From 2010 to 2011 I co-founded and ran a data marketing company called Qwerly, which appended metadata to email addresses and social media usernames. You'd ping our REST API with an email or a username, we'd do a bit of magic and serve back some JSON that listed social media profiles, location, photos, associated URIs. It wasn't so dissimilar to what a Rapportive or a Rapleaf did and it piggybacked off the available APIs, e.g. for finding friends by email addresses, as well as the rel="me" microformat to discover new websites and social media profiles.
The core bet that didn't really pan out was that the social web was getting more diverse and interesting and meta-profiles would have value. We figured this could be a cool consumer play, like a decentralized Friendfeed. But as we rolled this out, we saw the opposite was happening: 80% of our results were from Facebook.
And so we opened an API to see what people wanted to do with our results. And found that people really valued social data: to populate CRM systems, email lists, to make call center agents' conversations better, to target marketing. And so we pivoted to B2B.
Qwerly the B2B business was almost instantly profitable. Our biggest customer was Mailchimp, who used us to populate email lists with social profiles. But we also had political customers (an NDA prevents me from saying which campaigns). The volumes were impressive. For example, when we sold the company, the database had hundreds of millions of profiles. We never had to raise money. But we did get some inbound investor interest - from In-Q-Tel. Not surprising.
To their credit Facebook, while allowing us to do what we did, was keen to preserve privacy controls throughout the chain. So if a marketing manager somewhere had been blocked by his ex-wife, he shouldn't be able to see her information.
We prided ourselves on only serving "public" data, i.e. what you'd see searching Google or if you weren't logged into any one service. So that wasn't really a problem for us (though the constant re-OAuthing sucked). But of course the way we appended data relied on using private or frequently hidden APIs. We quickly realized this privacy stuff was harder than expected.
Well, they didn't. And I think most people in tech knew that. At least anyone that was in the "app" business knew that data was being traded quite openly. One way was via "acquisitions" of Facebook apps.
Today there are big vested interests in keeping Facebook the way it is. Businesses love the Facebook marketing machine and there seems to be very little demand elasticity to price increases in its advertising. Government, now posturing about the awfulness of it all, has always had a big interest in a more transparent populace and the means to control it. Heck, most people I know only know they kind of dislike Facebook, but they can't imagine their lives without it.
The fact that a WhatsApp co-founder is tweeting #deletefacebook is too little too late. WhatsApp was a company that said: we're never going to serve ads, we're going to show you that we can build a large social messaging app on subscriptions, we're going to do right by users. Well, that promise was worth less than $20 billion in case of these guys, so I have little sympathy here.
It is time. #deletefacebook— Brian Acton (@brianacton) March 20, 2018
I am not sure where we should go from here. I loved the internet without Facebook. I hate the monopoly on advertising and attention and M&A for social that is the Facebook vampire squid. But we seem to have done a poor job creating alternatives to Facebook. The application layer of the decentralized web seems years away. Mobile vertical communities can work but by definition they're more niche. The implicit messaging infrastructure of iMessage is cool, but Apple isn't good at building social. From Snapchat, Musically and Path to Meerkat and Periscope, Houseparty, and many others, long-term retention and engagement isn't very good on most social properties. And the ones where it's good or has potential to be good are acquired by Facebook.
Perhaps it's time to turn East to learn how to build new consumer platforms in the US and Europe. Anyway, here's some old slides from a presentation we did at a Techcrunch event. Knowing what you know now, it does sound more ominous than it did at the time...:
So you probably heard by now that an autonomous Uber killed a woman in Arizona. There was a human operator behind the wheel, but the car was in autonomous mode.
Of course we don't yet know what really happened (EDIT: preliminary police statement indicates likely not fault of self-driving car | UPDATE: the video is pretty damning - suggest huge problem for Uber).
The promise of autonomous mobility is fewer deaths. I'm very excited about self-driving cars for other reasons as well. I love the vision of living further outside of cities that are less full of cars, with fewer traffic jams and less wasted time commuting.
But what I don't understand, on the road to self-driving (sorry), is why we're letting proprietary tech take over such an important part of our lives yet again. Or, as a comment on Hacker News put it, "Why can a private, for-profit company "test" their systems on the public roads?"
Socialize the losses, privatize the profits? That doesn't seem right. And so, as with other software that has such enormous potential, as long as it's using public infrastructure we should probably insist that it be free software (as in freedom, not beer). You should be free to inspect, to modify and to share both the training algorithms and the machine learning models that result. At least as long as you're purchasing or leasing the vehicle. The same should probably be true for government licensing bodies above all.
Otherwise we'd have no idea what that software is meant to do. Which is crazy if you think about it.
Stop clicking on that 'bait!
But while you're here, I very much enjoyed this Q&A-style article about current knowledge in nutrition: "The Ultimate Conversation on Healthy Eating and Nutrition."
Maybe it's confirmation bias, but I do believe the best diet while building a startup is the one that cuts out all the things that actually _have_ nutrition labels. Also cheat days. You gotta have those.
On a more serious note, most founders I know don't self-care enough. Go work out. Eat healthy. Get more sleep. If you're feeling down, there's no shame in talking to a professional - be it a coach or a therapist. Both will make you a better founder.
I'm currently reading book #1 of Iain Bank's Culture series, Consider Phlebas. I bought the 25 anniversary box set, so will make my way through the three books. I'm much better versed in fantasy than science fiction, so if you have suggestions - please comment.
So far, Banks' work is delightfully weird and refreshingly written for what feels like deep genre. I also love the title, which is taken from TS Eliot's Waste Land:
IV. Death by Water
Consider Phlebas, indeed. Amazon announced in February that it has acquired the global television rights, making this a likely original series in a year or two.
Pär-Jörgen of Northzone and Fredrik of Creandum had great retrospective posts (here and here) on the Spotify story this week. I never saw Spotify at the Series A (I was at Atlas at the time - here's Fred Destin's Twitter thread reminiscing) and my current partners at Sunstone also passed on that round.
After Skype, ARM and Supercell, Spotify's debut (WSJ, paywall) is testament yet again to Europe's capability of building truly global, sustainable technology businesses. I am in awe of what the team there has built over the years - it just continues to be so much better than the offerings by Apple, Google, Amazon, Rdio, Napster and whoever else tried... (Jay-Z?).
I went back today to see what my first ever Spotify playlist was (I joined in 2009) and it's "Jazz and the 90s", a slightly histrionic, cheesy but ultimately catchy jazz cover album of various 90s hits. Yes, I'm showing my age here. Best enjoyed as background music. Here's the embed:
P.S. If there's one thing I could change about Spotify, it's their tendency to do zero-rating deals. Please, please support net neutrality and don't let yourself be instrumentalized that way.
The George Washington Bridge in Heavy Smog. View toward the New Jersey Side of the Hudson River, 1973 (Chester Higgins / EPA)
Detroit Lake the Dam, 09/1973 (David Falconer / EPA)
International Paper Company Mill at Jay on the Androscoggin River, 06/1973 (Charles Steinhacker / EPA)
1/ Everyone in consumer is trying to figure out how to play emerging CPG (<$15m in revenue). EVERYONE. Every public CPG company, every retailer, every consumer VC and PE firm, every meaningful public investor. EVERYONE.— Ryan Caldbeck (@ryan_caldbeck) March 14, 2018
Here is why.