OMG Ponies

Bullshit, Intuition, and Organics

Working in software in 2016 is to be buried in a mountain of bullshit so large you can see it from Mars. People don’t really like to admit it, but bullshit is what keeps the money flowing. Some people call it playing the game. Some call it politics as usual. Some call it a necessary evil. I call it bullshit.

Bullshit has a familiar smell. It smells like VCs, growth hackers, VPs of Global Sales, Ninjas, Rockstars, and Unicorns. It smells of over-confidence and hubris. It smells of results that seem like they’re just too good to be true. And it reeks of short-term thinking and lack of patience.

Every Venture Capitalist I’ve met firmly believes they know how to pick winners. They’ll look at a product’s financials, run it through some spreadsheets on their end, and declare that they’ve predicted the probable market size and penetration over the next ten years. But it’s secret sauce, and no, they won’t share. But they will happily dispense their advice on _what to do next_. Because they know with certainty that an influx of capital and their advice are the only thing your business needs to succeed.

But if you start to measure how well VCs are performing — measuring the _reality_ — the success of their portfolio companies in the public markets — you’ll see that VCs have not been very successful in technology. For every Facebook and Amazon, there are dozens of Zyngas and Groupons. These companies that were declared winners are failing on every front in the public markets. This hasn’t harmed VC’s returns because their success is mostly dependent on the structure of our financial system, not on the performance of the products they serve as advisors for.

VCs depend on the most common form of bullshit in technology — the confusion of correlation for causation. Venture Capital claims it knows how to run products because of it’s financial returns. Yet, our entire system is designed such that VCs often make (lots of) money on failed products. Acquisitions, IPO structures, and big late round investments prop up their returns, but say nothing about the success of the _products_. You’ll notice VCs rarely talk about the _revenue_ of their portfolio. It’s too embarrassing.

Do you think Facebook would have been less successful without VC advisors? Do you think a different set of investors could have saved Zynga? VCs firmly believe they have the power to save companies. I suspect the companies would have performed just about the same, regardless of what happened in the board room. Facebook was always going to succeed because its users love the product. Zynga was always going to fail because its products were built around extorting its users.

This idea — the idea that someone in the board room can have massive effect on a product’s success — has trickled down from VCs over the years. VPs, growth hackers, and unicorn hackers all believe that _they_ are responsible for their product’s success.

**This is bullshit.** It’s not so much that they _aren’t_ responsible for the success, it’s that we have no way of knowing what the cause of the success was. It could be that people with a history of success have just been lucky and picked companies that were always going to succeed. Our sample sizes are far too small and our measurement windows far too short. We’ve been making steel for over two thousand years and we’re still figuring out new ways to make better steel. Facebook has been open to the public for ten years.

We don’t really know much about software. We haven’t been building it that long. Its likely that the reasons some products succeed and some fail are completely unknown to us. We have observed that some techniques tend to work better than others, but our sample size is incredibly small, and our measurement windows pitifully short. How many software products have been successful for longer than a decade? How many founders have started more than three successful companies?

// scurvy

My Stepmom has owned an art gallery & frame shop for over thirty years, and one of my favorite stories from her is about paint. We take for granted today that you can paint something on canvas and know it will be around two hundred years later. But it wasn’t always that way. It took millennia to discover painting materials that would last that long. For ages and ages, artists would create art, and it would fall apart long after their death. The artists thought their painting would last forever — they lasted their entire life. But the success of painting materials is measured in generations, not lifetimes.

The moral of this story is that it’s easy to declare success, but difficult to know how lasting it will be. The world works much slower than one might anticipate. A lot of failures look like successes in the short term, but when viewed on a longer time scale are obviously failures.

We learn to smell bullshit because we want to build successful products that last. Too often my colleagues have taken bad advice and ruined their products because they’ve believed in shaky claims. That’s bad for everyone. We need to be more critical of extravagant claims. Look at it from another angle. How are they defining their terms? What is their definition of an an active user? What assumptions were made for this analysis? How did they measure long-term effects? What measures did they take to identify unintended consequences? When people say things like _retargeting always increases sales_ — my bullshit meter goes through the roof. Retargeting has only been around a few years, there is _no way_ anyone knows the long-term effects of its use. There’s _no way_ all audiences react the same to it. It’s a claim that’s too good to be true. Whether the claim that _retargeting always increases sales_ is true or not — I don’t know. But I know we can’t be confident that it’s true.

Once you’ve practiced smelling bullshit long enough, you’ll start to see some common themes. You’ll see that a lot of people use the same terms — like _daily active users_ — to compare products, yet every product defines a _daily active user_ as something different. You’ll see that people almost never measure long-term effects or know what assumptions were made in their analysis. You’ll see that a lot of our industry’s best practices rarely stick around for more than a couple of years. And you might start to suspect that we don’t really know as much as we say we do. You might suspect there are wholly different approaches that might outperform current best practices.


Technology has an intuition problem. Few people trust their intuition or even believe intuition to be a valid source of reasoning. In our data-driven world, we have convinced ourselves that intuition is a magical talent unbound to the laws of nature.

But intuition is not magical. It is a skill like any other that can be practiced and mastered over time. It’s a vital skill that many professions rely upon — firefighters, military personnel, rescue workers, extreme athletes — and a skill that humans require to function in this world. If you did not trust your intuition, you’d be a useless lump of meat. You’d never be able to pick up a fork, speak a word, or focus your eyes.

Intuition is our ability to immediately understand something without being able to consciously reason about it. It’s what allows us to vibrate our vocal chords at a certain pitch without knowing the frequency of the note we’re signing. It’s what allows us to jump across a gap without knowing our specific mass or the acceleration of gravity. Our intuition allows us to solve complex problems without being able to understand them. And that’s an important skill in a world filled with far more variables than we can solve for.

It’s always felt like people who work in technology act a bit like indentured servants. For as long as I can remember, there’s been a majority of designers that implement ideas they believe to be bad. We agree that users hate aggressive ads, but we implement them anyway. We agree that mouse-trap signup flows are bad, but we implement them anyway. We know that collecting too many personal details is dangerous, but we do so anyway. We know that users don’t want us spying on them, but we do it anyway. We know that user’s don’t want aggressive engagement emails, but we send them anyway. We know that many of these growth hacks feel unethical, but we ship them anyway.

You might think we’re being commanded to build these features, but in my experience it’s more likely people believe these bad ideas are somehow actually good ideas. Every ounce of their intuition is screaming not to do it, but a single shred of shaky analytical evidence and they’ll throw their intuition by the wayside. In technology in 2016, there is no more powerful church than that of analytical reasoning. Big data, analysis, graphs, and charts are seen as the one and only truth.

We’ve forgotten that the world is filled with millions of variables and that our analytical reasoning sits on a tower of assumptions. When I was doing Civil Engineering work, I had to state my assumptions down to the _direction and acceleration of gravity_. Analytical reasoning requires extreme diligence. A diligence not often found in product design.

In other professions, it’s common practice to rely on intuition when faced with highly variable situations that benefit from quick decisions. When a firefighter enters a burning building, they have no idea how the building is going to react, and if they want to save lives, they need to make split second decisions. Firefighters rely on their intuition to stay alive and save the lives of others. No rational person would rely on magic to do that.

But product design is a perfect match for intuition. Humans using computers create so many variables you might think they’re breeding them. We’re prone to irrational moves that escape conscious reasoning. We don’t make sense. And if there’s one thing we can actually agree on in software, it’s quick decision making is vital to successful product development. A quick bad decision is often better than a delayed good decision. Ship early, ship often. Iterate, iterate, iterate.

So why do we undervalue our intuition so much? I think it comes down to three challenges:

* Intuition is trained through meaningful experience, and it’s difficult to find meaningful experience in product design. Few people have the chance to see a successful product through from conception to stable business. Designers don’t know how to train their intuition, so they do not view it as a skill — they view it as a talent.

* We lack the skills or tools necessary to communicate intuitive reasoning. When asked to defend our intuitive solutions, we struggle to detangle the complex interactions happening in our brains and devalue our reasoning by calling it a gut feeling.

* Product design vocabulary has conflated *intuitive* with *obvious*. When most designers describe an interface as _intuitive_, they most often mean _obvious_. Our profession expects all intuitive reasoning to be immediately obvious to everyone. But our intuition is a skill and defined through our experiences, and there is no reason to expect everyone’s intuition to come to the same conclusion.

My intuition tells me that our industry approaches product design in an overly stressful way. We employ labor-intensive, questionably ethical, inorganic methods in the name of growth, while abandoning time-honored traditions of treating customers with respect and service. Modern product design puts too much emphasis on our role in a product’s success. We don’t trust organic processes to work on their own. We don’t trust our customers to grow our product. We have the hubris to believe that we are masters of our users, and that our decisions are the reasons our product succeeds or fails.

But intuitively, that philosophy smells like bullshit. My experience has led me to believe that there are much easier ways to approach product design. Methods that require much less effort, almost no stress, and leave you feeling like you’re treating your customers well. Methods that will immediately seem right, despite being unable to consciously reason about them.


My style of product design relies on organic methods. When we talk about organic farming, we talk about farming in harmony with nature and the environment we’re working in. A truly organic farm leaves the land more productive than it was the year prior. Similarly, organic product design is design that works in harmony with human nature and the world we live in. Organic product design declares that it matters how we treat our customers, and that treating people with respect and integrity can improve the world we live in.

Selfishly, I enjoy organic methods because it’s a less stressful way to work. Instead of feeling like you’re constantly having to prop up growth with continued effort, you design a system where customers grow your product for you. That means more vacations and less stress. It means more profit with less investment. It’s just a better way to work.


About this entry