‘Experimentation’ is a process, not a policy.

franksA week behind on my reading due to work, I recently came across this article in the Economist. Of particular interest is this quote:

…In the early 1900s Frank Gilbreth, one of the pioneers of industrial psychology, tried to raise his 12 children according to Frederick Taylor’s principles of scientific management. He discovered that you could cut the time it took to shave if you used two razors at once—but then abandoned the idea when he found that it took an additional two minutes to bandage the resulting wounds…

The remainder of this column is about guruism in business. The column talks about jargon and the re-skinning of old ideas with new language. It calls into question the wisdom of naming outstanding companies only to watch them fall on their faces. And most interestingly to me, it calls out the lack of rigor behind many of the assertions of business gurus.

I’ve seen a recent parallel to this latter point in the common social media refrain, “Don’t be afraid to experiment! You must experiment! If you’re not experimenting you’ll miss the wave!”

This is a very convenient rallying cry for all gurus. It removes them from accountability if something doesn’t work (‘failure is part of the experimentation process’), and rewards them when it does.

Not surprisingly, a willingness to experiment is being held up as the must-have policy of any organization seeking to profit from social media. Of course, experimentation is never defined as a scientific process in this instance. Instead, it’s presented as published brainstorming. The logic is that digital media is cheap, if you have an idea, just try it. Try everything. Load that social media shotgun and fire away.

This may be acceptable to companies with unlimited resources, but for most businesses profiting from social media is one of a thousand items they need to focus their resources on at any one time.

My father is a polymer chemist. I grew up around scientific process and even though my career path  took me elsewhere (marketing and business strategy), I have always found myself relying on science in addressing business challenges.

Scientific process begins with an Objective. Companies are familiar with objectives so this one should be easy to understand, I’ll
move on…

Making an Educated Guess
Next comes the hypothesis. This is an ‘educated guess’ as to how to achieve the objective. How does one make an educated guess about social media? Given that it’s inherently social, I might start with how different types of people group and utilize communities for support, action, belonging etc. I might investigate how they perceive and define government, family, institutions like banks, and their own communities. By knowing the context of the lives of a target constituency, I can formulate a hypothesis as to which social media tactics might be useful in achieving an objective.

That said, when was the last time you heard the word ‘hypothesis’ used in a social media discussion? More to the point, when was the last time a conversation about social media began with human behaviors and community constructs instead of technology-of-the-week gadget jabber? Yet without a hypothesis – an educated guess – we wind up shooting in the dark and hoping for the best. It’s that simple.

Experimentation should be focused.
From the hypothesis the scientists then design an Experiment. This experiment is specifically constructed to test the hypothesis. It’s very execution is reliant on the needs of the hypothesis. In creating the experiment the scientists must apply rigor. They must create a control group to benchmark results. They must define the metrics that will prove the hypothesis and the means to test for them experimentally. They must consider all the details of the experiment to make sure it does indeed test the hypothesis. They must take all necessary precautions not to pollute their experiment with contaminants that might skew the data or deliver misleading conclusions.

Once designed, an experiment is conducted under disciplined conditions. Data is carefully logged and critically analyzed. A good scientist is looking objectively at the data, resisting the temptation to cherry pick results to prove the hypothesis. In fact, scientists are careful in drawing conclusions to properly attribute them to the causal reasons for the outcome. Social media experts, in my opinion, don’t always do this.

The goal of any experiment is to come to a conclusion that leads to a theory. To do so, the experiment must be a repeatable process that achieves consistent results. Again, social media pundits are quick to run to the press with a single success story and call it ‘proof’ of ROI or paradigm change. Scientists would be appalled.

This brings us to failure. After ‘you must experiment’ the next most popular  refrain among social media advocates is ‘you must be willing to fail’. This is often presented superficially as simply not being afraid to fail. And its true, innovation requires risk taking and penalizing failure will rarely get you to meaningful innovation.

But…

The scientific benefit of embracing failure is the importance of the data failure delivers. Failed experiments that are simply abandoned offer no benefit. If an experiment doesn’t work, one must endeavor to figure out why it didn’t work. Scientists, working toward a stated objective and testing a hypothesis against that objective, will use the data of failure to tweak the parameters of their experiment before trying it again. Any experiment that’s not worth pursuing after a failure probably isn’t worth doing in the first place. That means the objective and a hypothesis are critical preconditions that necessitate experimentation (vs. just including it as a corporate cultural policy). The necessity of the experiment then makes the data of failure valuable which is what gives meaning to ‘embracing failure’ as a cultural value.

In terms of social media, this means resisting the urge to abandon something that fails initially. someone should be thinking, ‘But if we keep doing failed experiments we’ll run out of money and resources to fund them.’ Precisely. Which is why the objective and hypothesis are so important. They will help sift through the many social media options to find those most worth experimenting with.

Experimentation and embracing failure are indeed necessary elements if business is to benefit from any new technology, idea or insight. But experimentation and embracing failure are not an excuse for stumbling forward recklessly with a shotgun approach to solving problems – in social media or otherwise. Simply glomming onto the technology of the week and firing off an execution against it is not experimenting in the most useful sense of the term. At best it is tinkering.

Without an objective, hypothesis and reasoned experiment to test it, the chances of success are minimal and failure, when it happens reveals no new insight to make future experimentation smarter. Conversely, the cost of failure in wasted time, energy and resources increases when using the shotgun method. Eventually, these costs outweigh the very benefits of embracing the values and policies of experimentation and willingness to fail in the first place.

One Response to “‘Experimentation’ is a process, not a policy.”

  1. jan zlotnick Says:

    Corey, thanks for all the amazing, fun, engaging, useful, inspiring ideas through 2009. Your stuff is the perfect storm with a big fat beautiful rainbow that keeps lasting for days…good holidays, enjoy, recharge, we’ll be looking in 2010.

Leave a Reply