Wednesday, March 16, 2011

Learning from Your Mistakes - by the Numbers

There are good lessons in mistakes and “natural experiments” in events and across the media business. The lessons pose both emotional and conceptual challenges. That’s why running the numbers right is highly instructive.

Mistakes in the moment in any live media generate tension – you run low on food, cues are missed, wiring is wrong, the award that was supposed to be at the front of the room was left in the back – whatever. Somebody miscalculated, didn’t practice, didn’t plan or didn’t execute the plan.

When you are in the hot seat, fixing the immediate problem, it’s hard not to get a little curt, and it’s easy to blame others. This complicates the after-action analysis. Specific detail and quantification makes it less personal.

Bigger mistakes usually manifest more slowly. Marketing deadlines get pushed back for what seem like good reasons. Key indicators are ignored, or erroneously interpreted; your product drifts off the intended focus or markets change and leave your media product positioned for yesterday. Numerical discipline helps you recognize slow but persistent change.

“Natural experiments” are often hidden in benign variation. Business is better in some regions then others. A marketing effort does a little better or worse than average. An audience responds more positively in one instance than another. Numbers reveal opportunity.

While the race is on, the best response may be counter intuitive. If you are spinning out of control, sometimes you have to “gas it” to recover. In any case you may not have time to think much. Training is necessary for times of deep crisis.

However, trusting the numbers during a crisis or when planning your next move is often the best way to guide yourself out and learn from the mistake or natural experiment.

Often mistakes involve not trusting the numbers out of fear or wishful thinking. The most common scenario is judgmental adjustments – a little padding here and there for safety or a gratuitous allowance for a shortfall. These judgments are often repeated and compounded, killing margins and driving actions off track.

Natural experiments are situations where inputs are varied unintentionally. They can go unnoticed if you always look at overall numbers and not the details. And variation is often misunderstood. Causation and correlation are easy to confuse. Human nature and human interest get involved, e.g. successful salespeople see themselves, not their assigned accounts as their source of success.

You can’t assume accuracy in your numbers. People code and count things differently. Performance numbers get “gamed.” Conversions and aggregations introduce error. Things get mislabeled. Time periods don’t match. It pays to test any numbers you use until you are sure what they mean and how reliable they are.

But, especially where things go wrong, or results vary, your best source of new tactics and strategies to prevent failure, replicate success, grow or maintain your business and make it more rewarding and less stressful is in creative, quantified analysis of the facts.

Labels: , , , , , , , ,


Post a Comment

Subscribe to Post Comments [Atom]

<< Home

Powered by Blogger

Subscribe to
Posts [Atom]