Sukhinder Singh Cassidy founded theBoardlist, a premium talent marketplace that helps diverse leaders get discovered for board and executive opportunities. A technology executive and entrepreneur, board member and investor, she has 25 years of experience founding and helping to scale companies, including Google, Amazon and Yodlee.
After listening to others pitch me a few different job opportunities while still at Google in 2008, it became clear to me that I would make a better decision if I could fully explore the larger landscape of new companies emerging in Silicon Valley.
I had spent the last several years focusing on Google’s business outside the U.S., and I honestly felt out of touch with the startup world. Beyond my goal of becoming a CEO of my own company, I had two other ambitions: I wanted to help build a great consumer service that would delight people (potentially in e-commerce) and I wanted to build further wealth for myself and my family.
To better evaluate my options, I made the decision to quit Google first and find a way to study the wider ecosystem of companies before choosing where to go. Resolved to give myself a “blank slate” before making a final choice, I left Google when I was three months pregnant and joined Accel Partners, a top Silicon Valley venture capital firm and an investor in my previous startup, in a temporary role as CEO-in-residence.
In the months that followed, I helped Accel evaluate investment opportunities across a wide variety of digital sectors, with a particular focus on e-commerce, taking the opportunity to study those companies I might join or think of starting from scratch.
On Thursday, August 19 at 2 p.m. PDT/5 p.m. EDT/9 p.m. UTC
Managing Editor Danny Crichton will interview Sukhinder Singh Cassidy, author of “Choose Possibility,” on Twitter Spaces.
One of Accel’s key partners, Theresia Gouw, helped me brainstorm, joining my cadre of professional priests. We had known one another for over a decade (I originally met her as a young founder at Yodlee) and were at similar stages of our careers, so I knew she could identify personally with my career quandaries. Like me, Theresia was pregnant with her next child and at a similar life stage — yet another commonality.
Image Credits: Sukhinder Singh Cassidy
While at Accel, I spent a disproportionate amount of time testing my macro thesis that online shopping was about to explode in new ways. I had seen the rise of e-tailers at Google (many of these companies, such as eBay and Amazon, were Google’s largest advertisers at the time), but many of the leading e-commerce sites like Amazon and Zappos still had a utilitarian feel to them.
Meanwhile, new fashion and décor e-commerce sites such as Rent the Runway, Gilt, Houzz, Wayfair and One Kings Lane were popping up everywhere and growing rapidly. These sites sought to tap into a more aspirational and entertainment-oriented kind of shopping experience and move it online.
Expert investors like Accel and others were funding them, and my own observations suggested that this area would yield another big wave of online consumer growth. These lifestyle categories of shopping also appealed to me personally; I was the target customer for many of them.
I started to work on an idea for a new e-commerce service, a luxury version of eBay, while listening to the pitches of every e-commerce company that was looking for funding and talking to several that needed early-stage CEOs. I continued to listen to non-e-commerce pitches as well, simply to give myself a point of reference for evaluating online shopping opportunities.
At Yodlee and Google, I had been lucky enough to work with incredibly smart and talented people who shared my values, and I wanted to do the same at my next venture.
I wanted to work with great investors, too, and fortunately I had the ability either to work with Accel-funded companies, start my own or leverage other investor relationships I’d developed. I spent time with multiple company founders to try to discern who they were as leaders, in addition to what they were working on.
By this point in my career, I had a pretty clear idea of my own superpowers and values, so I looked to find companies that could make the most of my unique gifts and whose founders or senior leaders had strengths complementary to mine.
Specifically, I hoped to join a company with a very strong engineering and product management culture that needed a CEO with strategy, vision, business development, fundraising and team-building expertise. Applying these criteria, I turned down several opportunities at companies whose founders had skill sets too similar to mine, reasoning that this overlap might lead to conflict if I ever became CEO.
Finally, I used my time at Accel to think long and hard about the risks I would take in becoming a startup CEO and whether I could afford to fail. My biggest risk by far was ego- and reputation-related. Mindful of how precarious early-stage startups are, I feared that I would leave a successful role as a global executive only to suffer a very large and visible failure. But the more I thought about this, I faced this ego risk head-on and concluded that my reputation as an executive from Google would hopefully be strong enough to survive one failure if it came to that.
The personal risks of taking on a startup CEO role felt different but not greater than those associated with my job at Google. While I knew that serving as a first-time CEO while having another newborn at home (my son Kieran) would be immensely stressful, I would likely benefit from no longer traveling around the world for days and weeks on end and working across multiple time zones, as I had previously.
Last, I evaluated the financial risks of potential moves. Although my startup equity would have uncertain value for a long time, I judged this a risk worth taking, given how excited I’d feel to have more impact and responsibility as CEO. While I lost a large financial package in choosing to leave Google and switching to a startup salary, I could pay the bills at home while digging into my savings only slightly. Under these conditions, I was prepared to make the leap.
In early 2010, almost a year after I left Google, I finally found the right opportunity and decided to join fashion technology startup Polyvore as its full-time CEO. A precursor to Pinterest, Polyvore was based on the idea that women could “clip” online images to create fashion and décor idea boards digitally that were instantly “shoppable.”
Millions of young women (including influencers) were already using the service and loved it. The founding team was led by a rock star engineer, Pasha Sadri, along with three other product and technology folks he recruited from the likes of Yahoo and Google.
Pasha was known for his intelligence, and we had connected informally over the years for coffee, each time having great discussions about business strategy. In fact, Polyvore twice before had tried to recruit me to become its CEO, once when I was at Google and again when I departed that company in 2008. Back then, I’d spent a productive afternoon with the founding team, helping them think through their business model. I also knew Peter Fenton, one of Silicon Valley’s most successful investors and a leading funder of the company. Peter was the one who first introduced me to Polyvore and who continued afterward to passively court me.
Having spent so much time exploring my options from multiple angles, I was now poised to make a great decision. I felt convinced that e-commerce was starting its next wave of growth, and felt excited to be part of it.
Within that vision, Polyvore was among the companies best positioned to succeed, and I knew I could contribute in significant ways to building a service that would delight millions. I was impressed with the strengths of Polyvore’s founder and investors and anticipated that I would be able to complement their efforts nicely. Recognizing that my success as a startup CEO hinged on my relationships with the founder and board, I had also invested time to get to know them.
Meanwhile, I had faced my fear demons, taking financial risk but negotiating my offer aggressively to account for downside scenarios I imagined, and coming to grips with my ego risk. With all this work in place, I finally jumped.
After managing a multibillion-dollar profit and loss and leading a 2,000-person team at Google, I became the newly minted CEO of a 10-person fashion startup in February 2010.
As we tee up the bigger choices in our careers, we all face critical moments of decision. No choice we make will be perfect, and all the frameworks in the world won’t eliminate risk entirely. But we don’t need perfection or freedom from risk. We just need to take the next step.
By choosing thoughtfully, using all the tools at our disposal to maximize our upside and anticipate our downside, we can grasp the opportunities available to us while equipping ourselves to handle whatever challenges reality throws our way.
Lucas is still out for a few more days, so I’m filling in for him again on this week’s Week In Review. To recap: I’m Greg. I started at TC back when deciding who to put in your MySpace Top 8 was a very serious matter and being able to summon an Uber with a button press made people think you were a wizard from the future.
Before we dive into the news you might’ve missed this week, a heads up! Brian Heater’s much-loved robotics recap is becoming a weekly newsletter and getting a fancy new name: Actuator. The official launch date is still kind of up in the air, but you can already sign up for it right here.
And now, the news you might’ve missed this week.
The Big Thing
Whether or not you’ve ever dabbled with Unity, you’ve almost certainly interacted with something built with Unity. It’s the 2D/3D engine that powers so, so many of the video games out there, regardless of what console or platform we’re talking about. Studios use it to make animated movies. Automakers use it to help design cars.
This week Unity announced its intent to make a big acquisition — its biggest to date, in fact — with the $320M purchase of Parsec. And while one company buying another is hardly rare news around these parts, a Unity Senior Vice President suggests that it could be the beginning of a bigger shift for the co.
Parsec started its life as a way to stream games from your powerful PC to your less powerful devices — or to your far away friends, allowing for long-distance multiplayer in games that otherwise didn’t support it. And it still works for that!
When the pandemic booted everyone out of the office, though, the company found that many of the features it had built for playing games remotely (like low latency streaming, support for input devices beyond keyboard/mouse, etc.) were also super important to the folks building games remotely. They embraced that newfound audience hard, quickly rolling out plans and features just for creative teams. The shift worked out well for them; the company went from raising $25M to being acquired for nearly 13x that in less than a year.
According to Unity SVP Marc Whitten, this is the beginning of Unity diving deeper into the cloud.
“I think you’re gonna see that Parsec is a great foundational block for a broad sort of cloud ambition that we have as a company,” he says. “You’re going to see a lot more from us in that particular regards.”
Even only considering what Parsec immediately brings to the table, there’s multiple potential paths to the cloud here. They could use Parsec to help Unity developers more easily add multiplayer to their games; they could use it to build a Stadia/Amazon Luna-style game streaming service that showcases Unity-powered games sans downloads; they could use it to offer up beefier hardware-in-the-cloud rentals to help smaller studios iterate more quickly or test on a wider range of devices.
Oh, and if you’re one of those aforementioned people using Parsec to game with friends from afar, don’t stress: The company says its free app isn’t going anywhere.
The book written about the evolution of Fortnite will be a wild one. It began its life as a moderately popular tower defense game. When it shifted gears into a free-to-play Battle Royale game, it exploded into one of the biggest games the world has ever seen. Now it’s a remarkable example of how a game can be a place, and what can be when a developer has absolute control over their game engine and might-as-well-be-endless money to throw into content creation. Example #2,138,413: this in-game Ariana Grande “concert” in which players dogfight a demon, ride rainbows, and dance alongside a skyscraper-sized Ariana. The YouTube replay alone has already been seen by millions.
Twitter overhauled its website this week… and, as it tends to go when you change the look/feel of a popular thing, there was some user backlash. This time it went beyond the usual “I don’t like the radius of the rounded corners” complaints, though, with some users complaining that the new font Twitter chose gave them headaches.
Last week Apple announced that it would be rolling out a set of features meant to keep kids safe: One alerting parents if a minor is sending or receiving explicit images over iMessage, and another that would automatically compare generated hashes of iCloud Photos to detect and flag users who were storing known child abuse images. While protecting kids is undeniably and universally a good thing, security researchers have raised concerns about the potential for misuse by governments to scan for things beyond abuse imagery. It’s said to have caused a bit of a dustup within Apple, with “more than 800 messages” of back-and-forth posted by employees on the company’s internal Slack. The company has spent the last few days trying to clarify how the features will work, with Apple’s Craig Federighi admitting that it “got jumbled pretty badly in terms of how things were understood”. Read our interview with Apple’s head of Privacy here.
Boston Dynamics’ Spot is no longer the only creepy dystopian robot dog in town. This week Xiaomi announced CyberDog, a four-legged bot meant to flex the company’s robo knowledge and serve as a platform for developers to build upon. Xiaomi says they’ll be selling them for around $1,500, with a catch: they’re only making a thousand of these, initially, and they’re only selling them to select “Xiaomi fans, engineers, and robotics enthusiasts.”
Lowercarbon Capital, the climate-focused fund run by Chris and Crystal Sacca, raised $800M to pour into companies that are taking on the climate crisis. “It turns out that raising for a climate fund in the context of an unprecedented heatwave and from behind the thick clouds of fire smoke probably didn’t hurt,” writes Sacca. Can tech save the planet? TBD. But continuing to do nothing definitely won’t.
If your phone was shouting about a test of the National Wireless Emergency System earlier this week, don’t panic: It was, in fact, just a test. Didn’t get it? Don’t panic about that, either — the test system is opt-in. If there’s an actual test, everyone will get it. Hopefully.
Is your company trying to drag everyone back into the office sooner than you’d like? Maybe send’em this article from Insightly COO Karl Laughton, outlining some of the findings and data-driven upsides the company has seen since going remote.
You’ve found the perfect job candidate and want to make an offer, but there’s a catch: they need an EB-2 green card, and another company has already started the sponsoring process. Can you make it work? It depends. In the latest edition of Dear Sophie, Immigration attorney Sophie Alcorn breaks down the EB-2 process and potential snags.
Dr. Phil Wagner is the founder and CEO of Sparta Science. After receiving his medical degree from USC, he was frustrated by the lack of evidence-based approaches to performance and injury prevention within sports, the military and occupational health, inspiring him to create Sparta Science.
Political debate at work was not encouraged when I was training to be a doctor at the LAC+USC Medical Center in the early 2000s.
On the 13th-floor jail ward, we had a professional duty to care for drunk drivers and thieves just like any other patient and leave any opinions about criminal justice policy for their appropriate venues.
Medicine is not unique in this respect — we’re all better off when lawyers, soldiers and other public service providers place their duty to society over individual opinions.
Technology companies often aspire to fill similarly critical roles in society, but few have internalized the separation between professional duty and personal opinion. I have seen this firsthand as the founder of a tech company that serves a wide range of organizations, including amateur to collegiate and professional sports, occupational health and a growing roster of military commands.
Many founders fail to explore DOD opportunities because they do not want to be seen as engaging in the business of war.
During the last presidential administration, a handful of colleagues questioned whether serving the military was consistent with our mission of helping the world move better. Over the past few years, this stigma toward military work has roiled some of the largest companies in Silicon Valley, sometimes leading to contract cancelations, non-renewal pledges and a noticeable chilling effect toward work involving the United States military.
Partnerships between technology companies and the military are nothing new, but rarely have they attracted as much controversy as they do today. These partnerships were the norm throughout the 20th century, yielding war-winning technologies — like microwave radar, GPS and ARPANET — that pulled double-duty in peacetime as the building blocks of our modern connected world.
Military contracts have been traditionally viewed in Silicon Valley as a win-win — for the nation’s military superiority and for a company’s bottom line. Moonshot projects backed by the federal government’s financial resources also represented some of the most interesting workarounds for product-minded technologists.
That relationship has been knocked off its bearings over the past several years, with employees at Microsoft, Google and Amazon, among other companies, seeking to distance themselves from all federal projects due to the revulsion of the previous administration’s policies. But with new leadership in Washington, companies and tech workers need to determine if the stigma against military work will become permanently ingrained or limited to one chapter in an evolving relationship.
Before looking forward, one common misconception is worth correcting from the previous administration about the tension between employees and the military. Recent research challenges the notion that anti-military views are universal among the tech workforce.
In a survey conducted between late 2019 and early 2020, Georgetown University’s Center for Security and Emerging Technology found that less than a quarter of AI professionals view Pentagon work in a negative light, and 78% consider it a positive or neutral.
Companies that are open to pursuing opportunities with the Department of Defense should consider several advantages and differences between commercial and government clients.
Federal contracts are generally distinguished by large dollar amounts, low profit margins and long periods of performance. This can appeal to VC-backed companies that are valued based on revenue, and the unique structure of government contracts brings a welcome complement to the lucrative but highly volatile work in B2B and B2C markets. Blending the two extremes produces a stronger whole, not unlike mutual funds that balance stocks and bonds.
Many founders fail to explore DOD opportunities because they do not want to be seen as engaging in the business of war. We encountered a version of this at Sparta Science with colleagues who conflated our work to support federal employees with a full endorsement of all government policy.
Reality is more nuanced. The DOD has an annual budget of more than half a trillion dollars and a workforce of 2.8 million. Only a portion of these individuals are directly engaged in warfighting, and they rely on great numbers of administrators and knowledge professionals to accomplish each mission.
The DOD has approximately 1.3 million active contracts at any given moment, spanning fields as diverse as healthcare, apparel, logistics and software licensing. The military is rightly described as a cross-section of the United States, and supporting those who serve is a Silicon Valley tradition, good business practice and the right thing to do.
On this edition of the Gillmor Gang, Brent Leary shows off his new wireless adaptor for his live streaming studio. The result is a captivating view of his console as he switches between closeups and incoming feed from the rest of the Gang, all captured in a widescreen cinematic view. The underlying message is that live realtime video production has become accessible to virtually anyone as streaming becomes ubiquitous at the so-called citizen level.
Trailblazers like Brent and his CRMPlayaz partner in crime Paul Greenberg have been way out on the bleeding edge of this stuff; now we’re seeing something similar to what’s going on in the creator boom. Newsletters are becoming a baked in feature of the major social platforms, as is live audio streaming a la Clubhouse and Twitter Spaces. This week, Salesforce announced its Salesforce+ streaming network, and celebrated its completion of the Slack acquisition with several new enterprise spins on live audio (Huddles) and cross-company collaboration via Slack Connect.
Roll this up with the first wave of work from anywhere efforts to get back to school and the office, streaming as a service may be a key feature of the ongoing hybrid approach to fighting off the pandemic. The political struggle with vaccinations and masking seems destined for the long haul. How the tech community responds should be a more hopeful sign of progress.At the professional level, Disney and Scarlet Johannsen are trading lawsuit threats as month-old contracts are ripped up. Newsletter deals are chasing a dwindling population of hit authors as the New York Times puts most of their star-driven publications behind the paywall. The more things transform, the more familiar they seem.
Even the Gang newsletter sports a link to Om Malik’s post on Nam June Paik, the experimental video pioneer. I was his TA at California Institute of the Arts, and “borrowed” his associate Abe’s 3 Sony black and white portapacks to film a Firesign Theatre writing session. Civilization ho.
Thanks for reading Startups Weekly. Want the weekly digest in your inbox every Saturday? Sign up here.
This past week, edtech entrepreneurs, investors and analysts congregated at ASU+GSV, a yearly global edtech conference, to reflect on the sector’s newfound spotlight after the massive jolt of COVID-19. Beyond masked excitement to finally meet Twitter friends in real life, signs of bullishness were everywhere.
My takeaway from the entire week, though, was more than edtech is booming (which, it is). Instead, I felt like the general sentiment at the conference — even with polite disagreement — was that the sector getting spotlighted means it finally has the buy-in to go everywhere.
In other words, edtech is at a point where it doesn’t need to just rely on itself to accomplish its goals. It can operate outside of a silo, which feels like the needed follow-up to the sector’s 2020.
For example, if a platform brings fun UX to instruction online but doesn’t take into account how that move impacts childcare, mental health and the digital divide, the impact of its savvy solution will be immediately limited. Consolidation, which will continue to play out due to freshly minted unicorns, won’t just be edtech scooping up edtech, but you may see companies begin to launch products that take into account the full human experience.
BetterUp, a reskilling and coaching platform for employees before and beyond the C-suite, is signaling that it’s already happening. The edtech company announced that it is diving more into behavioral health with new products.
Operating beyond edtech insiders is the difference between creating products that reinforce the status quo and inventing ones that question the status quo on its head. Fiveable, a virtual space for high schoolers to study and express themselves, has turned dozens of its users into interns. The feedback loop there is brutal — high schoolers are harsh — but means that the people making decisions for them finally aren’t adults talking to adults.
Of course, the lack of training wheels means that it’s easier for startups to go rogue. As the pandemic unevenly plays out, remote school may become the normal once again. Companies have to be massively focused, and humble, about their reach. Reflection will be important in what distinguishes a Course Hero from a Codecademy to a Coursera — and when it makes sense to leave their own lane.
It was a refreshing, surreal week talking to the people behind the dollars and ideas of our future educational landscape. The jolt of the pandemic highlighted the inequities and work left to be done. Now, the spotlight will be part cheerleader, part accountability coach in helping edtech reflect its way to a better end product.
The relevancy of venture capital
Image Credits: Bryce Durbin / TechCrunch
As nontraditional investors get into private startups, a growing debate in Silicon Valley is if traditional venture capitalists can evolve into the new landscape. When Tigers eat your lunch, where do you look for competitive advantage and relevancy?
Here’s what to know: While regulations matter, it’s ironic that Uber only had the chance to become iconic because it ignored regulations during its early launch days. Disruption has a way of ignoring the rules.
You know what no one talks about? Maps. The medium is a powerful tool for consumers and companies to visually express space and relevancy. At the same time, the complexity of maps — from the curvature of land to how space takes up space — has limited how easy it is to just spin one up.
Here’s what to know: Felt, which just left stealth this week, wants to make maps mainstream. Its mapping software has raised $4.5 million to date and is a case study in how climate change can bring new energy to old products.
Welcome back to This Week in Apps, the weekly TechCrunch series that recaps the latest in mobile OS news, mobile applications and the overall app economy.
The app industry continues to grow, with a record 218 billion downloads and $143 billion in global consumer spend in 2020. Consumers last year also spent 3.5 trillion minutes using apps on Android devices alone. And in the U.S., app usage surged ahead of the time spent watching live TV. Currently, the average American watches 3.7 hours of live TV per day, but now spends four hours per day on their mobile devices.
Apps aren’t just a way to pass idle hours — they’re also a big business. In 2019, mobile-first companies had a combined $544 billion valuation, 6.5x higher than those without a mobile focus. In 2020, investors poured $73 billion in capital into mobile companies — a figure that’s up 27% year-over-year.
This Week in Apps offers a way to keep up with this fast-moving industry in one place with the latest from the world of apps, including news, updates, startup fundings, mergers and acquisitions, and suggestions about new apps and games to try, too.
A new Senate bill could put an end to app stores’ dominance
Image Credits: TechCrunch
A bipartisan group of three U.S. senators — Richard Blumenthal (D-CT), Amy Klobuchar (D-MN) and Marsha Blackburn (R-TN) — introduced a new piece of legislation called the Open Markets Act, could change the way mobile software is distributed. The bill would give developers the right to tell their customers about lower prices outside the app stores (without fear of punishment), and permit alternative payment mechanisms, sideloading and third-party app stores where developers could avoid platform fees. It would also bar platform makers like Apple and Google from using non-public information they collect via app stores to build out competing products, or rank those products more favorably.
The bill is being applauded by Apple critics, including the Coalition for App Fairness and its members, Epic Games, Spotify, Tile and others, who are now urging Congress to swiftly pass the legislation to level the playing field.
As regulatory pressure on platform makers has intensified, the companies looked for ways to better cater to smaller developers with drops in commission rates, as well as increased privacy and security measures — the latter which could help boost their arguments that the app store model is favorable to consumer interests.
Such a bill is a notable first step toward some sort of market changes, but it’s still too early to know if or when the bill will gain traction, much less be passed into law.
Tech giants Google, YouTube and TikTok follow Instagram with increased protections for minors
Google and YouTube (as well as TikTok) this week rolled out a series of changes to their products and services to increase the privacy and security of accounts belonging to teenaged users under the age of 18. The specific changes vary a bit from service to service, but are largely focused on making younger people’s accounts more private by default, ensuring they’re making an intentional choice when shifting accounts or content to become public, and limiting to what extent advertisers can target them. TikTok went a bit further to restrict push notifications after “bedtime” hours for its teen users, while YouTube chose to turn on its “take a break” and “bedtime” reminders by default instead.
Image Credits: TikTok
The changes follow similar moves announced just weeks ago by Instagram, and follow increased pressure from the U.S. Congress to do more to protect younger users from the harmful impacts of using technology.
One piece of legislation, which tech companies may be trying to get ahead of, is an update to COPPA that would expand some protections to children under the age of 18, instead of just under 13. What’s missing from all these initiatives, however, is any plan to more strictly verify children’s ages on an app. Since many kids already know to lie about their birth year at sign-up, it’s unclear how effective these measures will be in the long term.
Samsung this week hosted its Unpacked event, where it debuted the company’s latest mobile products. This time, the smartphone maker showed off a new crop of foldables, including the Galaxy Z Fold 3 and Galaxy Z Flip 3 (clamshell, woo!), which will give app makers even more device styles and formats to consider when designing their apps. Well, if these foldables ever gain market traction that is, instead of existing in consumers’ minds as a gimmick. (For what it’s worth, Microsoft hopped on the bandwagon.)
Samsung also introduced a new smartwatch powered by Google’s WearOS (if you can’t beat ’em…), the Galaxy Watch 4, and entry-level wireless earbuds, the Galaxy Buds 2.
Weekly News
Platforms: Apple
Apple released the fifth betas of iOS 15 and iPadOS 15 to developers, which offer some more minor tweaks and stability improvements as the platforms head toward a fall release. New additions include an updated weather icon, shading around the tab interface in Safari on iPad, an option to use larger icons on iPad, a new warning pop-up that reminds you the iPhone is still findable when off, new splash screens for Apple’s apps, the integration of TestFlight info in the App Store and more.
Apple released a new developer tool that allows app makers to test how their app behaves when the device is connected to 5G instead of Wi-Fi. The tool is necessary because iOS 15/iPadOS 15 devices can automatically prioritize 5G over Wi-Fi when the latter’s performance is slow.
Apple settled a 2019 lawsuit with Corellium, a company that builds virtual iOS devices used by security researchers. Apple had said Corellium was infringing on its copyright, selling its product indiscriminately and compromising platform security. A judge dismissed Apple’s claims as “puzzling,” noting Corellium established “fair use.” The settlement terms were not disclosed.
Apple’s Find My app in iOS 15 will use Bluetooth technology to precisely locate AirPods (Pro and Max) devices, and will tie AirPods to users’ Apple ID.
Platforms: Google
Image Credits: Google
Google launched Android 12 beta 4, whose biggest new feature is that the platform has now reached stability. Developers can now test apps before the public release, without having to worry about future breaking changes. Android 12 offers a big redesign, with a more personalized “Material You” design language and increased privacy protections.
Google banned the location data firm SafeGraph, funded by a former head of Saudi intelligence, which was paying developers to include their data collection tools in their apps so they could resell the data to other companies or government agencies. Any apps working with SafeGraph will have to remove the code.
E-commerce
DoorDash recently held talks to buy Instacart, according to a report from The Information. The $40 billion-$50 billion deal would have combined two top food delivery apps — one for restaurants, the other for groceries — but talks fell through.
Weedmaps added in-app cannabis purchasing for iPhone users. Thanks to looser App Store restrictions, Weedmaps users can now browse, select and purchase cannabis and have it delivered or set for pickup directly within the app.
Instagram is testing ads in its Shop tab, which allow brands to feature either an image or image carousel. The ads will launch with an auction-based model and will only appear on mobile devices.
Augmented Reality
Snap hired a Facebook AR executive, Joe Darko, The Information reported. The new AR leader, who previously launched the Spark AR Partner Network at Facebook, will now oversee Snap’s AR Developer Relations.
Fintech
Image Credits: Venmo
Venmo announced it would allow its credit card holders to automatically buy cryptocurrency with their card’s cashback, through a new feature called Cash Back to Crypto. Cardholders can select between Bitcoin, Ethereum, Litecoin and Bitcoin Cash, which will be purchased monthly with no transaction fees.
Social
Some Snap creators have left for other platforms as the company’s creator bonuses dried up, CNBC found. Snap had been paying $1 million per day in prize money for creators posting to its TikTok competitor, Spotlight. Now, it’s paying “millions” per month — and creators are chasing bigger bonuses elsewhere.
Instagram’s TikTok competitor, Reels, added a new feature that allows users to search for audio to include in their short-form videos, and the pages for the tracks will show the other videos that used them — like TikTok offers.
Instagram rolled out new anti-abuse features after high-profile incidents of racism took place on its platform following the Euro 2020 final, where angry fans attacked players. New tools include Limits, which lets you restrict certain groups from DM’ing and commenting for a period of time; plus an expansion of Hidden Words to include new strings of emoji; and a more aggressive “Hide More Comments” feature.
Instagram took down a website, LikeUp.Me, selling fake likes and followers. The site, which was served a C&D from Facebook, had earned around $100K in the past year.
Reddit is rolling out a TikTok-like video feed button on its iOS app. The feature has reached “most” iOS users and drops them into a full screen experience where users can upvote, downvote, comment on, gift an award or share the video. You can also swipe up to see more videos, also like TikTok. Surprisingly, the company claims its acquired Dubsmash IP was not a part of this project.
Messaging
Image Credits: WhatsApp
WhatsApp will gain the ability to transfer chat history between mobile operating systems. The feature is coming to Samsung devices first, followed by a broader Android rollout, then iOS. Samsung customers can use the company’s transfer tool, Smart Switch, which already copies other personal data between devices, to also now move their encrypted WhatsApp chat history, including voice notes, photos and conversations.
Google is pushing mobile Hangouts users to switch to Google Chat through an in-app message that reminds users that Hangouts will be discontinued. Disgruntled users took to Google Play to express their anger with dozens of one-star reviews. How’s that mess of a messaging app strategy looking now, Google?
Messenger delves into the design of its recently launched “Soundmojis,”which pair an emoji and sound together to create a new form of expression that would be universally understood and surprising.
Facebook is bringing end-to-end encryption to Messenger calls, noting that E2E was the industry standard and what people now expect. The company said also it would begin testing E2E for group chats and calls in Messenger and Instagram DMs.
Photos
Apple’s next iPhones will reportedly allow users to take “video portraits,” Bloomberg reported. Other additions will offer the ability to record video in the higher-quality format called ProRes and a new filters system to improve the look of photos.
As Instagram’s photography community is getting increasingly frustrated by the app’s shift to video, a new app called Glass launched into beta to serve photographers of all sizes. As reported by Om Malik, who has been testing the app for nearly six months, subscription-based Glass is beautiful and fresh, and reminiscent of early Instagram, with support for comments and followers, but differentiates itself by not chasing clout though public likes.
Image Credits: Glass
Dating
Facebook Dating is gaining an “audio chat” feature, which will allow matches to have voice chats to get to know one another. It’s also adding a Lucky Pick, to suggest daters outside someone’s typical preferences, and Match Anywhere, which allows users to consider locations outside their current city.
A Match beta test is targeting users’ most common dating app complaints, like too much swiping and ghosting. Now, Match will offer weekly “Matched by Us” recommendations and will prompt users to unmatch or respond, instead of leaving conversations hanging. The company also hinted that it may roll out a human-led matchmaking feature in the future, as well.
Tinder’s interactive feature, “Swipe Night,” is coming back after a 20 million user turnout from its “season 1.” The new version won’t be a choose-your-own-adventure, but rather a “Gen Z whodunit,” the company said, and will use the quick chat feature that allows users to chat without having first matched.
Streaming & Entertainment
HBO Max added free episodes to its platform, including its app for mobile devices. Users can sample 13 episodes from top shows and originals without paying, including “Euphoria,” “Game of Thrones,” “Lovecraft Country,” “The Flight Attendant,” “Veneno,” “Warrior” and more, as well as browse the catalog to see what else is offered with a paid subscription.
A Spotify representative told users in the company’s forum that its work to add support for the nearly 3-year-old AirPlay 2 technology was being abandoned for the time being, citing “driver compatibility issues.” The post gained a lot of attention from disgruntled users and press, leading Spotify to clarify that it “will support AirPlay 2,” without offering a time frame. The company, a staunch Apple critic, has been hesitant to support other Apple products, including HomePod speakers, where native support isn’t available.
YouTube’s Android app is trying out a new gesture that will allow users to navigate its video “Chapters” by double-tapping with two fingers.
A new U.S. streaming report by Penthera found that 71% of viewers stream video on 2 devices per day, on average, including a connected TV device and mobile. And 80% said they watch videos at home. But now 92% (up from 88% last year) now say re-buffering is the biggest problem they face while streaming.
Gaming
Image Credits: App Annie
App Annie released its 2021 gaming report, which estimates mobile gaming will reach $120 billion in consumer spending by year-end, or 3.1x more than consoles. Other highlights include:
4 of the top 10 most downloaded subgenres across all games are in the hypercasual category.
“Happy Glass” is the fastest ever hypercasual game to break 100 million downloads.
The pandemic pushed gaming to new levels. Weekly game downloads topped 1 billion in March 2020, and have stayed there ever since.
In H1 2021, there were over 810 games surpassing $1 million in consumer spending per month, up 25% from 2019.
In H1 2021, per week there were over 1 billion downloads, $1.7 billion spent and 5 billion hours spent on mobile games globally.
The U.S. topped the mobile games market by App Store consumer spending.
AppLovin topped the charts for worldwide downloads, while Tencent dominated consumer spend.
U.S. mobile game usage skews female (64% of gamers are female). That’s not the same in other markets, where more mobile gamers are men — like Japan (56% male) and South Korea (53% male).
U.S. mobile tabletop game spending rose by 40% to $704 million over the past 12 months, a Sensor Tower report found. The top grossing game was Solitaire Grand Harvest from Supertreat, followed by Solitaire TriPeaks from GSN, then Yahtzee with Buddies Dice from Scopely. Downloads, however, declined by 12% YoY, with 202.7 million installs over the past 12 months, versus 230.7 million in the year-ago period.
Epic Games CEO Tim Sweeney discovered a surprise in a set of recently unsealed court documents in the Epic Games v. Google antitrust case. He learned that Google had once mulled acquiring his company at some point — which Sweeney said was related to Epic’s decision to compete with Google Play. The documents also refer to the “frankly abysmal” sideloading experience on Android.
Food
Google-owned navigation app Waze announced a partnership with Too Good To Go, an organization that works with small businesses and organizations to decrease food waste. The partnership will highlight independent restaurants and grocery stores in select U.S. cities that are taking steps to reduce food waste by showcasing them on the Waze map. These businesses also sell “surprise” bags of food that contain three times more food than the cost of the bag at the end of the day. The food is perfectly good, but can’t be sold the second day, so would otherwise be thrown out.
OpenTable’s app added a new Direct Messaging feature that lets diners and restaurants communicate directly after a reservation is made, instead of having to place a phone call. The feature can be used to clarify a diner’s requests, or other changes, or even message after the reservation has ended in case of items that were left behind, or other needs.
Image Credits: OpenTable
Adtech
Tapjoy launched MobileVoice, a market research solution for surveying mobile-first consumers where researchers bid for each response. Higher bids will give users more virtual currency for their game, which motivates consumers to share their opinions.
Government & Policy
WhatsApp, Facebook, Instagram, Messenger and Twitter were restricted in Zambia amidst ongoing general elections on Thursday, polling day, through Sunday, when votes counts are expected to have ended.
Facebook’s acquisition of GIF database Giphy has come under fire from U.K. regulator, the Competition and Markets Authority, which announced a preliminary finding that the deal “will harm competition.”
India’s government says Twitter is now in compliance with the country’s new IT laws, which required the company to appoint a chief compliance officer, a nodal contact person and a resident grievance officer in the country.
Security & Privacy
Recommended Reading: TechCrunch Editor-in-Chief Matthew Panzarino interviewed Apple Privacy head Erik Neuenschwander about the company’s plans to detect CSAM and Apple’s new Messages app safety features. Apple’s announcement stirred controversy in the security community because of how the company is implementing the technology, which some have argued could leave the door open for other governments or agencies to compromise for their own ends. Neuenschwander explains the system’s protections that make it less useful for doing so — meaning that its technology itself, and not just Apple’s word, could prevent this type of abuse. And in terms of user privacy, there is an opt-out — you just turn off iCloud Photos.
Privacy-oriented search app Xayn raised $12 million in Series A funding led by Japanese investors Global Brain and KDDI (a Japanese telecommunications operator) for its app that fuses together search, a discovery feed and a mobile browser. The app will focus on Asian markets and Europe.
Social banking app Kroo raised $24.5 million in Series A funding led by Rudy Karsan, a high-net-worth tech entrepreneur and founder of Karlani Capital. The London-based fintech offers a prepaid card service but is moving toward offering expanded banking services in its mobile app.
Pokémon GO developer Niantic acquired the iPhone and iPad app Scaniverse for an undisclosed sum. The Scaniverse app allows users to scan objects and environments into high-res 3D, and will remain live on the App Store, and the founder will join Niantic’s AR team.
Car ownership “super app” Jerry raised $75 million in Series C funding led by existing backer Goodwater Capital, valuing its business at $450 million. The app uses automation to give consumers customized quotes from more than 45 insurance carriers, but is expanding into areas like financing, repair, warranties, parking, maintenance and more.
Mobile field service startup Youreka Labs raised $8.5 million in Series A funding co-led by Boulder Ventures and Grotech Ventures. The company simplifies development of mobile service applications with a no-code authoring studio and one-click deployment to Apple, Android and Windows.
Reddit confirmed it has raised $410 million of a planned $700 million Series F funding round, led by Fidelity, valuing the business at $10 billion. The funds will be used to further build out community and advertising efforts, as well as increase headcount.
U.S. grocery delivery service Gopuff acquired U.K. competitor Dija, which was only eight months old, to expand into Europe. Gopuff had previously acquired a similar startup, Fancy, just three months ago.
India’s VerSe Innovation, makers of Dailyhunt and Josh apps, raised over $450 millionin a Series I funding round led by Siguler Guff, Baillie Gifford, affiliates of Carlyle Asia Partners Growth II and others. The company’s new valuation is now “nearing $3 billion.” Dailyhunt now has over 300 million MAUs and Josh has 115 million MAUs.
Social calendar app Saturn raised $35 million led by General Catalyst, Insight Partners and Coatue, bringing its total raise to date to $44 million. The app allows high school students to manage their schedule, track assignments, chat with friends and more across web and mobile devices.
Fintech Robinhood acquired Say Technologies, a company offering a communications platform that allows smaller shareholders to pose questions to companies in which they invest. The $140 million all-cash deal is Robinhood’s first major purchase since going public in late July.
Medal.tv, a short-form video clipping service and social network for gamers, entered the livestreaming market with the acquisition of Rawa.tv, a Twitch rival based in Dubai. Deal terms were not revealed, but the deal was in the seven figures.
Turkey’s Trendyol, an e-commerce website and app serving over 30 million shoppers, raised $1.5 billion in a round that valued the company at $16.5 billion. General Atlantic, SoftBank Vision Fund 2, Princeville Capital and sovereign wealth funds, ADQ (UAE) and Qatar Investment Authority, co-led.
Argentine fintech Ualá raised $350 million in Series D funding, valuing its business at $2.45 billion. The company offers a Mastercard and app, where users can access bill pay solutions, investment products, personal loans, BNPL installments and insurance. To date, the startup has issued more than 3.5 million cards in Mexico and Argentina.
Fintech Chime Financial raised $750 million in a round that values the business at $25 billion, ahead of a planned IPO next year. The round was led by new investor Sequoia Capital Global Equities. Chime today offers credit cards and no-fees banking services through a mobile app.
Mexico’s Orchata, a mobile app for getting groceries delivered via micro-fulfillment centers, raised $4 million in seed funding from investors including Y Combinator, JAM Fund, FJ Labs, Venture Friends, Investo and Foundation Capital, and angel investors Ross Lipson, Mike Hennessey, Brian Requarth and Javier Mata.
Public Markets
Krafton, the South Korean maker of PUBG, closed 9% down on its first day of trading on Tuesday after first debuting at $432 per share. Analysts said the company tried to go out with a valuation that was too high. At closing, the valuation was $19.32 billion. To date, PUBG Mobile has generated $6.3 billion in player spending across the app stores.
Crypto app Coinbase’s stock jumped 7% on Wednesday after better-than-expected earnings, where the company reported $1.6 billion in net profit for the quarter (earnings per share of $3.45), beating analyst estimates. Coinbase trading volumes were also up 38% to $462 billion in Q2.
TikTok owner ByteDance is considering a Hong Kong IPO, The FT reported. The Beijing-based tech company may list in either Q4 2021 or early 2022. As of its last fundraise of $5 billion in December 2020, the company was valued at $180 billion.
Mobile marketer and game provider AppLovin’s stock jumped 4.2% in after-hours trading Wednesday after the company reported 123% revenue growth to $669 million year-over-year from $299 million, beating analyst estimates. EPS was 4 cents versus a loss of 10 cents in the year-ago period.
The Disney+ streaming service beat analyst expectations to reach 116 million subscribers in Disney’s fiscal Q3. Disney now has nearly 174 million subscribers across Disney+, Hulu and ESPN+.
Downloads
Reelgood (update)
Streaming guide Reelgood has historically offered a great service for discovering new things to watch, and keeping track of where you’ve left off with favorite shows. Now, the company is introducing a new feature called Search Party, which makes it easier for two or more people to find something to watch that they all agree on. Using a familiar swipe left or right mechanism, users try to find a “match.” The feature also lets you set other filters, like release year, IMDb rating, genre and more, to narrow its suggestions. When one or more matches is detected, Reelgood notifies the group and displays the matches in a new tab where they’re organized by the total number of “Likes.” Reelgood is a free download on iOS and Android.
PairPlay
PairPlay is a clever new app from Jonathan Wegener, previously co-founder of Timehop and product designer at Snap, which turns a pair of AirPods into a two-person adventure game. You and one other player will split the pair, with one person taking the left AirPod and the other taking the right, to start the audio challenge. The players will hear different sides of the audio adventure story at the same time, which they can then play out together. Storylines may have you playing as secret agents, ghost hunters, robots and more. The game is family-friendly and can be played with kids as young as 7, though arguably some adults will have fun with this one, too. The app is a free download and requires AirPods.
Welcome back to The TechCrunch Exchange, a weekly startups-and-markets newsletter. It’s inspired by what the weekday Exchange column digs into, but free, and made for your weekend reading. Want it in your inbox every Saturday? Sign up here.
Happy weekend, friends. I am writing to you on Friday afternoon after powering through a grilled cheese. But as I have a huge iced coffee on deck, we can dodge a food coma and get right to work. Today we’re talking about a pretty neat venture round, chatting with a founder about verticalization and riffing on Marqeta’s earnings report. So, we have fintech and SaaS and public company notes for your enjoyment. Let’s do it!
HiHello’s ambitious Series A
You may be familiar with Manu Kumar. He’s a venture capitalist at K9 Ventures. But he’s also building a startup at the same time, and it’s the latter effort that we’re interested in today.
The company, called HiHello, raised a $7.5 million Series A, it announced recently. Foundry led the investment, Lux Capital took part, and a host of angels also kicked in checks. So far, so ordinary. But the round is not the interesting bit of the HiHello story. Instead, it’s what the company is building.
A question: When did you last order business cards? I can’t recall, frankly, but somewhere between my last job and coming back to TechCrunch I forgot to get new cards. And not simply thanks to COVID or the fact that I now live far from San Francisco. I just didn’t think of it as they didn’t seem too useful.
HiHello is building something akin to the future of business cards for the internet. Per Kumar, everyone still needs a way to show their identity and introduce themselves, even in a digital world. Sure, for scheduled gatherings, he argued, you can do prep. But for meeting folks in a more unplanned manner, having a way to share your identity is useful.
So, HiHello lets you create virtual business cards of a sort for yourself. But not just one, the idea is to have several, one for each of your personas. Kumar said that I could have one for our podcast (Equity), one for TechCrunch proper and so forth. You can make them for your personal life as well.
I figured that business cards were dead. And that we didn’t need to rebuild them. Kumar doesn’t agree. He sees a future where HiHello can create what are, in effect, personal social networks around context. It’s bold and it’s counterintuitive. Good startup fodder, in other words.
HiHello is monetizing off of consumer revenue today and has a business product as well. Let’s see how quickly the startup can grow. It’s about time we got excited about a new sort of social product.
Going vertical
I’ve written about Skyflow a few times. It’s co-founder, Anshu Sharma, is someone I’ve known for ages. We met when he was at Storm Ventures. Since then he’s invested as an angel and founded a few companies, one of which is Skyflow. The software startup sells a digital vault that allows for PII and other critical information to be secured on customers’ behalf and accessed in a safe manner, allowing companies for whom information security is not their core focus to avoid breaches.
The model is working, with Skyflow raising capital at a pretty aggressive rate. And Sharma seems chuffed thus far with customer progress. (Sharma also provided notes that helped me ground an essay the other day.)
Recently Skyflow announced a particular flavor of product for the healthcare market. Given that I’ve been tracking the company since it first launched, I was curious. So I got Sharma back on the phone to explain his verticalization strategy — I was curious how he was picking markets to pursue and where he might take his company next.
Sharma said that his company’s plan is to prove its technology in complex markets, and then expand its remit over time. Hence the healthcare push and Skyflow’s work with storing financial data. By solving hard problems and selling to complicated customers he said, Skyflow will earn market permission to offer its tech to other folks.
From the CEO’s perspective, we need to “rewire” the internet from the ground up with a privacy focus. Citing a Marc Andreeseen riff about how not building payments tech into the internet from the start was an error, Sharma argued that two things were forgotten in the early days of the web: payments, yes, and privacy.
The verticalization strategy of Skyflow, then, is to tackle the hardest problems that it can — healthcare data is privy to all sorts of rules and regulations — and then broaden its focus until PII is safer for everyone. It’s a fundamentally optimistic take on where the internet could be heading. Not a Facebookian world where privacy is theoretical and adtech is persistent, but a world where your data is yours and is safe, stored and out of reach.
The competitive landscape that Skyflow plays in will harden. But so long as even some of the startups in the market that want to return privacy to individuals wins, I will be content.
Marqeta’s first earnings call
Somewhat lost amidst the wave of IPOs that we’ve seen this year was Marqeta’s debut, a fintech unicorn working in the card issuing space. It reported its earnings publicly for the first time this week, so I got on the phone with its CEO Jason Gardner to yammer about the results.
In brief, Marqeta grew quickly in the second quarter, easily besting expectations. The company lost more money than the markets anticipated however, leading to its shares shedding effectively all their post-IPO gains.
A few notes from the call. First, Gardner seems content to be past the public offering. He said that he’s had the chance to fall back in love with running the company now that his 18-month IPO market is complete. And he said that swapping yearly board-level planning for quarterly reports has been enjoyable, as having more regular disclosures brings a sense of urgency to the company’s work.
As we usually hear private company CEOs worry about distracting earnings calls and the like, it was somewhat refreshing to hear a public executive praise floating their company. It reminded us of the comments that we heard from BigCommerce CEO Brent Bellm on the same topic, even if they like being public for different reasons.
More important to our understanding of the world of startups, however, was Marqeta’s notes on the BNPL market. In the wake of Klarna’s rise, Square buying Afterpay and a zillion startup BNPL rounds, seeing Marqeta note the buy now. pay later space as a growth market for its work caught our eye. Why was BNPL helping a card issuing platform?
Well, it turns out, the virtual cards that Marqeta and others can spin up for customers are often used as part of the software sinew that makes BNPL transactions possible. The fintech world is always more interconnected than you expect. So, when we consider BNPL as a category, we’ll do well to also keep tabs on what other boats its growth may be floating. That expands the number of startups that could be riding the BNPL wave.
One tip before we go. The fastest way to get an explanation of a market dynamic that you are not familiar with is to ask a public CEO to explain it to you. The downside to this particular educational method is that if you were close to understanding the concept before but missed a single key element, you will feel pretty silly when said CEO tells you in small words what you previously failed to grok.
However, as I am, in fact, very dumb, I refuse to be red-faced about not knowing things. Alright, that’s enough for today. There’s an extra Equity episode out today, and The Exchange is back on Monday morning!
Hello and welcome back to TechCrunch’s China roundup, a digest of recent events shaping the Chinese tech landscape and what they mean to people in the rest of the world.
A sexual assault case at Alibaba has sparked a new round of #MeToo reckoning in China. Industry observers believe this is a watershed moment for the fight against China’s allegedly misogynist tech industry. Meanwhile, social media operators are still undecided on how to deal with the unprecedented public uproar against the powerful internet giant.
In other news, more Chinese tech companies have delayed plans to go public overseas after Didi’s fallout with Chinese regulators over its rushed IPO, including Tencent’s music streaming empire and one of China’s highest-valued autonomous driving startups.
Call for justice
Just past midnight last Sunday, an Alibaba employee posted on the company’s internal forum a detailed account saying her manager and a client had sexually assaulted her on a business trip. She took the case public after failing to obtain support from her superiors and human resources.
The post quickly made its rounds through China’s social media platforms. People stayed up blasting Alibaba’s ignorance, toxic business drinking, and the pervasive objectification of women in the Chinese “tech industry,” which has grown so far-reaching that it’s just the contemporary corporate world.
A day later, on August 9, Alibaba swiftly fired the alleged perpetrator. Two managers resigned and the firm’s head of HR was given a “disciplinary warning.” Alibaba’s CEO Daniel Zhang said he felt “shocked, angry and ashamed” about the incident and called on the company to work with the police to investigate the case.
This is arguably the most high-profile #MeToo case embroiling a major Chinese tech company by far and one that seems to have beckoned the toughest response from the company involved. Alibaba is formulating company policies to prevent sexual assaults, which surprises many that the global tech behemoth didn’t already have those in place.
The case managed to garner widespread public attention in China thanks to social media. Within the first few hours, it seemed as though discussion around the incident was propagating organically and uncensored on microblogging platform Weibo, in which Alibaba owns a majority stake.
But people soon noticed that despite the severity of the event, it took days before the case climbed to the top of Weibo’s trending chart, a bellwether for the most talked about topic on the Chinese internet. The perceived delay recalls Weibo’s censorship of an extramarital affair involving Alibaba executive Jiang Fan last year.
Talang Qingnian, roughly “Surfing Youth,” a social media column under state paper People’s Daily, blasted in an article:
The slow buildup of discussion again raised suspicion over whether Alibaba has manipulated public discourse.
Ever since the Jiang Fan case, the country’s attitude has been very clear that capital must not control the media.
As the basic infrastructure for truthful news in China, Weibo should not be a tool for any stakeholder to manipulate public opinion.
The article fanned up more public outrage but was soon taken down, likely because its wording was too strong. The Chinese state media apparatus is vast and only a few outlets, such as Xinhua, consistently convey top-level leaders’ official opinions. It’s not uncommon to see the less authoritative state-affiliated publications back down on reports that have cause backlashes. Last week, an article from a state-affiliated economic paper removed a piece calling video games “spiritual opium,” a loaded description that had earlier tanked the stocks of Tencent and NetEase, and republished the article with a softer tone.
Smaller war chests
Regulatory uncertainties have always been flagged as a risk by Chinese companies seeking overseas listings, but it was largely up to foreign investors to decide whether they were worthwhile investments. China’s recent regulatory onslaught on its tech darlings, however, has become a real deterrent for Chinese firms’ IPO dream.
This week, reports arrived that NetEase Music, a popular music streaming service, and Pony.ai, an autonomous vehicle startup last valued at $5.3 billion, have respectively postponed their plans to list in Hong Kong and New York.
Beijing has become warier of its data-rich companies getting scrutinized by U.S. regulators. Last month, the U.S. securities regulator said Chinese companies that want to raise capital in the U.S. must provide information about their legal structure and disclose the risk of Beijing’s interference in their business.
Many Chinese tech firms have learned from Didi’s fallout with the government, which had reportedly told the ride-sharing company to hold off on its listing until it sorted out a data protection framework. Didi went ahead regardless, triggering a government probe into its data practice and tanking its shares, which now stand at $8 apiece compared to $16 around its debut in early July.
Beijing’s crackdown has affected every major player in China’s consumer tech sector, wiping as much as $87 billion off the net worth of the country’s tech billionaires, including Pony Ma of Tencent and Colin Huang of Pinduoduo, according to Financial Times. The government wants “hard tech” like semiconductors and clean energy, so it has made it clear to future entrepreneurs where they should allocate their energy. The new generation of startups is listening now.
Liz O’Sullivan is CEO of Parity, a platform that automates model risk and algorithmic governance for the enterprise. She also advises the Surveillance Technology Oversight Project and the Campaign to Stop Killer Robots on all things artificial intelligence.
Advocates of algorithmic justice have begun to see their proverbial “days in court” with legal investigations of enterprises like UHG and Apple Card. The Apple Card case is a strong example of how current anti-discrimination laws fall short of the fast pace of scientific research in the emerging field of quantifiable fairness.
While it may be true that Apple and their underwriters were found innocent of fair lending violations, the ruling came with clear caveats that should be a warning sign to enterprises using machine learning within any regulated space. Unless executives begin to take algorithmic fairness more seriously, their days ahead will be full of legal challenges and reputational damage.
What happened with Apple Card?
In late 2019, startup leader and social media celebrity David Heinemeier Hansson raised an important issue on Twitter, to much fanfare and applause. With almost 50,000 likes and retweets, he asked Apple and their underwriting partner, Goldman Sachs, to explain why he and his wife, who share the same financial ability, would be granted different credit limits. To many in the field of algorithmic fairness, it was a watershed moment to see the issues we advocate go mainstream, culminating in an inquiry from the NY Department of Financial Services (DFS).
At first glance, it may seem heartening to credit underwriters that the DFS concluded in March that Goldman’s underwriting algorithm did not violate the strict rules of financial access created in 1974 to protect women and minorities from lending discrimination. While disappointing to activists, this result was not surprising to those of us working closely with data teams in finance.
There are some algorithmic applications for financial institutions where the risks of experimentation far outweigh any benefit, and credit underwriting is one of them. We could have predicted that Goldman would be found innocent, because the laws for fairness in lending (if outdated) are clear and strictly enforced.
And yet, there is no doubt in my mind that the Goldman/Apple algorithm discriminates, along with every other credit scoring and underwriting algorithm on the market today. Nor do I doubt that these algorithms would fall apart if researchers were ever granted access to the models and data we would need to validate this claim. I know this because the NY DFS partially released its methodology for vetting the Goldman algorithm, and as you might expect, their audit fell far short of the standards held by modern algorithm auditors today.
How did DFS (under current law) assess the fairness of Apple Card?
In order to prove the Apple algorithm was “fair,” DFS considered first whether Goldman had used “prohibited characteristics” of potential applicants like gender or marital status. This one was easy for Goldman to pass — they don’t include race, gender or marital status as an input to the model. However, we’ve known for years now that some model features can act as “proxies” for protected classes.
If you’re Black, a woman and pregnant, for instance, your likelihood of obtaining credit may be lower than the average of the outcomes among each overarching protected category.
The DFS methodology, based on 50 years of legal precedent, failed to mention whether they considered this question, but we can guess that they did not. Because if they had, they’d have quickly found that credit score is so tightly correlated to race that some states are considering banning its use for casualty insurance. Proxy features have only stepped into the research spotlight recently, giving us our first example of how science has outpaced regulation.
In the absence of protected features, DFS then looked for credit profiles that were similar in content but belonged to people of different protected classes. In a certain imprecise sense, they sought to find out what would happen to the credit decision were we to “flip” the gender on the application. Would a female version of the male applicant receive the same treatment?
Intuitively, this seems like one way to define “fair.” And it is — in the field of machine learning fairness, there is a concept called a “flip test” and it is one of many measures of a concept called “individual fairness,” which is exactly what it sounds like. I asked Patrick Hall, principal scientist at bnh.ai, a leading boutique AI law firm, about the analysis most common in investigating fair lending cases. Referring to the methods DFS used to audit Apple Card, he called it basic regression, or “a 1970s version of the flip test,” bringing us example number two of our insufficient laws.
A new vocabulary for algorithmic fairness
Ever since Solon Barocas’ seminal paper “Big Data’s Disparate Impact” in 2016, researchers have been hard at work to define core philosophical concepts into mathematical terms. Several conferences have sprung into existence, with new fairness tracks emerging at the most notable AI events. The field is in a period of hypergrowth, where the law has as of yet failed to keep pace. But just like what happened to the cybersecurity industry, this legal reprieve won’t last forever.
Perhaps we can forgive DFS for its softball audit given that the laws governing fair lending are born of the civil rights movement and have not evolved much in the 50-plus years since inception. The legal precedents were set long before machine learning fairness research really took off. If DFS had been appropriately equipped to deal with the challenge of evaluating the fairness of the Apple Card, they would have used the robust vocabulary for algorithmic assessment that’s blossomed over the last five years.
The DFS report, for instance, makes no mention of measuring “equalized odds,” a notorious line of inquiry first made famous in 2018 by Joy Buolamwini, Timnit Gebru and Deb Raji. Their “Gender Shades” paper proved that facial recognition algorithms guess wrong on dark female faces more often than they do on subjects with lighter skin, and this reasoning holds true for many applications of prediction beyond computer vision alone.
Equalized odds would ask of Apple’s algorithm: Just how often does it predict creditworthiness correctly? How often does it guess wrong? Are there disparities in these error rates among people of different genders, races or disability status? According to Hall, these measurements are important, but simply too new to have been fully codified into the legal system.
If it turns out that Goldman regularly underestimates female applicants in the real world, or assigns interest rates that are higher than Black applicants truly deserve, it’s easy to see how this would harm these underserved populations at national scale.
Financial services’ Catch-22
Modern auditors know that the methods dictated by legal precedent fail to catch nuances in fairness for intersectional combinations within minority categories — a problem that’s exacerbated by the complexity of machine learning models. If you’re Black, a woman and pregnant, for instance, your likelihood of obtaining credit may be lower than the average of the outcomes among each overarching protected category.
These underrepresented groups may never benefit from a holistic audit of the system without special attention paid to their uniqueness, given that the sample size of minorities is by definition a smaller number in the set. This is why modern auditors prefer “fairness through awareness” approaches that allow us to measure results with explicit knowledge of the demographics of the individuals in each group.
But there’s a Catch-22. In financial services and other highly regulated fields, auditors often can’t use “fairness through awareness,” because they may be prevented from collecting sensitive information from the start. The goal of this legal constraint was to prevent lenders from discrimination. In a cruel twist of fate, this gives cover to algorithmic discrimination, giving us our third example of legal insufficiency.
The fact that we can’t collect this information hamstrings our ability to find out how models treat underserved groups. Without it, we might never prove what we know to be true in practice — full-time moms, for instance, will reliably have thinner credit files, because they don’t execute every credit-based purchase under both spousal names. Minority groups may be far more likely to be gig workers, tipped employees or participate in cash-based industries, leading to commonalities among their income profiles that prove less common for the majority.
Importantly, these differences on the applicants’ credit files do not necessarily translate to true financial responsibility or creditworthiness. If it’s your goal to predict creditworthiness accurately, you’d want to know where the method (e.g., a credit score) breaks down.
What this means for businesses using AI
In Apple’s example, it’s worth mentioning a hopeful epilogue to the story where Apple made a consequential update to their credit policy to combat the discrimination that is protected by our antiquated laws. In Apple CEO Tim Cook’s announcement, he was quick to highlight a “lack of fairness in the way the industry [calculates] credit scores.”
Their new policy allows spouses or parents to combine credit files such that the weaker credit file can benefit from the stronger. It’s a great example of a company thinking ahead to steps that may actually reduce the discrimination that exists structurally in our world. In updating their policies, Apple got ahead of the regulation that may come as a result of this inquiry.
This is a strategic advantage for Apple, because NY DFS made exhaustive mention of the insufficiency of current laws governing this space, meaning updates to regulation may be nearer than many think. To quote Superintendent of Financial Services Linda A. Lacewell: “The use of credit scoring in its current form and laws and regulations barring discrimination in lending are in need of strengthening and modernization.” In my own experience working with regulators, this is something today’s authorities are very keen to explore.
I have no doubt that American regulators are working to improve the laws that govern AI, taking advantage of this robust vocabulary for equality in automation and math. The Federal Reserve, OCC, CFPB, FTC and Congress are all eager to address algorithmic discrimination, even if their pace is slow.
In the meantime, we have every reason to believe that algorithmic discrimination is rampant, largely because the industry has also been slow to adopt the language of academia that the last few years have brought. Little excuse remains for enterprises failing to take advantage of this new field of fairness, and to root out the predictive discrimination that is in some ways guaranteed. And the EU agrees, with draft laws that apply specifically to AI that are set to be adopted some time in the next two years.
The field of machine learning fairness has matured quickly, with new techniques discovered every year and myriad tools to help. The field is only now reaching a point where this can be prescribed with some degree of automation. Standards bodies have stepped in to provide guidance to lower the frequency and severity of these issues, even if American law is slow to adopt.
Because whether discrimination by algorithm is intentional, it is illegal. So, anyone using advanced analytics for applications relating to healthcare, housing, hiring, financial services, education or government are likely breaking these laws without knowing it.
Until clearer regulatory guidance becomes available for the myriad applications of AI in sensitive situations, the industry is on its own to figure out which definitions of fairness are best.
This week Danny and Alex and Chris took to Twitter Spaces to chat about the current state of the crypto economy, and hang out with friends in a live Twitter Space. We’re doing more of these, so make sure that you are following the show on Twitter.
As a small programming note, I forgot to tell the folks who chimed in during the chat that we were recording it, so we had to cut most the Q&A portion of the show. We got Ezra’s permission, thankfully. The mixup was a bummer as we learned a lot. In the future, we’ll not make that mistake and keep all the voices.
And everyone wants to fund the next Coinbase. You can understand why. The company is printing cash lately, helping drive more investment into localized exchanges in different markets.
Ok, we’re back Monday with your regularly scheduled programming!
Equity drops every Monday at 7:00 a.m. PST, Wednesday, and Friday morning at 7:00 a.m. PST, so subscribe to us on Apple Podcasts, Overcast, Spotify and all the casts.
Privacy-oriented search app Xayn 
Pokémon GO developer Niantic 
