Link to the original article by Forbes

The key matter of discovering meaningful data is the right approach in carving out findings. Let’s be honest here, in every profession, there are deceitful practices and there are authentic ways.

Naively, I believe the larger group of people are trusting their datasets as it is and accepting the output of the analysis, for better or worst.

“Any tool can be good or bad. It’s really the ethics of the artist using it.” – John Knoll –

How to reduce bias in data

To help reduce biases/cherry picking data within your company, here are some helpful (and ethical) practices:


Ethics, Integrity, Honesty – Any organization that carries values such as high integrity, great ethics and truly honest, does not have to massage the dataset till the dataset reaches conformity with whatever plan that was set out prior.

Falsifying the outcome of the dataset is meaningless and pointless and one should not even kick off with a data-driven initiative if the intention is to deceive. I mean, really, what is the point here. If the dataset analysis is aligned then that strengthens your case.

However, if there is a wallowing gap, that just means you have opportunities to address the gap. Either way, you are guided to make a better-informed decision.


To stay guided and not misguided, there are a few areas to observe;

  • Understanding the Data
    It is not necessary to mass harvest everything and anything under the sun. The reason is, it’s impractical to do so. Therefore, the sampling method is an acceptable and approved research method. It is also found that an excessive dataset with irrelevant noise might lead to incorrect findings while a cleaner and smaller quality dataset sample will bring greater insights.
  • Knowing the Target
    It is important to chart out the desired and intended achievement. With each and every dataset offered to you, it is unnecessary to reach out to every space. Set a clear target audience, a distinctive distribution or channel tiering and identify green and mature markets. Knowing this would set a guide on what dataset to focus on.
  • Know Yourself
    Before promising growth and extension, you ought to know the current standing. For any data-driven analysis, the very first to-do action is the preliminary findings – Study the numbers. You need to know the maximum, the lowest, the average, the mean, etc. Once you progress with the execution, any deviation from the initial findings, good or bad, shall be a guide for the next course of action.

Media Insight Agencies

It is not the end of data-driven analysis if all of the above fails. You should be able to hire independent media insights experts to assist you. Being an independent party, the data-driven analysis is presented with true hard facts, no sweetener but just tough love.