Montage Sequence #2 - bubbles, loonshots and the OA

Welcome to the 2nd ever “Montage Sequence”. Most of this was half-written last week but between my first week back at Twitter after paternity leave and the logistics of handling a baby at home, I had to wait for a flight this Friday to finish it. I miss the days of the mid-2000s when you would see people update their blog apologizing for not having written in a while and then hop on over to to see who had written recently. Those were some fun days.

Without further ado, onto this installment. Please send me any thoughts or feedback to or tweet me at @sriramk.

Tech Bubbles

I had a tweet go surprisingly viral last week. It was a collection of various press pieces over the last decade or so on the implications/predictions and fear of a bubble in Silicon Valley. It’s always hard to reverse-engineer why something suddenly catches fire but this seemed to hit a nerve in the midst of many competing narratives in tech and media. What I truly enjoyed was the conversation that ensued, both on Twitter and via private DM from people who both loved and hated it.

I came away with a few thoughts.

1/ The trauma of the dotcom bust has left a deep-seated fear of a return to the bleak years of the early 2000s. This, combined with the focus in recent years on wealth inequality and the impact that tech companies have on the world in multiple ways all result in a lot of anxiety. This bubbles (no pun intended!) up in many ways and often not well-understood by those of us who work “inside” the tech world.

2/ It’s worth examining whether we have over-rotated after the dotcom bust on who can invest in companies. There have been multiple articles on how a lot of the upside on the recent IPOs (say Lyft going from $4m to $20b+ in valuation) have gone to a select few institutions and individuals. It is worth asking then: why was this not available to the regular mom-and-pop investor? Is there a way to structure it so people don’t risk losing their entire savings or pensions over a scammy fly-by-night bust?

I’ve been starting to see some interesting ideas in this space including a) having investors take a test similar to the DMV driving license one before investing b) private company investments restricted to only a % of your net worth. There are some interesting ideas here coming from the crypto space after the spree of shady behavior in the ICO world.

The issue I care about here: how to get normal people here in the US ( not just institutions or high-net-worth individuals) access to one of the strongest drivers of wealth creation in recent times. In Forrest Gump, buying stock in “a fruit company” famously makes Lt. Dan and Forrest rich beyond their wildest dreams. How do you get  Forrest and Lt. Dan access to that in 2019 without the risks of the dot com bust?

3/ The unicorns going public now are fundamentally different structurally from the companies in 1999-2000. Software may not yet have eaten the entire world but has definitely eaten a lot of industries. First, these companies have been around a long time. Lyft, Airbnb, Uber, Pinterest, etc have all been around for a decade. Second, it’s clear that these companies are now here to stay driven by their underlying network effects and economies of scale: it’s hard imagine a world where Airbnb is not around a decade from now.

It may be time to stop grouping them together as “tech” and instead looking at each on its own merit as just regular ol’ boring companies.

4/ Finally, I was struck by how all the companies mentioned in the “bubble” pieces - often in reaction to the sticker shock of what seemed like a large valuation back then - went on to bigger and bigger valuations as time went on, often many times over. Which leads me to wonder:
a) why did so many commentators miss the growth that was going to happen to these companies?
b) is the same mistake being made now in the narrative around valuations?.


One of the best “business” books I’ve read in recent times is Loonshots by Safi Bahcall. When I first head the title, I was skeptical and assumed it to be yet another mediocre work promising the secrets of Google-esque innovation to a mid-level exec. But what I found was a real delight and a great framework on how to think about risk and innovation inside businesses.

Let’s start with the author.  After reading the first couple of chapters, it was clear this was no ordinary management non-fiction writer trying to expand a blog post into a book with a catchy title. I set the book down and Googled the author. You immediately see why the book has the gravitas it has. Bahcall is a well-respected entrepreneur in the pharma world and someone who has actually built companies in a challenging space.

This guy is the real deal.

Onto the book itself. Bahcall walks through multiple famous business case studies (Pan Am vs American Airlines, Apple, Kodak, etc). He does it with flair and a focus on the quirky characters that make up these companies: the book is entertaining for that alone. But deeper in, he reveals his real masterpiece.

Bahcall splits all “loonshots” (think of them as pattern-breaking innovation) into two kinds: P-Type and S-Type. The P (“Product”) Loonshots are those companies that competed by focusing on product innovation. Pan Am famously built planes no one else could. But what gets interesting is the “S-Type” Loonshots. “S” for Bahcall stands for “strategy” but I believe a better name for this would be “Operational Loonshots”. He breaks down cases where innovation doesn’t happen from “product” (new kinds of planes) but “operations” (yield management on pricing, the invention of frequent flier miles, optimizing routes). He walks through multiple case studies where companies fail by focusing on just one-kind to another. 

Over the years, I’ve come to realize the same in the tech world: it is key to understand what the dynamics of the market you’re in and then figure out whether you should be innovating on product or strategy/operation. The most common mistake I see here is the wrong team fit: asking a team focused on pattern-breaking innovation to focus on optimizing an existing process (or vice versa) is a recipe for failure that I see all too often.

There’s a lot more here in this framework. I highly recommend this as a read. 

The OA

The OA’s Season Two showed up on Netflix about a week ago and let me just say - IT IS AMAZING. My wife and I watched Season One and had the same reaction as many others: interesting, weird, uneven but very different from anything else on TV. And a lot of interpretive dance. Season Two is something else altogether.

First, unlike the JJ Abrams-Lindelof trope of having a mystery that is never answered, S2 starts with resolving the central mystery of S1. But then it gets weirder and weirder and more fun with every episode. I have to resist from giving away details but by the end, Aarthi and I were messaging about  the previous night’s episode  during our workdays.

Second, it is bold. The OA takes risks, goes places no show has gone before and takes the “weird” feeling of S1 and dials it up. I have rarely seen a show break so many patterns so quickly.

Third, I love the San Francisco-sci-fi vibe. It makes San Francisco look beautiful, mysterious and interleaves in modern tech in a way that feels very 90s cyber-punky. 

And finally, it might have the biggest “WTF …did they just do what I think they did?” season finale of all time. It’s up there with the classics like Lost’s flash forward. However, unlike many other finales, they attempted something truly audacious and somehow made it…work. Hats off to Brit Marling and Zal Batmanglij.

I loved it and highly recommend it as a watch (you will have to watch S1 first).

That’s it for now. Please send thoughts and brickbats my way at or tweet me at @sriramk. Until next time!