Breaking the Stranglehold of Prediction – Jaffer Ali

I have been in business a rather long time…long enough now to be called an "Oldtimer". I started in my own business right out of college WAY back in 1979. I was trained in the business mythologies of the day. I believed most of them back then. But reality has a way of crushing certain myths if we allow ourselves not to be blinded by them. But breaking out of the cave is not easy.

When I gravitated toward direct marketing, the myths proved overwhelming because they made so much "sense". Predicting behavior (AKA response rates) was the overriding goal. In MBA speak, "visibility" was the objective of every business. This meant we worked very hard to eliminate uncertainty. Uncertainty was/is an anathema to most plans. There has never been a shortage of methodologies to sell to us over the years created to propping up the predictive meme.

I believe that the way forward in our overwhelmingly fast-paced environment is to embrace uncertainty and ditch prediction as much as possible. Not only is our environment fast paced, it changes direction at a dizzying pace. Ask Sears and the Boston Globe if you doubt me on this point.

Big Data is just the latest term for what is essentially the same theme. Big Data promises to tame uncertainty and actually help us predict the future. And if we are to be honest, we can see the glaring failures of Big Data. The NSA collects plenty of data, yet according to the deputy director, maybe one terrorist attack had been detected in advance. (He also admitted that this was not certain either)

But the purveyors of Big Data misdirect us from the failures. Most of the trades accommodate the misdirection. Why? These charlatans buy advertising from the trades. This is a working hypothesis.

I am currently running a Big Data test (I continue to test things in which I do not believe) and am getting a CTR of .35% and am being told that this is really good…even though we still lose money. For anyone not good at math, the predictive algorithm is failing 99.65% of the time… So we have an algorithm failing 99.65% of the time AND we are losing money, yet are being told that the results are really good. I am not making this up. But I digress.

The most beneficial thing we have done as marketers is to develop tactics that do not require us to be right all that often. The cost of failure has been driven down to allow us to keep our hook in the water. Keeping our hook in the water with good enough bait eventually lands a big fish or two for us. I do not know if this can or will work for anyone else, but I thought I would share the notion. It is not a terribly new idea, but neither is the mysterious, dark art of prediction and that does not stop fools from rushing in. Anyway, I spend most of my time these days trying to find ways to break the stranglehold of prediction and embrace uncertainty. It is working for us and making me a lot happier.

Jaffer Ali is CEO of PulseTV.com and Vidsense. Jaffer has been a pioneer in online media and marketing doing both since 1998. PulseTV.com is an e-commerce site that utilizes the power of video todrive sales. He is a rare blend of theory and practice. He has written over 500 essays ranging from politics and media to philosophy and marketing and delights in actually implementing the things he writes about. A collection of essays from Jaffer are available in book form and can be downloaded for free for both the Kindle and Nook:
Chasing String In The Digital Era – Kindle free version

Read all Jaffer’s MediaBizBloggers commentaries at On the Other Hand….

Check us out on Facebook at MediaBizBloggers.com
Follow our Twitter updates @MediaBizBlogger

The opinions and points of view expressed in this commentary are exclusively the views of the author and do not necessarily represent the views of MediaBizBloggers.com management or associated bloggers. MediaBizBloggers is an open thought leadership platform and readers may share their comments and opinions in response to all commentaries.

[Image courtesy of cooldesign/FreeDigitalPhotos.net]