Fake Cues: Why The Next Photo Innovation May Be Helping You Tell a Lie

I want to share something which might be a bit unsettling – whenever I’m looking at you, I’m judging. Whether it’s in person or a static photo. I just can’t help it. It’s how I’m wired. By the way, it’s how you’re wired too. Perhaps our most basic survival skill is the near constant assessment and reassessment of other people based on their facial structures, their demeanor, slight changes in their movement. Which makes one wonder: can I “hack” this evolutionary necessity and influence the way you react to me?

For example, “microexpressions” are defined as “brief, involuntary facial expression shown on the face of humans according to emotions experienced. They usually occur in high-stakes situations, where people have something to lose or gain.” Perhaps you remember “Lie to Me,” the Tim Roth tv show where he played the world’s leading expert of microexpression-reading (now THAT would be a good LinkedIn Endorsement).

Screen Shot 2017-09-18 at 2.27.58 PM

But it doesn’t even require the fidelity of a realtime interaction for our monkey minds to start forming an opinion of someone else. All it takes is a face, even a static picture or artistic representation. Did you know we typically find facial symmetry more attractive because it potentially signals high breeding status? Side note – I’m fairly certain I’m asymmetrical AF.

So when hearing that the new photo sharing app Polygram reads the facial reactions of the viewer to tell you whether they liked or disliked your photo, well, that set off a bunch of ideas. We’ve undoubtedly already trained machine learning models to predict the “attractiveness,” “honesty” etc of people depicted in a photo. What happens when we start running this software not just in post-photographic analysis but in photo creation and editing? For example:

  • Selfie mode of a camera could let you select what emotions you want to provoke in the average viewer, give you some basic facial motions to mimic and then shoot a burst of photos, selecting from the burst the pictures which are most likely to work.
  • Photoshop could have buttons for “honesty,” “attractiveness,” “happiness,” etc and move your facial bits around in the smallest ways needed to “enhance” this aspect of your person. The changes likely wouldn’t need to be very significant – they’re called “microexpressions” for a reason.
  • Run software against models in commerce site, Tinder profiles, realtor photos and so on. ID any photos where the viewer would have real but imperceptible negative reaction. Prime them to buy, to swipe right.

So yeah, if software is going to help us read emotional reactions you can be sure it’s going to be used to manufacture them as well.

The Markets Ignore Inputs But Over Time, Not The Means

When asked for advice about a career in tech my feedback varies based on the individual’s goals but always ends with “…you can recover from any failure except one where you ruined your reputation.” Basically, trying to make sure that the desire to succeed, to impress, doesn’t overwhelm someone young in their career and cause them to compromise their ethics or values.

My friend and frequent co-investor Semil Shah wrote a good post this weekend called The Market Ignores Inputs. He writes,

Screen Shot 2017-09-18 at 10.33.55 AM

Absolutely true. Satya and I often describe how we’re thankful of the solid personal relationships we maintain with our investors and the shared values, but also know that ultimately our returns are the primary point of evaluation for the continuing business relationships.

The majority of investors I respect (Semil included) are output-focused but with a clear and consistent morality. For them the ends don’t justify the means, if that includes operating in a manner which is poisonous, ethically troubled or illegal. And it’s not about getting cute and trying to avoid knowing (ie “The doctor turned around so he could have deniability“). The Hippocratic oath of Venture Capital extends to helping founders maintain their true north, sometimes in the face of immense pressures and economic consequence. That’s an input you don’t compromise on.

Thinking About Bodega

As investors, we hope to help founders see around corners in order to assess and predict outcomes. Startups, especially early stage startups, are in the weird position of simultaneously trying to reduce known risks while embracing new risks. Today, one of our portfolio companies launched and was met with a strong, unexpected reaction, one I didn’t anticipate. This post is primarily about trying to understand that lack of anticipation on my part, and secondarily, providing some insight into why I believe in this company.

The company is called Bodega and their goal is to use technology to extend commerce to spaces where today you can’t put a store, or where an existing impersonal vending experience can be improved upon. Bodega’s vision is that these flexible devices can do several things at once (and here’s some more in their own words):

  • Provide better item inventory driven by local preferences, data analysis and curation
  • Support not just sales, but item rental, item exchange and so on
  • Be price competitive with delivery-service alternatives
  • Give store owners, property owners and individual entrepreneurs a chance to run a micro-business in an updating of the classic vending model

Bodega’s vision is not, and has never been, to compete with or replace the urban corner store. Bodega doesn’t want to disrupt the bodega. Some instances of today’s press coverage suggested that element, a soundbite which, exacerbated by Bodega’s naming, pissed people off as another example of tech startups being at best tone-deaf, and at worst, predatory. This article in Forbes explains Bodega well while appropriately critiquing the challenges they face in succeeding. We – and the other investors – committed capital, sweat and reputation behind a team that absolutely is working on a problem that matters to them and we believe can be meaningful to customers. And that’s the founder-market fit we seek.

So, About Our Name is CEO Paul McDonald’s explanation and introspection of the Bodega brand. Let me tldr by saying I agree with Paul’s commitment here to listen to and understand the feedback. And I believe they want to build a durable and thoughtful company where the decisions they make – brand included – represent their values.

But even though I looked at this name for several months pre-launch, why didn’t I anticipate the ways it could be interpreted? When I first heard it my biggest concern was, would anyone outside of NYC understand it? The early part of my childhood was in Queens, NY and bodegas were beloved, but did it translate outside of the five boroughs? The team’s market research suggested that it did – and in doing so they also spoke with consumers of different ethnic and cultural backgrounds.

Once that research came back positive, I was sold. It didn’t occur to me that some people would see the word and associate its use in this context with whitewashing or cultural appropriation. I heard it in a different way than some others are hearing it today. And that leaves me wondering why, because as an investor, and even more importantly as a human being, it’s an awareness that I need. So like the founders, I too want to listen and better understand the lines between homage and respect versus exploitation and insensitivity. Today tells me it’s a personal blindspot and to assist founders, to help them see around corners, I need to see clearly.

The Bot Registration Act of 2017 Could Improve Twitter For Us Humans

This sounds like an X-Men storyline, but Twitter needs to ask all bot accounts to register as such & then badge. Bots can be very useful but users should know they’re following a bot & bots should follow certain ToS. Treat bots like a developer ecosystem, not like user accounts.

Twitter hasn’t been able to effectively police their bot ecosystem – I don’t know if it’s will (desire) or way (ability, prioritization) – but it seems that they’d want to understand what percentage of their activity is machine-driven versus accounts piloted by actual human beings. The company’s policy lead published some general “we take bots seriously” thoughts earlier this summer but the post doesn’t clarify whether Twitter believes bots should be identified in a consistent user-facing manner. One of my favorite Bot-ologists, Renee DiResta, also recently suggested that bot accounts be labeled with a robot emoji, or some similar demarcation in the name or bio. (In that article Renee also distinguished between Bots and Cyborgs, both automated accounts but where the latter spoofs human behavior).

So want to follow a bot which tweets whenever NYTimes Maggie Haberman publishes a new article? Awesome. Whenever USGS detects a big earthquake? Yes! Which combines two current news events into a single headline? I don’t, but you do you. These and many more wonders exist on Twitter, but without a Bot designation and Bot directory, who knows what the average user assumes. It also gives Twitter license to shut down Bots which don’t register and more easily monitor account behaviors. As it stands, who knows what percentage of Twitter accounts are bots and what impact these bots are having on user experience. Well, I guess we’re incrementally answering this question without Twitter’s help….


The One Question Investors Should Ask Founders On Demo Day

“What you’ve accomplished is really great and we’re excited to introduce ourselves and chat a bit more. Would you be willing to share your application to [incubator/accelerator] with us so that we can use it as a way to collaboratively discuss what you’ve learned since then?”

For many accelerators and incubators, demo days kick off a fundraising process for a full round or additional capital based on inbound demand. Homebrew’s focus is ~8 or so seed commitments per year, and we’re putting dollars, sweat and reputation behind those companies ongoing. As a result our investment process is geared towards giving teams a chance to understand what it’s like to work with us and, on our end, confirming that we can be a good partner to the founders. Sometimes we already know the team but what happens when you’re meeting them for the first time on demo day and want to accelerate towards familiarity?

Well, instead of just focusing on where the company is today and what the go-forward plan looks like, I’d add a perhaps counter-intuitive discussion: where they’ve come from, using their accelerator/incubator application as a starting point. Why “waste” time looking backwards? Two main reasons:

  1. Evaluate Learning, Not Just Milestones: every demo day presentation – complete with ‘up and to the right’ curve – is marketing. It’s the best attempt to showcase where this company is today and where they hope to go. Reviewing their application, which is usually 5-6 months old, provides a sense of who they were at that time, what they thought was important and how they ‘pitched’ themselves then. I wouldn’t expect it to have fully or accurately represented the future any more than a Five Year Plan at seed stage is a true roadmap, but it does serve as another “dot” for us to connect the dots and “draw a line.”
  2. Cultural Alignment: Are we going to have a long, productive relationship where we can trust one another and Homebrew can put 100% of itself behind helping you build a company that you’re proud of? That’s probably the central question we look to answer during a fundraise and equally, we want to display to a founder that we’re going to be worthy of their trust and commitment ongoing. Reviewing the application together shows you how an investor will react to founders making decisions with incomplete information. Will your investors want everything to be neat and tidy all the time (and cause you grief when it isn’t) or do they understand that building a company is difficult and messy? Are they investing in the glossy Instagram version of you, or are they investing in the real you? The messy version. The high-potential version. The version that sometimes needs a pat on the back or a kick in the ass. That’s who we invest in. So, to me, finding a way to ‘get real’ during an accelerated fundraising process creates mutual signal as to whether we’re going to be a good fit for one another. Because if not, let’s find that out now, not 1, 10, 100 days after we’ve lashed ourselves to each other.

In reflecting, perhaps this isn’t just the “one question” investors should ask founders, but one founders want to offer up. Take your potential lead investors through your evolution, not just your pitch deck. My bet is anyone you “lose” wasn’t the right investor for you to begin with. And you might just end up with better investors who are committed to the journey with you, not just the demo day hype.

“My editors value impact stories” – Recode’s Jason Del Rey on Evaluating His Success

I think his colleague Peter Kafka introduced us? Or at least that’s my best guess for how Recode’s Jason Del Rey and I started hanging out a few years back. Jason rocks the Commerce beat but his reporting chops are only the start of my admiration. He’s a dad to two young kids and it’s been fun seeing that side of him develop. In fact, I ask about it below and Jason’s answer is a great reminder about the importance of us all seeing one another as humans. So here’s Five Questions…

Screen Shot 2017-08-20 at 11.01.13 PM

Asa Mathat for Vox Media

Hunter Walk: The way media companies make money (or measure success) can sometimes be inconsistent with what I think leads to good reporting. For example, page views and being “first” versus in-depth analysis. How does Recode measure a reporter’s contributions? How do you know if you’re doing a “good job?”

Jason Del Rey: I have quarterly check-ins with our managing editor, where we discuss my goals for the coming quarter, what I accomplished in the previous three months and how I can improve. My editors value impact stories—articles that typically start a conversation and/or serve as a jumping-off point for a lot more coverage on a specific topic within a beat.

Oftentimes, impact stories are well-read. And if they’re not, there’s usually something that went wrong, whether it be the headline or some other part of the packaging. It’s important to try to suss out what that is.

Impact stories don’t just come overnight, though. They are usually the result of a steady stream of smaller reports on a topic, which are also important for a site like Recode to maintain momentum and freshness each day. The never-ending battle I have with myself is how to best balance the newsier items with the big, step-back impact pieces. It’s not a science and, when in doubt, I ask one of my editors.

HW: Continuing with the “making money,” Recode has a very valuable event businesses alongside their site. You run Code Commerce, which relies upon getting interesting tech folks to serve as guests. How do you maintain objectivity in your reporting given these potential conflicts?

JDR: My reporting and writing is what has gotten me to this point, and is what will hopefully allow me to enjoy several more decades in this profession. Most smart people in the industry realize and respect that, and will sit down with me onstage whether they view my coverage as “positive” or “negative,” so long as they believe it’s well-researched and fair. There are exceptions, of course, but you won’t see them on my stage. It’s a big industry and I don’t have time for that strain of BS.

HW: Similarly you’re a conscientious reporter but also maintain friendships with many of the people in our industry. Has that put you in awkward situations? You and I are close—how do you think I’d react if you came out with a really negative piece about one of our portfolio companies?

JDR: I’m friendly with lots of people in the industry, but I spend the vast majority of my non-work time with my family and with my friends that I’ve known since college or earlier. I don’t know that that’s intentional, but it probably does help me avoid some awkward situations that clubbier reporters may face.

As for you, I think you’d understand it if you thought the piece was well-researched and fair. You’d be surprised how many times someone on the wrong side of a harsh story sends me a text that says something like: “Sucks you had to write that, but thanks for being fair.” You should try it sometime!

HW: You’re a dad now. How has having kids changed your approach to work generally?

JDR: Is there a word limit on this? The first thing that comes to mind is how I judge the people I interview. I’ve always been a reporter who over-analyzes everything about a subject the first time I meet them: What is their tone with the waiter at the cafe? How real was their apology when they show up 15 minutes late? Why did their body language change when I ask about their co-founder?

I still think reading people is a really important part of the job, but I now understand that I’m not seeing their whole life in front of me in a 30-minute meeting. Maybe their one-year-old threw up on their shoulder on the way out the door. Maybe they’re in the middle of a rough patch in their marriage and they’re not being cold—they’re just distracted. In short, I probably give people the benefit of the doubt much more than I did pre-kids, even though I still bring a healthy dose of skepticism to every interview.

The other thing I certainly have had to adjust is how and when I get my work done. I work as hard as I ever have but I refuse to work a ton on weekends if it’s not a breaking-news situation. That means being much more productive during weekdays than I may have been earlier in my career. It also means working on stuff later on weeknights than is ideal a lot of the time.

HW: When a source leaks news to you, how do you decipher what their motivation might be and how does your estimation of their trustworthiness factor into how you use that information?

JDR: A lot of times people tell me or it is just obvious. If I can’t figure it out, I just ask. Either way, knowing someone’s motivation is crucial to weighing if and how to use the information they are sharing.

Trustworthiness is critical, too. I, like other reporters I respect, probably have missed out on stories because it’s coming from a source I don’t know well or don’t fully trust. On the other hand, that approach greatly reduces the chances of getting duped, looking like a fool, and ruining my credibility. If you want to have a long career, I think this is the way to go.

Thanks Jason! Follow him on Twitter!

This Former Hillary Staffer Is Building Aloe, A Self-Care App. Meet Amber Discko.

I know Amber via her kind and helpful Twitter presence (I *think* @Iano introduced us). So when she launched a recent Kickstarter for Aloe, a self-care app, it felt right to jump in and support her. As the project gained momentum I was surprised by two types of reactions: first, the large number of teens who really were into the idea, and second, the harassers who belittled the concept. As a friend, and now Aloe backer, I wanted to learn a bit more about Amber and Aloe, so Five Questions.


Hunter Walk: Besides the holder of an awesome name, who is Amber Discko?

Amber Discko: Professionally I’m a freelance digital strategist who most recently worked on the Hillary for America campaign, a creative strategist at Tumblr, and fun fact I was the community manager for Denny’s Diner back when it was innovative for a brand to be good on Twitter. Personally? I’m still figuring that one out. I live in Brooklyn, NY with my cat @ScotusCat and partner. I’m a huge fan of using social media to create social change and filling my followers feeds with wholesome memes and positivity.

HW: You spent a few years working at Tumblr. What was your gig there and what did being on the inside teach you about how people were using Tumblr?

AD: I was on the creative strategy team there which involved helping brands understand Tumblr and all the wonderful communities that make it so special. I’d also sometimes go to different agencies to talk about using the Tumblr platform and working with its creators. Talking about these communities constantly and seeing how they interacted and grew was really cool to watch. It definitely made me feel like I was witnessing something unique.

HW: Should platforms like Twitter or Facebook take stronger stances on harassment and hate speech or is it too slippery a slope to centralize this type of judgment instead of giving users the tools to moderate their own experiences?

AD: Yes I do believe social networks have a responsibility to protect its users from harassment and hate speech. As someone who was a target of online harassment for many years, the self moderating tools are nice to see. It just sucks that they allow people to cause so much pain to others. It would be nice to see more action and holding people accountable versus just making the community do the work. The people at the top of these platforms shaped their communities and allowed hate and harassment to thrive, it’s their responsibility to make it better if they truly care about their users.

HW: I backed your Kickstarter for Aloe, a self-care app. What’s the origin story behind the project?

AD: Thank you so much for backing!!! Long story short, I was not taking care of myself while working on the Hillary Clinton campaign. I realized that I kept forgetting to do basic tasks like drink water, give myself a break from the news cycle, and even eat. I created a self-care survey tool which I used to help me check-in with myself at various times throughout the day by setting alarm reminders on my phone. It was the only thing that seemed to be working for me. After the Inauguration, I decided to share my tool and start building this new community around a topic that is so important, especially now. I noticed a lot of people were using the check-in tool on mobile, and I kept getting people asking me for an app, so it only made sense.

HW: Since I follow you on Twitter, I’ve seen you RT a number of teenagers who seem really excited about Aloe, why do you think it’s connected with that demographic so intensely?

AD: Growing up I always felt an intense pressure to be “normal” or “fit in” and in return was always too hard on myself for how hard it seemed to be this way. I would have loved to have had a pocket cheerleader encouraging me on days that seemed extra hard. There’s so much pressure to be online and using social media, so having an app to remind someone to take a break when they need it is proven to be something that works. Since bringing this idea to life, hundreds of young people have come to me expressing how much they love Aloe and how excited they are for this app to come to life.

The reality though is that these teens can’t afford to back this app. That’s why I’ve made it so people can donate their beta ($30) or alpha tester ($60) account to someone in need. After the campaign ends I’ll be matching people with a sponsor and it will be really wholesome and good for all.

Thanks Amber! You can follow her on Twitter