How To Not Fail At Failing

The danger of unclear hypotheses and the step-by-step approach to reaching actionable conclusions.

Sisyphus was the King of Ephyra. He was an evil, conniving king. As Greek mythology tells, he tried to escape death from the gods and was ultimately punished for his misgivings. As his punishment, he was forced to roll a massive boulder up a hill. Every time he’d get close to the top, the rock’s weight would be too much for him, causing it to roll back down to the bottom. Up and down he went, for all of eternity.

One who performs a laborious and futile task that can never be completed is said to be Sisyphean.

In modern times, we revere failure, as a necessary step in the process towards achievement and greatness. As seen on your favorite guru’s Instagram:

Michael Jordan: “I’ve failed over and over and over again in my life… and that’s why I succeed.”

Henry Ford: “Failure is the opportunity to begin again more intelligently.”

Oprah: “Think like a queen. A queen is not afraid to fail. Failure is another steppingstone to greatness.”

C.S. Lewis: “Failures, repeated failures, are finger posts on the road to achievement. One fails forward toward success.”

Brené Brown: “There is no innovation and creativity without failure. Period.”

Richard Branson: “Do not be embarrassed by your failures, learn from them and start again.”

I think you get the picture.

If you’re reading this, chances are you’re currently failing at something. I hope you are. And as evidenced by the quotes above, not only is it OK, it’s necessary.

Yet, there are two kinds of failure in life — the Sisyphean, indiscriminate style of failure and a successful failure, which is the result of deliberate action and intention. It may feel like pushing a boulder up a hill is hard work, but it’s actually an easy way to fail, whereas the latter is hard.

It is here that we should reflect and ask ourselves in what way are we failing. Are we failing at failing?

Every opportunity for failure, considering the uncertainty of outcomes, is an experiment in and of itself. Some experiments seek to make a previously unknown discovery, whereas others may aim to demonstrate a known fact.

The scientific method tells us to start with a question or purpose, conduct research and then formulate a hypothesis, experiment, analyze and ultimately reach a conclusion. And while it may be more intuitive to use this process on an academic or business problem, we can certainly adapt and apply to personal situations, as well (after all, people are startups too).

When applying lean methodology, in particular, we are testing leap-of-faith assumptions to determine whether the solution to a problem is viable (from a commercial perspective). This requires that entrepreneurs test clear hypotheses and use actionable metrics to reach a conclusion — to pivot or preserve.

As entrepreneur and lean startup guru Eric Ries writes —

When an entrepreneur has an unclear hypothesis, it’s almost impossible to experience complete failure, and without failure there is usually no impetus to embark on the radical change a pivot requires.

…the failure of the “launch it and see what happens” approach should now be evident: you will always succeed — in seeing what happens. Except in rare cases, the early results will be ambiguous, and you won’t know whether to pivot or preserve, whether to change direction or stay the course.

According to Ries, when testing a clear hypothesis and using actionable metrics, a disproven hypothesis nonetheless leaves us a step closer to validation. Whereas the “launch it and see what happens” approach is failing at failing.

Consider a personal scenario — I’ve recently found myself increasingly distracted by my cell phone and wish to reduce the amount of time I pick it up as a method of procrastination. Ultimately, I want my phone to be less of a distraction, but beyond that, I haven’t set a clear hypothesis. So far, I’m trying a few different things out and just “seeing what happens”.

If I continue to do the same thing, indiscriminately, will I reach my goals? What are my goals, anyway? How do I quantifying (or qualify) “less distraction”? Are these actionable metrics?

Ultimately, the purpose of my experiments is not just to reduce distractions but subsequently increase productivity. So, let’s go step-by-step together through the scientific method so I can stop failing at failing.

Setting a clear hypothesis

My initial hypothesis, more or less, has been “If I turn off most notifications on my phone, it will be less distracting”. While it is a hypothesis, it’s not a very good one.

A good hypothesis should be specific and falsifiable. To qualify, it should specifically include the following components:

1. the change that you are testing
2. what impact we expect the change to have
3. who you expect it to impact
4. by how much
5. after how long

It also should be related to a specific experiment or process, not an overarching goal. As Scott Adams writes in How to Fail At Everything And Still Win Big,

If you do something every day, it’s a system. If you’re waiting to achieve it someday in the future, it’s a goal. [O]ne should have a system instead of a goal…In the world of dieting, losing twenty pounds is a goal, but eating right is a system… Goal-oriented people exist in a state of continuous pre-success failure at best, and permanent failure at worst if things never work out.

Turning off notifications on my phone is a great start. Testing specific and falsifiable hypotheses regularly is even better!

A great hypothesis in this instance may be —

If I schedule time throughout the day in which I am allowed to look at my phone and turn it on airplane mode when not allowed to be on it, I will reduce my screen time by 25% this week compared to the week prior.

What kinds of experiments should we run?

There are two kinds of research — evaluative and generative.

Evaluative research tests a specific hypothesis, like the one above. I know exactly what kind of experiment I need to run to test this hypothesis.

Generative research, on the other hand, may not test a specific hypothesis but instead may help identify a problem which is worthy of developing a hypothesis to test.

In our personal scenarios, such as my iPhone distraction situation, we clearly know what the problem is, so we’re going to run generative experiments that test our specific and measurable hypothesis.

There are many experiments I can run to test solutions to my iPhone distraction — I can put it on airplane mode; turn it off; leave it in my bag or the car or at home; message everyone who I know needs to reach me and notify them that I am only available by phone at a certain time; turn my phone on grayscale; give my iPhone away and buy a flip phone.

However, it is important, as we’ll see, that we are conducting experiments that are measurable, and generating the right kind of metrics.

Are we measuring the right things, in the right way?

In The Lean Startup, Eric Ries distinguishes between vanity metrics and actionable metrics. Vanity metrics may look or feel good, but do not actually bring us any closer to proving or disproving our hypothesis. Actionable metrics, on the other hand, must demonstrate clear cause and effect.

In business, vanity metrics are things like pageviews or app downloads. They may be trending upwards, but they don’t necessarily inform subsequent action.

Similarly, in my iPhone situation, I may reduce my screentime by 25%, but what does that tell me? If reduced screentime leads to greater productivity, then that’s great! However, if despite the iPhone specific experiments I run, if I fill the void with another distraction, perhaps I was measuring the wrong thing in the first place.

The purpose of my hypothesis is ultimately not just to reduce the amount of time that I’m distracted by my iPhone, but to achieve greater productivity. Reducing screen time is a testable hypothesis within my larger personal development objectives, just as conducting A|B tests to verify a specific (and falsifiable) hypothesis fits within the larger business objective of increased sales.

Pivot or preserve?

Ultimately, after testing hypotheses, we must make a decision to pivot or preserve. To preserve is the easy route — so often the status quo bias and inertia keep us on the same perpetual path. But as they say, the definition of insanity is doing the same thing over and over and expecting different results.

The Lean Startup speaks of the requisite courage necessary to pivot, just as it takes courage to make a change, take a risk and venture into the unknown.This decision is made more difficult if and when vanity metrics are telling us what we’re doing is working. Yet, actionable metrics more clearly demonstrate the necessity to pivot, and embolden us, whether in business or in life, to try a new experiment.

Are we really failing?

There’s another category of inspirational Instagram quotes — those that argue that failing is not really failing.

Thomas Edison: “I have not failed. I’ve just found 10,000 ways that won’t work.”

Susan B. Anthony: “Failure is impossible.”

Tony Robbins: “There is no such thing as failure. There are only results.”

These takes are instructive, as this mindset undoubtedly helps address the cognitive considerations — and sensitivity — surrounding failing.

Is invalidating a hypothesis a failure anyway? I would argue no — after all, what makes a good hypothesis in the first place is that it’s falsifiable. In other words, it’s supposed to fail.

So, when we’re failing at failing, we’re actually failing at trying.

I recognize that it may not be entirely practical to hold our life experiments up to as much scientific scrutiny and measurement as described above. If something is working in our lives, we should be doing more of that. It doesn’t always require extreme rigor to make personal improvements.

Yet, if there are two types of failing, there are two types of trying — Sisyphean style, and with deliberate intention. Aspirational individuals understand that through deliberate experimentation, risk-taking; pivoting and change; through actionable metrics compelling actionable decisions; through so-called failure — the good kind — we can create the outcomes we seek to create, and become the versions of ourselves we seek to become.