An Experiment Gone Useful

Artypapers Homepage

About Artypapers

Factoids
- So far there are 5,299 artypapers members, 56,109 scribble links and 251 CSS Help Pile posts.
- Become a member today!
Fresh Links +
Features

Valuable Coupon Offer

Free

Web-based job tracking, invoicing, reporting & project management software for the part-time independent contractor.

Get Your FREE Side Job Track Account Today!

There’s this idea that anything can be represented by an equation that produces a number. The newsworthiness of random information. The love between two people. The strength of a familial bond. A specific person’s curiosity, the very boundaries of their imagination. Nothing is off limits, nothing out of reach. It’s dedication to the precision and accuracy of those numbers that has had a critical role in recent mega-IPOs like Facebook and Google. Acting in concert, those numbers drive the companies that in turn have defined the very interface of the consumer internet itself: An algorithmically ordered list. That list may be of stories or search results or friend suggestions or ads, it doesn’t matter.

These companies have recursive virtuous scaling because the bigger they grow, the more data they have to refine and improve the equations generating those numbers. The better sort of the list of options develops increasingly well understood and valuable behaviors. It’s not easy to copy, either. If you're a number company all the work to generate those numbers is so distributed throughout the organization, no single employee has too much power. This is why there’s no real risk when the head of feed or search goes to a competitor, it’s simply not possible for one human to contain the entire ruleset within their head. They understand important chunks and strategy, sure, but not everything that goes into creating the best order in all contexts. Even if someone managed to surmount that impossible mountain of information, the product they apply that information to would be different[1][2]. Different network, different humans, different motivations.

Not all companies care about numbers in the same way. Apple, for example, cares about experience. The unboxing experience. The experience of holding an object in your hands. The unique experience of software crafted specifically for a particular hardware. Snap, Airbnb, Stripe, Uber, each try to set themselves apart based on end to end experience which helps build brand[3]. These companies are ostensibly easier to copy because their complexity is hidden under outrageously unreasonable amounts of hard work, focus, and determination. The way they fight off competition is by moving quickly to capture marketshare, by successfully predicting the future, by obsessively going ten times beyond what's reasonable and then going another hundred more just for kicks[4]. They aren’t explicitly trying to reduce abstract concepts or human behaviors to a number that can be sorted and compared to another number objectively. The subjective experiences they create can be measured and those measurements compared, but the aim is always the experience, not the measurement itself.

Number companies tend to fail to recognize what they are. Experience companies tend to fail to recognize their internal definition of “unreasonable” isn't. A company understanding its nature is critical to its execution in context over time. Because that nature drives all manner of decisions big and small, from what people to hire and the relative importance of various roles to core product strategy. Look at how Netflix suggests movies. Their determination to give their customers the very best ordered list of movies revealed to them the value of creating their own content. Netflix doesn’t care about the art of film or anything like that, it's beholden to making the algorithms better and better and that knowledge guides what gets filmed. This understanding will have tremendous impact on the future as folks grapple with the popular acronyms of today. AI (ML, NLP, etc.) VR AR… these are all technologies in search of a use case, in search of a market.

An open question is: If the UI for the consumer internet has thus far been an ordered list, what, if anything, replaces that list? Could anything possibly? That's a big question for number and experience companies alike but it’s a far more pressing concern for the number companies that are bound to its fate. Conversational UIs were the first best guess. Voice is currently the strongest contender at this point. When you’re a number-driven company, the obvious future is no longer producing an ordered list from which to choose, but to provide the singular discrete answer[5] via some form of artificial intelligence. That is an incredibly daunting task for a market that may not actually materialize. The real question for Google and Facebook is not: Can AI become more advanced and human-like? But rather: When do humans actually want to talk to a computer as opposed to another human?

This is where experiential-focused companies have an advantage because they aren’t bound to a strict interface dictated by a foundation based upon numbers powering the order of a list. And it’s not because of any inherent value of experience over numbers, it’s that as of right now there’s no obvious next step. This ambiguity is pure opportunity. The next punctuation in the equilibrium is on the horizon and it will lay a foundation on which the next generation of companies will be built. Like mobile identity (attention) before it, the fundamental truth powering the next generation of technologies won’t be obvious and it possibly won’t be apparent until looked at in hindsight. Whoever discovers it, however, will have the run of their lives ahead of them.

________

[1] You see this with high-profile name brand hires all the time. Why are these people unable to single handedly recreate success of company A at company B? Is the person not as good as advertised? Maybe, but also because company B is not company A and the same rules don’t apply for company B’s product. I suppose if the name brand person were truly great, they would see that and adjust but company A was never the creation of a single person.

[2] This is why Instagram had to sell to Facebook. Without direct access to Facebook’s years of experience scoring images, Instagram may not have been successful in transitioning to a number company after they ran out of gas in terms of innovation. Filters were great when phone cameras were weak, but now that they’ve improved? What’s the next innovation there? Instagram would have had to answer that question on their own and given how small that company was at the time of the sale and the expertise of the founders, they were not in a good position to do so.

[3] If the #deleteuber cycle is any indication, Travis Kalanick’s resignation from the advisory council is a perfect example of the importance of brand for that company. There’s no magic number they produce, there’s a feeling—and this feeling is true for all sides of the marketplace; for riders and drivers alike. If that feeling is too closely aligned with a bad feeling, that’s going to hurt business. Self-driving cars can solve for one side of the market when they’re ready for prime-time, but not both, and Uber needs to survive long enough to see that future.

[4] Snap—as noted in its S-1—understands the need to keep pushing beyond what’s reasonable and to keep innovating. Facebook has basically ceded the future to Snap by not demonstrating any ability in terms of innovation for a very long time. What helped Facebook kneecap Twitter, will allow Snap to slip through its fingers. Because unlike Twitter, Snap knows exactly what kind of company it needs to be and while its metrics may sound similar to Facebook’s, its methods and what it’s aiming to produce are not. Facebook is Facebook because every other major competitor committed suicide and they scavenged those remains for the best bits, not because it was some product visionary. The thing with photocopying an innovator is that by the time you’re happy with the copy, the original creator has already learned more about their product and moved on to something new based off those learnings.

[5] However, voice, like text, has no real facial expressions, no subtly in its ability to communicate. The voice “emojis" will probably initially be a standard set of tones programmed into the artificial voice. Why have the AI doctor say it's 88% confident in its diagnosis when you can have a confident tone in its speaking voice?

Execution in Context Over Time

essays - 0 Comments

Dec. 30, 2016

What is the difference between something that is world changing and something that is a change in the world? Execution in context over time.

Execution is how directly you can leverage the right tools toward some material, meaningful, and fundamental aim. The measure of the execution’s impact is related to how effective those tools can be in the present context. Over time, as the context shifts from other changes in the world, the execution is likely to require some adjustments in order to relate most directly to that new context.

Here’s an example: In early 2010 I, like many folks, thought there was an opportunity for a privacy-focused social network. But if you were to look at the obviously privacy-focused apps that were launched around that time—Path and Google+ being the most hyped—nothing really took hold. That is, of course, until Snapchat’s launch in late 2011. Both Path and Google+ took a reasonable, straightforward view of privacy. The content that people wanted to share was static text, photos, and video, and folks just wanted to share that same content with a different, smaller set of people. The broad expanse of Facebook was too cluttered with “friends” so why not create a new space with real friends like Path? Or, categorize and label existing contacts (note: not people) properly like Google+’s circles? Problem solved, right? Neither Path nor Google+ leveraged the right tools and both failed to address the current context most directly. Namely, that for better or worse Facebook was good enough at people-centric privacy.

What Snapchat revealed, however, was that there was room to innovate in terms of privacy but it wasn’t to be found in any straightforward view of privacy that only saw so far as the clumps of people standard content was distributed to. Rather, privacy could be viewed in terms of the content itself. The right tool in this case was a temporal primitive assigned to every bit of content and the notification when that primitive was violated. This created a completely new view of privacy applied directly to the content. Privacy wasn’t only about the network of any specific group of humans, it was about the network that got created when there’s trust that the content itself has strict limits[1].

Another social network that grew explosively just before Snapchat launched was Instagram in 2010. Instagram had almost no privacy controls whatsoever due to borrowing Twitter’s username, public default, and follow model. The early photos posted to the network were of landscapes and buildings, not people. But the speed and ease with which you could access the phone's camera and the filters compensating for the early iPhone camera’s poor image quality—those tools flourished in the context of the abysmal speed and layout of the Facebook mobile app at the time. As Facebook pushed the world toward openness, Instagram shifted too, and more people were willing to post photos of themselves and of other people more publicly. Filters that once made sunsets look their best were changed to those that would make people look their best.

This was a simple example, but ultimately, very complex when you think about it[2] because what Facebook, Instagram, and Snapchat seem to prove is that there are opportunities for pseudo public, public, and private social networks to succeed at the same time… for now. As each network grows its internal context will shift and change as well as the context set by external networks as time progresses. There is no clear winner, only an ever-changing list of survivors. Losing is just a failure to adjust and relate to the present context most directly.

Looking forward, thinking through a product's execution in context over time is the least amount of work one can do to judge its potential impact. What’s its aim? Is it something that is fundamental? What tools does it use? Are those the right tools to work toward that particular aim? How can those tools be best leveraged in context? How directly are those tools applied? How flexible are those tools to shift with changing context over time? Which tools should never shift?

________

[1] The sexting that early Snapchat was known for was an asset because 1/ what’s more fundamental than fucking?, and 2/ if you trusted a product with matters that private, surely you would trust it with the more mundane aspects of your life.

[2] Here is an actual simple example: Food delivery companies are easy to pick on. Overfunded, overhyped, and, ultimately, underwhelming. Ostensibly they should be good businesses but none has yet to succeed. Why is that? They meet Ev Williams’ basic outline for a billion dollar business: human desire that’s been around for a long time (eating food) and make it more convenient (stay at home, select options from phone). You could also argue that eating food is fundamental truth as much as it is a human desire. So, what went wrong?

The execution in context over time. And not the execution in terms of the app design or branding or whatever else, the end to end experience is lacking because of the fact that not all food travels well and deep cultural expectations surrounding eating as a ritual. There’s a reason pizza and Chinese food restaurants have had their own delivery services for years and years and fine dining hasn’t: the first pizza delivered on any given run survives about as well as the last. Carbonara, as it's currently prepared, never travels well. French fries—unless they are the ultra processed fries from a fast food establishment (and even those are hit and miss)—don't travel well, either. So why does quality beat convenience in this case? Because of cost. Because of personal preferences. Because of the weather that particular day. Because of context. Humans require food to survive but its abundance made it a commodity and its requirement in our daily lives gave it cultural meaning. Do you really want to serve your guests lukewarm so-so enchilada or would you rather meet at the restaurant over a pitcher of margaritas? Even when food is warmed, perfectly packaged, and delivered by robots—instead of sad underemployed delivery people who have insane quotas to keep—the experience will still be worse than a restaurant.