Tuesday, February 7, 2017

How to Change the Ratio Now

One of our core values at 500 Startups is diversity and inclusion. It’s well-documented that women are dramatically underrepresented in the technology industry including startups and venture capital.

So in Silicon Valley today, how can you get 50/50 (or more) gender balance in a growing organization of 100+ people?

The technology industry pipeline problem exists for women with barriers at every life stage, from being discouraged to pursue math and science as children, to lack of support for balancing work and parenting as mid-career professionals and many steps between. No one can eliminate every source of structural discrimination overnight.

But what if you still want to change your organization today?

I manage the 500 Startups Growth Marketing and Series A team. Here’s what we do. It’s very simple.

We aim to hire the best candidate for the job and we focus 100% of our recruiting effort on women.

The nature of our work means that we have large industry networks and receive  a consistent flow of high-quality people through referral. Logically, this inbound flow reflects the tech industry and growth-marketing average demographics. To counteract the averages, we primarily hire through active, outbound sourcing of referrals from our network and focus our effort on women. We work extremely hard to meet 3 or 4 women for every man when we are trying to add to our team. And this gets us very close to a 50/50 ratio.

Do this well enough and your organization will develop a reputation that precedes you. The most recent person we  hired told me that she only considered joining after hearing from trusted friends about our commitment to diversity and inclusion.

I’ve described one method that we use to diversify our staff at 500 Startups, but here are many other dimensions to diversity and inclusion.  

Changing the ratio is hard but it’s not complicated. Just start doing it.

You can follow me on Twitter here: https://twitter.com/matjohnson

Friday, January 27, 2017

Profit Pools in SaaS and Infrastructure

Much has been written about Network Effects and Platform Potential and New Monopolies in consumer and/or social products. It’s easy to understand how a marketplace like Airbnb or a messenger like Slack build defensible moats through network effects. It’s even fairly obvious how an infrastructure platform product like Segment can build self-reinforcing momentum that takes profit share away from their analytics partners.

Elsewhere in SaaS and infrastructure, it’s less clear which layers and products in the tech stack will be the ones that profit margin naturally flows to, and which layers will become commodities.

Opinion (and bets) within the sophisticated venture and startup community is always divided. In fact this question is one of the most central to venture investors: given equivalent funding and team quality, which products, approaches and categories have an unfair structural and strategic advantage?

I like to think of this question as analogous to core value investing concepts like Warren Buffet’s quip about preferring businesses that even ‘your idiot nephew could run’

For example, a very smart Series A investor I know told me a couple of years ago that he had a strong thesis that test automation services would tend towards commodity, but that margin and pricing power would pool around server monitoring services.

Smart investors in test-automation companies have a different thesis.

One reason that test automation products may tend toward commodity relative to monitoring products is that a series of passing test runs and successful code integrations does not generate the same uniquely valuable time-series data asset that server logs do.

As an investor, I follow analytics and BI products pretty closely. An interesting conversation that I  have been having with experts recently is the relation between Amazon AWS products like Redshift and Quicksight and the current generation of venture-backed BI companies like Periscope, Chartio, Looker, and Domo. Many of these products pull customer data into Amazon Redshift and then build visualizations and dashboards from there. Now Amazon offers a similar native product: Quicksight. The platform and associated pricing power risk to these startups is obvious. The alternative thesis is that the Amazon product will always need to offer 100% configurability and complexity so won’t fulfill the fundamental promise of the startup BI products to democratize data analysis down from engineers to line business analysts and managers.

As in the first example, there are very smart venture investors on both sides of this bet.

The core value-investing literature on businesses with defensible moats and lasting franchise value with margins of safety from Buffet, Munger, Klarman, et al. provides amazing insight for venture investors operating in what appears superficially to be a completely different asset class. For further reading, some of the smartest writing on technology investing like Bill Gurley’s Above The Crowd and Tren Griffin’s 25IQ really connect the dots between the two disciplines.

You can follow me on Twitter here: https://twitter.com/matjohnson

Sunday, January 1, 2017

Founders Are My Religion

My parents raised me atheist. The first time I attended a wedding, I felt deeply uncomfortable sitting in the church. I didn’t find my religion until last year and it took several years of experience as an investor to develop a way to articulate it.

I’ve come to realize that founders who I deeply believe in are my religion.
Charlie Munger and Warren Buffet say they like to invest in management teams they ‘love and admire’. I haven’t been able to get that phrase out of my head for years. George Zachary says he ‘backs revolutionaries who are changing the world’. I love that phrase too.

As an early stage venture investor, belief in the founder is really all I have. I back entrepreneurs at the beginning of many years of ups and downs. The product will change a lot, the market will change a lot, the customers will change a lot, the world will change a lot. I can’t predict any of that. What I can see today and what will continue to be true in the future is the grit of a founder and her ability to articulate a big vision.

At the end of the afternoon on Friday, the last business hour of the year, I met a friend whose company I backed. Five minutes after arriving, he had to step out and take a call to close one last deal for the year. When he came back, my friend told me his company had grown 50% in the last quarter and expected to grow 50% again next quarter on multiple millions of annual revenue. Definitely someone I love and admire.

Last month, I caught up with another CEO I backed. I invested a couple of years ago. His company has gotten to some scale and doing well, but ever since investing, I’ve thought that a major inflection point is always 4 to 6 months away. Still do. My job is to give the advice I can, but most of all to suspend disbelief, eliminate the doubts I can, and back the founder to the hilt. 

Sometimes it takes a lot of effort and reflection and self work to steel my belief through the ups and downs in advance of sufficient proof. That helps make me the better person I’d like to be.  

To kick off 2017, I’d like to thank the visionary, gritty founders who let us at 500 Startups join them on their hazardous journeys. You make me a better person. Thank you!

You can follow me on Twitter @matjohnson

Wednesday, September 25, 2013

2013 Tech IPOs - Healthy and Power Law Distribution

The tech IPO pipeline has been healthy through Q3 2013. 15 companies are now actively trading on US public markets at a collective market capitalization of over $20B at the time of writing. A number of notable venture-backed companies like Chegg, Violin Memory and RingCentral have not priced yet. This is plenty of public-market liquidity to keep the traditional technology venture and growth capital industry thriving. There is much talk of 'disruption' to the VC industry, but this discussion almost always only means changes in the seed-stage deal-sourcing environment. The back-end IPO market is healthy and consistent.

The most notable feature is the excess returns that accrue to the winners even within the select group of companies that IPO. Similar to 2012 when Facebook's market capitalization dwarfed the rest of the 2012 IPO companies, Twitter's estimated IPO value of around $15B compares favorably to the $20B of combined market cap of all tech IPOs through the first 3 quarters of this year.


Tuesday, September 24, 2013

Probability Vs Possibility

People who know me well know about my fierce and longstanding intellectual allegiance to Nassim Taleb of Fooled By Randomness and Anti-Fragile fame. I claim indie-fan cred for jumping on the bandwagon after reading the original Malcolm Gladwell article about his anti-fragile hedge fund, Empirica - before Black Swan was published.

Most people, lay, journalists, experts, economists still have a hard time with the difference between what's probable and what's possible.

Today, I noticed a small example from daily life, that almost anyone should be able to understand:

I was doing work on my laptop at a coffee shop this morning and had to use the bathroom. I thoughtfully waited until the bathroom was empty, then left my laptop on the table, got up, and made sure to return quickly. The time I was away wasn't more than 30 seconds. In this situation, I am quite confident that it is very unlikely - improbable - that someone steals my laptop, so I don't feel like it's a big risk.

However, this in no way makes it less possible for someone to steal my laptop while I am gone. Those 30 seconds could coincide with a thief coming into the shop just as easily as any other 30 seconds in the day. A thief who sat near me and waited and watched for me to get up with the express intent of stealing my laptop could easily take it. The possibility of theft is not reduced, even though I feel good about the low probability of theft occurring.

Monday, August 5, 2013

MSFTs New Office Suite

Microsoft’s new office suite:
Knowledge work has changed; it has become more oriented around communication, and less focused on document production.

Spreadsheets have become automated dashboards
Word documents have become living texts published on the web.
Powerpoint presentations have given way to live and recorded video talks.

The new office suite is made up of collaboration tools, not just new versions of solo-worker focused productivity apps. These types of apps are on the decline. The next 'word-processor' will probably look a lot more like Quip than like Word. Microsoft is refreshing its brand and cash cow boldly and correctly.

Ray Ozzie was ahead of the curve by bringing Groove into Microsoft and focusing on collaboration. Groove was just a little bit too wonky too early, and Ray did not have time for Microsoft style internal executive competition.

One aspect of the consumerization of enterprise technology has been the ability of workers to cherry-pick a set of the best products for their needs. This trend has created best of breed brand names with network effect businesses. These important companies rightly command premium valuations.  

Microsoft has not shied away from placing large, accurate bets on acquisitions at non-consensus price levels. This strategy is going to pay off in a positive way for Microsoft’s continued leadership in its core markets.

Microsoft knows that, going forward, enterprise buyers will not settle for inferior copies of excellent and widely used category-winner software like Skype andYammer.

In the productivity market, Microsoft is executing on the investment thesis that only a few winning technology companies really matter, and those are worth a significant premium to their also-ran competition. The antilog M&A strategy is buying runners-up and market failures with good technology and teams and banking on pushing those products out through the acquirer's bigger channel. With app switching costs low, and network effects available to category winner collaboration apps, there is no reason to expect that losing collaboration products will suddenly outcompete winners due to a change in ownership.

This is the one strategy that a large, cash-generating incumbent can employ to avoid business disruption by new innovative startups: Buy the category-winning product and brands.

Pebble Watchface Tutorial - Arabic Numbers

I just created my first Pebble Watchface. I have had my Pebble since February, but it has taken me a little bit of time to think of a good idea for a Pebble Watchface that only required swapping image assets from an example Watchface. The idea: Big Arabic - Arabic-script numbers substituted for Western-script numbers in the Big Time example Watchface. The whole project took me 30 minutes, and I love my new, custom Watchface. Big Arabic looks like this:

First, I created a new project called 'Big Arabic' in CloudPebble, the dead-simple hosted Pebble App IDE. When I created the new project I chose to use the 'Big Time' template available in CloudPebble:

Second, I looked through the files and file structure pre-populated from my template. This is a really simple Watchface. All of the numbers for the Big Time template are individual PNGs that take up a quarter of the Pebble screen (72x84px.)

Next, I found a large image on the web containing Arabic-script numbers 0-9. I wanted to find the absolute simplest way to cut this image up into individual 1/4-Pebble number images, and found this easy tutorial on Preview in OSX: http://mac.tutsplus.com/tutorials/app-training/preview-os-xs-secret-weapon-image-editor/ 

I used Preview to generate 10 72x84 Arabic number images, then I went back to my CloudPebble project and replaced all of the number image resources with my new Arabic-script images. After this I went out on the web and found an icon-sized image of Arabic script, resized it to be the same size as the example menu icon in my CloudPebble project using Preview and replaced the example image with the new icon.

Last, I opened big_time.c in CloudPebble for the first time and made two changes to the code. First, I changed the display name within PBL_APP_INFO from "Big Time" to "Big Arabic". Second, because the original Big Time Watchface is white numbers on black background, and my Arabic number images are black numbers on white background, I changed  window_set_background_color from GColorBlack to GColorWhite

After saving my changes in CloudPebble, I clicked over to the Compilation tab and ran a build. When the build completed, I got a short URL linking to my new Watchface, Big Arabic that i opened on my phone and loaded on to my Pebble.

Super-easy, fast and with barely any coding, I have a cool new custom Watchface. You can get it here yourself: http://builds.cloudpebble.net/4/2/42cc68d25c4c4a3bbe2ab3053dd05b1c/watchface.pbw

Next up, something more involved - I cannot wait for access to the accelerometer in the SDK.