A chat with Josh Schwarzapel, early web entrepreneur & fintech executive

Building consumer products, leading growth for Yahoo Mobile, macro shifts in financial services, price-value storytelling in fintech, and more

Nov 23, 2020

Sar : You were an early product manager at the photo-sharing startup Cooliris. During your time at Cooliris 2005-2010, you built out the first engineering team, led the main web plugin product, and and played a key role in the distribution partnership with Mozilla. Even though Cooliris was acquired by Yahoo in 2014, it feels like generations ago when you consider how different the world is now just 6 years later. What are your most cherished memories and hard-learned lessons from your time building the product in that era? Startups like Cooliris are a good reminder of how there are always lots of teams going after the same idea at any given time and how history gets written by one or two teams that make it.

Josh : Throwback time! What was crazy about that era is that tech wasn't nearly as hot as it is now. Most of my classmates wanted to go to Goldman Sachs rather than work in tech.

For context, we started Cooliris to build a better browser. Internet Explorer had ~90% market share, and we all knew it sucked. We started with a browser extension for media viewing that took off like a rocket, with over 30M downloads in just a couple of years. But we were unsuccessful in making the jump to a full fledged browser.

What we did amazingly well at Cooliris was hire. Cooliris alumni went on to found or become very senior executives at Instagram, Slack, Twitter, eightfold.ai, Filecoin, Facebook, Oracle, and a bunch more. My biggest takeaways are:

It's way easier to hire people when you have an ambitious vision.
Run your hiring process like a sales funnel.
Even the best people won't make you successful if your strategy is f*cked.

For #1: Building a browser was a hard technical problem, and a product that everyone on the internet needed. The ambition and swagger we had as a team was a big draw. That's why places like Tesla or Amazon can hire such an amazing team. There is serious ambition behind those projects.

For #2: We were SO aggressive at finding and recruiting the best people. We'd meet candidates over at Kleiner Perkins and walk them past John Doerr and Al Gore's desk. We'd pitch multiple times before we ever interviewed. We moved really fast once we were at the offer stage. We'd get to know the candidate and their spouse, and made sure they felt like (a) the opportunity was huge and (b) the team really wanted them. It's amazing to me how many big companies don't do this well. They rely on inbound, and there's very little selling that's done up front or at the end of the process. Managers outsource all coordination to recruiters, and the process drags on forever. If you look at hiring like a sales process, you'll completely change how you approach it.

For #3: Despite having an absurdly talented team, and over 30M downloads of our software, we couldn't find a viable path to a product and business that was successful in the long term. You see this pattern ALL of the time. Quibi, Color, Magic Leap, you name it….they all had amazingly talented teams and all of the funding in the world, and it doesn't matter. What's worse, it's always harder to pivot and maneuver once you have a large team and tons of funding. The best businesses I've seen keep their teams REALLY constrained until product/market fit is obvious. Instagram only had 13 employees when they were acquired - that to me is the ideal.

Sar : Folks at Yahoo at the time must really like you! You left Cooliris in 2010, co-founded OnTheAir in 2011 and got acquired by Yahoo once again just a year after the launch! Here’s what Josh Constine said in his coverage at the time in 2012 :

“The app lets people set up adhoc webinars that can be watched over live stream by large audiences. The real differentiator was that OnTheAir could serve as a sort of video call-in radio show for the Internet age. A host could take requests from viewers to come “onstage,” then broadcast a split screen chat with them.

However, with Airtime unable to gain traction and big players like Google investing heavily in the space, being successful while staying independent may have been tough for the team. An early acquisition bid before having to raise more money may have been the smartest move.”

Once again, you were working on a product well ahead of its time. What can you tell us about how you were thinking about what you were trying to do at the time? What was and wasn’t working? I was in High School back then and had no idea what TechCrunch was! You were building products during a very formative time in Silicon Valley history. Most of the social and mobile juggernauts of today were getting off the ground. Would love for you to walk us through what was happening in the SV zeitgeist at the time! In 2012, Facebook was in the middle of transition to mobile, Snap was less than two years old, and Instagram was bought by Facebook. Tumblr got bought by Yahoo just a year later.

Josh : When we were in undergrad, we were really lucky to get access to famous entrepreneurs we admired in college. We went out to dinner with Reid Hoffman, and had Vinod Khosla come in to do a case study with our class of 12. Getting to know these folks in intimate settings inspired a TON of us to enter tech and to try and start companies.

Our mission at On The Air was to try and recreate this magic, and make the most interesting people in the world more accessible. If you imagine a spectrum of intimacy with Twitter on one hand, and a private dinner on the other, we wanted to push as close to the "private dinner" end of the spectrum as we could.

Our product allowed celebrities and influencers to "go live" and talk with fans, who could go live on video and ask questions or hang out. Hangouts launched Hangouts On Air shortly after to do the same thing, as did Jeff Fluhr - the founder of StubHub - who launched Spreecast.

It turns out all of us had the wrong idea :). Specifically:

We were all overly focused on synchronous, 2-way live video. It turns out that most regular consumers are shy, don’t want to be seen on video, and are kinda boring to listen to! Asynchronous publishing was the killer behavior, with YouTube and Instagram taking most of the fan/follower media habit. Masterclass also captured an interesting niche here, as did Periscope. The point on Periscope leads to...
We were all desktop products. In 2011 when we started, it was still too early for mobile live video because of broadband speeds. It turns out that you really needed the notification channel to make a live video product work well - “XYZ is live now!” is a great push notification. Having a camera on your phone reduced the friction to publish as well. That allowed Periscope to find success a few years later.

The big lesson for me here was a repeat of a famous Jeff Bezos saying: Be rigid on the vision, but flexible on the details. Our vision was largely right. But we got the details wrong and it killed the product. The other lesson is the same as Cooliris: super smart, ambitious people often have the wrong hypothesis. It wasn’t just us… the Hangouts team and the Spreecast team whiffed too. Fortunately our company ended well with the team and investors doing well in our sale to Yahoo.

Sar : I think most people would still say we haven’t cracked synchronous, 2-way live video in the US. 1:many video streaming has thrived. And there are a lots of incarnations of internet radio products out there today.

You spent nearly 4 years at Yahoo on mobile growth and emerging products as a senior product leader post acquisition from 2012 to 2016. Yahoo has had an amazing track record in buying hot startups but has struggled to integrate and grow most of them. What can you tell us about the prevailing philosophy on product development and acquisitions at the time? Both Google and Facebook who also were on an acquisitions spree during this era.

Josh : My time at Yahoo was super fun. We were the second acquisition of the Marissa Mayer era, and we helped build the mobile team and business from zero. Our team scaled from 4 to 40, and we helped Yahoo Mobile grow from $30M in revenue to $1B in revenue in 3 years. More importantly, an insanely talented group of entrepreneurs all came into the company at around the same time. It really was a lot of fun, and I’m super proud of the work we did with the Mobile team.

I do think there was a fatal flaw in Yahoo's broader strategy though. In particular, there was no way we were going to beat Google at Search or Email. And there was no way we were going to compete with Facebook, Instagram, Netflix, and YouTube for entertainment. All of these are insanely difficult, expensive habits to win, and we were starting way from behind. We ultimately spent aggressively to try and catch up, but only made small inroads on market share while our profitability cratered. This led to a lot of activist shareholder activity, and the executive team was ultimately forced to sell the company.

In retrospect, the right (albeit painful) move would have been to cede search and mail to Google, and to lay off 80% of the company. We could have doubled revenue over night by outsourcing search and mail, and cut costs by 80%. We would have been throwing off billions of dollars in free cash flow every year, and the team would have bought 5-10 years to make a few big acquisitions or take a number of bets on categories that weren't so competitive. Hindsight is 20-20, but that would have been a savvier strategy in my opinion.

Sar : You’ve been working in fintech for almost 5 years now, across multiple products and customer types. What are some of the patterns you see?

Josh : First off, I think we’re seeing a massive, 20 year platform shift in financial services from paper and branches to software. And we’re only in the early innings, with incumbent banks still owning most of the customers and revenue. I suspect we’ll see several fintech companies value in the hundreds of billions of dollars in the decade to come.

What’s interesting is that this transition looks like a classic low end disruption in the Christensen sense of the definition. Almost ALL of the mega-winners in fintech are focusing on down-market customer segments that big banks have traditionally overcharged. Chime, Robinhood, Dave, Earnin, Acorns, Digit, Affirm, AfterPay, Shopify, Gusto, Kabbage…all of them started by aiming down-market of their legacy counterparts.

It’s worth noting that this isn’t because big banks are heartless, greedy sociopaths who only care about rich people. They simply have a higher cost structure per account - due to physical branches and labor intensive processes - that makes serving low income customers less profitable. This renders them unable to profitably target low-income customers with good products. It’s also worth noting that the converse is true… almost no fintech startups have been successful in serving up-market customers, and incumbent banks have largely defended this turf to date.

In terms of some finer points around what’s working, I’ve seen a few patterns emerging…

Pattern 1: the frequency / urgency paradigm is alive and well for fintech

Kamo Asatryan, one of the smartest consumer product people I know, shared a framework with me for consumer products that I continue to come back to. The framework suggests the best consumer products address use cases that are both frequent and urgent.

When it comes to lower income consumers, short term liquidity is the most urgent problem, and card spend is the most frequent use case. Lo and behold, the most valuable private fintech company (Chime) addresses both. They help with short term liquidity through SpotMe and early pay, and their customers spend on their card every day.

This pattern is playing out all over fintech, with higher frequency use cases winning out. AfterPay is bigger than Affirm, Robinhood is bigger than Wealthfront/Betterment, and I suspect Dave will blow past Sofi in the next few years.

Pattern 2: creative framing around price and value

Many breakout fintech products frame price and value in creative ways.

Take the small dollar lenders like Earnin, Brigit, and Dave. They’ll offer you a 0% APR loan for up to $100. It’s free if you wait for the ACH to clear, or you can get your funds instantly for $5. $5 to save a few days doesn’t feel outlandish, especially if you need that money now. However, if the life of that loan is 1 week, the $5 fee works out to a ~260% APR! It just doesn’t feel that way because it’s not marketed as such, and the fee is optional. This is even more pronounced with something like instant deposit on Venmo where you pay 1% to get your money ~2 days faster. It’s definitely worth debating the ethics of this, but the fees are very transparent and easy to understand, and consumers seem to really love these offerings without resentment.

More examples are everywhere…..Brex markets “3-7x rewards” on various categories, but we know they are covering those expenses with 250 bps of interchange or less. That feels better than “1% cash back” or whatever the weighted average rewards math works out to. Robinhood offers Stock and crypto trading for “free”, but we know that they’re making the money in other ways behind the scenes. Wealthfront gives you “$5,000 managed for free” if you refer a friend, which is really just $12.

The lesson to me is that the way you frame price and value are super important, even more so than the “actual” dollars and cents.

An anti-pattern: LOTS of distribution models are working

It’s common that only 1-2 distribution methods will work for a given category. In social, EVERYONE was focused on address book invitations and social embedding for years. VCs exacerbate the problem with truisms like “you can’t build a successful business with paid acquisition” or “B2B2C rarely works.”

Fintech has broken a lot of these mental models, and a wide array of distribution tactics have led to multi-billion dollar companies...