. Computing is done on a large number of differentiated devices connected to a central OS endpoint via the cloud.
Abstraction away from source code: Everything done on software ultimately goes back to the 0s and 1s communicating with the CPU. However, services are increasingly built multiple layers above the binary. Think CPU -> server -> AWS -> Intercom

Containers, and the software they enable, are empowering companies to deliver focused digital solutions to singular customer use cases. And networks like Instagram are enabling consumer brands to do the exact same thing!

I wrote last week that brands can use Instagram in a targeted manner to create feedback loops within their desired demographic. These brands capitalize on Instagram’s uniform consumer experience to sell a solution to a specific use case (#firstworldproblems). A few varied examples:

  • Bevel (started with a razor designed to protect the skin of black men, built into a cosmetics brand for people of all colors)
  • Bonobos (started with one product, a better men’s chino pant, recently sold to Wal Mart for $310 million)

Andy Dunn, the CEO of Bonobos, has a fantastic piece about the need for a brand to nail one product: “If you don’t start with a relentless focus on an amazing first product, odds are you won’t even get a seat at the table” + “the best way to get volume is to sell a lot of one thing, not a little of a lot of things.” + “Make one thing great. Get one thing right. That earns you the right to go from product one to product two.”

Containers -> software; Instagram -> consumer brands.

Life many software startups, modern consumer brands use web platforms to reach users trying to solve specific use cases – often these use cases are ones they didn’t know they need solving.

To Dunn’s point – most don’t earn the right to go from product one to product two. More on that soon.



Instagram is the Ultimate Consumer Accessory

Much has happened since I last wrote a piece here.

I started working at Intercom in a challenging Sales Ops role. It’s been a whirlwind, and I love working there. We have a world-class blog and podcast that you all should read and listen to if you don’t already.

But now I’m back to thinking and writing critically about the important things happening in the tech world.

I first mentioned Instagram in this post from January. At that time, Snap was two months away from their IPO and one month away from filing their S1. I wrote, in more or less certain terms, that Snap was in trouble because a) Instagram Stories were crushing it and b) brands weren’t getting effective results from their ads on Snap.

These two trends have continued seemingly unabated.

I also noted that Snap’s confusing design was going to be a hindrance to more people using it. Sure enough, Snap announced a redesign of their app to make more clear the separation of content from friends and content from publishers.

We’ve also since learned that it’s not actually Snap’s core messages business that is the issue. The Daily Beast published an article yesterday with leaked numbers from a five-month period in 2017. During the time in question, daily messages sent on Snapchat actually rose from 80 to 87 million, which is a >20% annualized growth rate. That’s not bad. However, Snap’s other products saw flat to declining usage over the same time period.

At this point, Snap is the owner of a very popular chat app with millions of young users who each send hundreds of image-data-rich messages every day. Without a separate source of concrete data about its users, I don’t see a clear path to monetization.

However, 2017, to me, was less about Snap failing to execute and more about Instagram executing at the absolute highest level. 2017 saw Instagram become the most relevant social network outside of China.

Instagram has become the best tool to build a brand, ever. People who are champions of a brand not only get targeted ads from that brand directly in their feed, they actively sculpt and tailor their accounts into the hyper-realized version of that brand’s vision. On top of that, Instagram feeds these images back to its users in a cycle of self-perfection and realization.

It’s as if a brand held a focus session for a new product, and each of the participants, regardless of whether she truly liked the product, changed her opinion to suit the product, and then went out of her way to purchase that product and modify her life to include more of it.

I speak in extremes not for gratuitous effect but to try and illustrate just how effective this platform is. We’ve never seen anything that links aspirations, products, and people so effectively.

A vibrant Instagram has become the ultimate consumer accessory.

I don’t think this is a good or healthy thing.

Two major shareholders of Apple recently pushed for the company to create greater controls for parents over their children’s device usage. They did this out of concern that these devices have created a rapidly worsening health crisis of addiction to electronic devices.

I think that 2018 will see more of this – the largest tech companies struggling to grapple with the adverse side effects that their products have on consumers. 2018 for me will see me writing shorter posts more frequently.

In the words of Scott Galloway –

Life is so rich –



  1. Intercom website:
  2. Intercom Blog:
  3. Intercom Podcast:
  4. Instagram and Snapchat:
  5. The Verge – Instagram Stories is now more popular than the app it was designed to kill:
  6. CNN – Snapchat redesign:
  7. Daily Beast – This is the Data Snapchat Doesn’t Want You to See:
  8. Jana/Calpers Push Apple to Study iPhone Addiction in Children:

Abstracting Innovation

I just finished The Upstarts by Brad Stone. The book details the founding and first few years of Airbnb and Uber. Its coverage of Uber’s founding is particularly poignant given the recent anger towards that company. If you want to read more about that story, go check out my post from last week or read Daniel Compton’s post on the potential ramifications of Google’s lawsuit against Uber and Otto.

The Upstarts also got me thinking about the ideation of new technological innovations. In this post, I’ll try to illustrate a framework for thinking about the growth and maturity of similar new technologies. (As I started writing, it quickly became apparent that there was more content here than I originally thought. As such, I think this will be a three-part post that I’ll flesh out over the next couple weeks).

The Innovation Cycle

All new technologies go through a cycle whereby they are invented, introduced as a unique feature, and commoditized. After commoditization (ubiquity), they either go extinct or are reborn in a new iteration with the benefit of underlying technological improvements. I’ve attempted to illustrate this relationship below.

Innovation Cycle.jpg

In this model, new technology arises in a few different ways. It is then launched as a unique feature controlled by a small number of stakeholders. The technology then becomes pervasive and permanent through commoditization, which means that it becomes an interchangeable good with no meaningful qualitative differences across its providers. Finally, the feature either goes extinct, finds new use cases downstream, or is reinvented due to new upstream technologies. Let’s break down each step of the process.


Historically – although there are exceptions – new consumer and enterprise technology came about in one of three ways:

  1. Government research
  2. New social norms
  3. Invention out of necessity

Government research is responsible for most of the ‘general-purpose’ technologies that the modern world couldn’t live without: GPS, the internet, RAM, and microprocessors are just a few examples. However, most of these technologies were invented many decades ago. Recently, innovation has shifted to the private sector. There are a number of reasons why this happened – I’ve provided a few below:

  • Decreasing cost of computing power
  • Distribution of computing resources available to everyone
  • Proliferation of graduate research students including many from formerly developing countries starting private companies
  • Changing nature of corporations to include research labs and venture departments internally

Basically, the tools needed to create innovative products have been democratized to the point that they are freely available to all (the innovation cycle itself has been commoditized!). Today, new innovations are either born out of changing social norms and generational shifts or are created by companies to solve their own internal problems.

A quick look at the raft of large private technology companies today provides a decent backdrop for thinking about how changing social norms are responsible for the formation of many of today’s most impactful and innovative brands. Three of the largest private U.S. tech companies have a product that offers access to shared goods and services. These are, of course, Airbnb, Uber, and WeWork. Ignoring the necessity of the smartphone/internet to these products, I contend that each of these business’ products would have been largely unfathomable 50 years ago. If you asked consumers in the 60s whether they would be comfortable sharing their home or car with strangers, or if you asked companies whether they would be willing to work in an open shared space next to their competitors, I think all parties would have responded with a resounding no.

The Great Recession created ~2 generations of consumers that are fundamentally averse to debt and ownership. While the seeds for the sharing economy were likely planted with the growth and maturity of millennials and Generation Z, the lasting scars and slow recovery of the economy since 2008 have created millions of consumers that are resistant to the commitment of ownership and the debt that comes with it.

Modern innovation is also born out of necessity. If we continue looking down the list of valuable private companies, we can see a few examples of companies whose products were created as a tool used internally within a company. Slack, the popular team messaging tool, was created within a gaming company to help improve communication and collaboration internally. Cloudera, the database software company, was founded by Yahoo, Google, Facebook, and Oracle engineers to help bring Hadoop to industries other than technology. Hadoop is an open-source distributed database tool that was created within Yahoo.

While many of the primary innovations that occurred in decades past came about because of hardware innovations funded by public entities, the explosion of software and IT technologies coupled with social and commercial trends is largely responsible for the ideas impacting our world today.

New Features

Shortly after its invention, a new product or use of technology enjoys a short-lived period as a vertically-integrated product/feature unique to its inventor. During that period, users flock to that service primarily due to the novelty of the new product. After the initial exuberance, one of three things happens:

  1. The initial surge of users and interest is only hype and the product slowly dies out. There may remain a dedicated base of loyal or niche users, but the product remains essentially unique to its creator. I think Houseparty is at this stage in its lifecycle.
  2. The product slowly gains users but adoption isn’t as widespread as the initial trajectory suggested. Total users are asymptotic to a decent and sustainable number. Competitors may offer similar products but new companies aren’t formed to copy the new feature. The product reaches permanence and acceptance in popular culture but continued attention doesn’t substantially grow its user base. I think Twitter is a good example of this.
  3. The product gains traction outside the initial surge of early adopters. Adoption continues and new use cases are invented regularly. There is a strong mix of niche use case, hardcore users, and casual users. Competitors spring up to copy the initial concept and add their own differentiating factors, but the core product remains essentially homogeneous across all providers. The product becomes a standard feature in many different use cases and is no longer seen as a differentiating factor. Most people will eventually have some sort of contact with the product. The product will eventually go off the market when the next generation is introduced. I think the video chat that Skype helped pioneer is a good example of this.


The third case above is, of course, commoditization. When a good or service is commoditized, it becomes readily and cheaply available from a wide range of providers. While different providers may offer slightly different features to entice consumers to buy from them, the product from one provider is largely indistinguishable from that of a competitor.


The chart above illustrates the process whereby a feature is invented and then becomes a commodity. Most technology hardware has been commoditized. Cloud computing is in the final stages of commoditization right now. Software providers have to constantly reinvent and improve their products to avoid commoditization.

Commoditization means that the field is level for all players. Everyone gets to start from the same point because everyone has the exact same resources available to them (except human capital). The commoditization of IT and cloud computing has enabled the rapid growth and massive scale of today’s tech leaders. This trend will only continue, although the next generation of technologies like machine learning and AI might resist commoditization given their need for massive data sets to function.

Snap is at the point in the lifecycle of its product Snapchat where it is being tested by the forces of commoditization. Snapchat has two core features: 1. disappearing picture and video messages and 2. stories. While it has launched other features like Discover, I think that the core of Snapchat’s user base primarily uses the service for one of the two reasons listed above. So, then, given that Snapchat’s main features are simply software products, why would they resist the forces of commoditization when so many other products haven’t?

The crux of that question is whether Snap is a brand or a feature set. There are hundreds of clothing brands that make black men’s t-shirts. I have my favorites, and I like them for often bizarre reasons. So, is Snap an enduring brand, much like J. Crew has held on to a category of products for so long? Will consumers continue to use Snap when its main products are available elsewhere? Or is Snap’s feature set simply a new invention that others will implement in their own way (say, on iMessage or Instagram)? Did Snap simply shift the model of media consumption for others to implement on their own channels?

My next post will talk about why some products and companies resist commoditization.


New technology innovations may seemingly come out of thin air, but shifts in generational values and corporate models are largely responsible for most of the new technologies we have today. Once a new product is introduced, its owner enjoys a period of time where it owns the entire market. Once the novelty wears off, the product goes through a period where it either achieves mass adoption or sees rapid extinction. In the former case, imitators move in and the product becomes a low-cost commodity good.

In the next post on this topic, I’ll go through some case studies of technology commoditization. In the final post, I’ll explore why some technologies resist commoditization trends.

If you like reading these posts, please be sure to subscribe to receive new posts by email. You can do that at the top of the page on desktop and at the bottom on mobile. I would also love to hear your feedback via email or in the comments section below.


  1. Daniel Compton: The Uber Bombshell
  2. TED Blog: Mariana Mazzucato
  3. Beating the Commoditization Curve
  4. Forbes: Shenzen’s High-Tech Empire
  5. HBR: It Doesn’t Matter
  6. BPD: Search and Such
  7. BPD: Snapchat vs. Instagram
  8. TechCrunch: Introducing Instagram Stories

Tech’s Hidden Bubbles

I’ve been dismayed by the situation unfolding over the past week in response to Susan Fowler’s incendiary blog post. If you haven’t read the post, you should, but if you haven’t, it’s helpful to know that Susan was an engineer at Uber from late 2015 through the end of 2016. Susan contends that Uber not only turned a blind eye to a repeated sexual offender but that the company actively rewards those who exhibit Machiavellian behavior to the detriment of other coworkers.

Continue reading “Tech’s Hidden Bubbles”

Spending in the Future

Americans have stagnating real incomes and seemingly endless rising housing prices. However, spending has decreased notably across a wide range of goods. Ther are a number of innovative tech companies tackling the underlying issues at work here, and the next 50 years will see a growing selection of higher quality goods and services at more affordable (or at least more equitable) price levels.  Continue reading “Spending in the Future”