Paws and Take a Deep Breath

Believe it or not, we have the same neck size.

After two-and-a-half incredibly memorable years at Scopely, last Friday was my last day. While there are dozens of coworkers that I already miss quite dearly, another hard thing to walk away from was having my dog alongside me every day of the week.

Since we first brought him home as an anxious, 3-month-old puppy, my French Bulldog "Mochi" spent every weekday of his life at Scopely HQ. All day long, he diligently rotated between a self-prescribed set of tasks: roving throughout the office demanding copious scratches from my coworkers, hunting assiduously for fallen taco scraps, forcing himself into any warm lap to nap in our quiet room, and rolling around on his back while making improbably adorable dragon noises.

As my friend David Adelman wrote in Forbes, there are a number of great reasons to have a dog at a startup: mood/morale, forced breaks, ice-breaking, entrepreneurial skills ... but for me personally, it really comes down to simply, "perspective." Whenever things were at their absolute worst, Mochi was a constant reminder that everything was going to be just fine. We could be sitting in the office at 2 am on a Saturday facing a looming deadline, or discovering a crippling flaw with our data integrity—but with Mochi sleeping upside-down and snoring like a furry chainsaw, it was easy to remember that life is so much bigger than API calls or even major product launches.

To look at things from a more clinical perspective, Swedish professor Kerstin Uvnas-Moberg took blood samples from both pets and owners before and during a petting session.  As the Professor noted in the PBS special Dogs Decoded, "we had a basal blood sample, and there was nothing, and then we had the sample taken at one minute and three minutes, and you could see this beautiful peak of oxytocin."   To which the narrator followed up, "Oxytocin has a powerful physiological effect. It can lower the heart rate and blood pressure and may lead to reduced levels of stress. Research indicates that owning a dog could even extend your life."

I'm incredibly excited for this next chapter in my career, but there's no question I'll miss Mochi's constant reminders to take everything in stride and just enjoy the ride.

(paws for effect)

Heisenberg, Hemodynamics and Direct Response TV

Reading this Forbes piece about Biz Stone and conceptual blending made me realize that one of the things I enjoy most about working with people from diverse backgrounds is the parallels that inevitably get drawn to entirely separate fields. Just this week, I had conversations about mobile gaming that improbably ended up spanning neuroscience and theoretical physics as well. Despite my utter lack of fluency in all things scientific, it's hard not to to be tickled by such discussions.

Let me set the stage: I've recently been thinking about different approaches to measuring the success of television ads for mobile games. Unlike viewers of traditional DRTV campaigns, mobile game consumers are loath to respond via direct response channels (vanity URLs, SMS short codes, etc) and instead choose in overwhelming numbers to head straight to the App Store or Google Play and download apps directly. Further confounding measurement is the fact that the first point in the “funnel” at which one can effectively record an app install is when a user first launches an app, which can easily take place anywhere from minutes to days or even weeks after they initially download it.

I first found myself chatting with Scopely’s data scientist Gabe Donnay. In addition to being half of the killer electronic duo Satchmode, Gabe’s research background in neuroeconomics and neuroimaging surfaced some interesting parallels. In particular, he cited the challenge of measuring hemodynamic response (HDR) via fMRI, given the overlapping responses to a series of stimuli. In other words, measuring how our neurons fire in response to triggers. One common solution is to utilize block design, which essentially entails presenting similar groups of stimuli in quick succession and discrete, alternating blocks—ideally resulting in more easily-perceptible HDR since the amplitude resulting from each block are additive. Or as applied in a simplistic way to DRTV, alternating between running a tightly-grouped set of ads and none at all.

Later that same day, I was discussing DRTV mobile app measurement with the CTO of a direct response agency who also happens to hold a PhD in theoretical physics. In his case, he pointed out that the very addition of a call-to-action (e.g. “text GOLF1 to 77123 for a download link!”) in and of itself affects the underlying signal we’re trying to measure—which has nice echoes of the Heisenberg uncertainty principle or the observer effect (I leave it to the more qualified readers among you to determine which one is more applicable). That is, adding a direct response call-to-action to a TV ad fundamentally changes its response rate, complicating any sort of measurement of the underlying ad.

The “other” Heisenberg Uncertainty Principle: trying to determine if your partner in crime betrayed you in and of itself influences his propensity to do so. There's truly a meth to his madness.

Even though both of these concepts are far too complex for me to fully understand in their native contexts, each of them has thought-provoking and practical implications for mobile app advertising that even my decidedly science-averse brain can appreciate.

Plus, situations like these remind me why I love working with people from diverse backgrounds, from luchadoras and screenwriters to recovering lawyers and ice truck drivers (names redacted to protect the wonderfully guilty). Any team that truly aspires to disruption or innovation would do well to place a premium on hiring for and encouraging diversity of background and thought.

Happiness Can't Be Linear Programmed

Captain Obvious here, reporting for duty with breaking news: happiness can’t be linear programmed. “But, Bill,” you ask, being a sane and rational person, “who would utilize a mathematical model to optimize an emotional outcome?”  I’ll tell you who--optimizers like me.

Whenever I’m faced with a big decision in life, my first instinct is to encode every possible consideration in a byzantine, unmanageably complex linear program with constraints, sensitivity analyses, and the like. Although I’m not talking about literally firing up Excel and building a linear optimization with numerical values (don't tempt me!), that’s a pretty solid analog for what does transpire inside my brain.

Let me give you an example. Once upon a time, I was finishing up graduate school in Philadelphia.  It had been a long two years away from my beloved New York, but in less than a month I would be blissfully wrapped in her crazed, urban arms once again. I could practically see the sun setting from the Brooklyn Bridge, hear Mimì coughing at the Met, and taste #78 with Beef at Saigon Grill. There was only one problem: I had been dating a wonderful, intelligent, and beautiful classmate for a scant two months, but she was moving to Los Angeles.

Famous "optimizer hats" from cinematic history.

You can guess what happened next. I instinctively donned my “optimizer” hat and cerebrally cycled through a crippling cacophony of cases, miring myself mightily in mental mud. The more I thought about it, the murkier the decision became. What if we break up? How much better will life be without snow? How often will I get to see my newborn little cream puff of a niece? Will 24/7 streetside access to lengua tacos outweigh the always-dependable chicken and rice carts at 53rd and 6th? And so on and so forth. Every fiber of my being fiercely compelled me to carefully and methodically weigh every possible factor and determine the rational, optimal solution.

Of course, that type of analysis is entirely absurd and doomed to fail from the outset: it’s too complex and too broad, individual outcomes are far too unpredictable, there’s some implicit premature optimization … but above all, anything involving emotion is slippery and amorphous by nature. In reality, any meaningful decision of that sort and scale that doesn’t have an obvious answer a priori simply has to be made on gut.

Happiness can’t be linear programmed.

So my advice for the like-minded optimizers among you?  Do the analysis. We both know you’re going to do it anyway. Then, once your brain has exhausted itself, consult your family … your friends … your mentors … and I’m guessing the answer will be clear as a sun-soaked Los Angeles day.  For my part, I had the benefit of some fun advice, including a handwritten note from a dear friend on his wedding day:

I’d stick around to hear you guess where I landed, but I have a date with my wife and some lengua tacos.

Love the Ones Who Listen

The first time I performed as a street musician was one of the scariest moments of my life.

As I stood in the middle of Rittenhouse Square, guitar draped from my neck like a sandwich board and my fingers frozen to the strings, I could visualize streams of Philadelphians flowing by with nary a glance nor ounce of hesitation, their utter indifference a pendulum drifting inexorably downwards and cruelly sanding away my ego, particle by brittle particle.

I hadn’t even played my first note, and already I was petrified.

Then it struck me: yes, the vast majority of pedestrians would probably stroll by without a second glance, let alone a dollar tossed or a quarter flung … but every pedestrian that did stop would do so because they wanted to hear me.

Busking wasn’t self-inflicted humiliation … it was liberation. Unlike a concert, at which every attendee was irreversibly bound to their seat by decorum, these “concertgoers” could and would walk “out” on a whim. But the tiny minority that did stop and stay did so because they wanted to hear me play. It was in that wonderful, freeing moment that I stopped caring about all the hundreds of people that wouldn’t dream of stopping and started focusing on the ones who would and did ... and kicked off a few years of incredibly rewarding time performing on the streets of Philadelphia, New York, and Los Angeles.

Since then, I’ve tried to look at my life, family, and work the same way. It can be all too tempting to think about how some broader group will evaluate or judge you or some aspect of your life or work. Former classmates … distant cousins … that nasty coworker. In the final estimation, none of them will matter. Invest your time and energy in the people that care about you and want to see you succeed.

Optimize your life around the people that stop to listen … or even better, the ones that tip.

The Game of Telephone

Ever play “Telephone” when you were little? You whisper a phrase to someone, they whisper it to someone else, and so on and so forth until the last exuberant kid joyfully screams the phrase aloud in all its mangled, mistranslated glory and milk squirts out of your nose because you’re nine years old and life is freaking hilarious. Now imagine playing Telephone as an adult, but drawing from a decidedly less enjoyable lexicon filled with dry terms like “database schemas” and “EBITDA,” and having the game’s final result be money squirting out of your company’s coffers instead. Sound familiar?

In some offices, a “best guess” will be passed from person to person with increasing confidence until finally some invisible threshold is crossed and the next recipient takes what was once a wild, finger-in-the-air assumption as gospel. For example:

PERSON #1: “With these ludicrous PLACEHOLDER assumptions, we’re going to net $100M in annual gross revenue.”

PERSON #2: “Person #1 predicts that we’ll net $100M in annual gross revenue.”

PERSON #3: [hmm, I remember hearing the word “net” …]  “Person #1 is projecting $100M in net revenue!”

PERSON #4: “We’re guaranteed to do $100M in net revenue next year; Person #3 validated it.  Teslas for everyone!”

At Scopely, this sort of miscommunication was never an issue when we were a small team crammed into a shoebox of an office above the Wiltern Theater … and from my personal experience at medium-sized companies, it wasn’t as much of an issue since carefully-qualified statements were the fortuitous albeit unintended byproduct of cultures that incentivized covering one’s own ass and disincentivized devil’s’ advocacy. But for fast-growing teams worshipping the trite-but-true startup mantras of “Fail Fast” and “Ask Forgiveness, Not Permission,” the problem can be particularly insidious.

For what it’s worth, Scopely’s gotten really great at minimizing this sort of miscommunication as we’ve grown from 15 to 85 people.  While there’s not much you can do the closer you are to the “receiving” end other than be healthily skeptical, here are some tips I’ve picked up along the way for those closer to the “giving” end:

TIP #1: BE REALISTIC.

Assume Telephone will happen. Because it will.

TIP #2: BE EXPLICIT.

Particularly when putting things down on screen or paper, make sure to be explicit and mark assumptions with visual cues. Gross vs net? Linear vs log scale? Empirical vs predicted? Outliers excluded?

TIP #3: BE PROPORTIONAL.

The importance of the conclusions that will be drawn from the data should dictate the amount of effort you invest in validation or making assumptions explicit. If you’re putting together a shopping list to stock the communal fridge, I suspect you’ll be just fine if you wing it.  If you’re modeling profitability for a go-or-no-go contract negotiation, it’s worth a few extra minutes.

TIP #4: BE SELECTIVE.

People can only process so much data--present them with too much and they’ll tune out most of it, potentially drawing the wrong conclusions. You might also consider dumbing down the specificity of the data to produce a more generalized conclusion that implies a lack of granularity.  As Chris Stucchio puts it, take your inspiration from xkcd.

TIP #5: BE HONEST.

If you’re caught off guard and don’t have the luxury of time, tread carefully. If you’re forced to make a guess and find yourself succumbing to a cognitive bias (e.g. anchoring) or logical fallacy (e.g. argument to moderation), sometimes it’s better to simply say, “I honestly don’t know and anything I tell you will be a complete guess.” After all that, you may still need to produce a guess if it’s your job to do so, but you’ll have done all you can by instilling some healthy doubt.

TIP #6: BE SKEPTICAL.

Anyone that gives you a list with fewer than six tips is not to be trusted.