Skip to content

Mass Personalization Is Coming. Are We Ready For It?

2014 January 29

The modern world can be a dehumanizing place.  Long gone is the sweet little old lady at the drugstore counter, replaced by big box retailers, brand logos and barcodes. We’re more often“handled” than serviced, calculated, rather than cared for.

That’s about to change in a big way.  When I recently spoke to Bernie Meyerson, IBM’s Vice President of Innovation, about trends for the next five years, he repeatedly stressed personalization as one of the most important things the company is working on.

Yet, instead of the sweet little old lady behind the counter who has known you for years, the new personalization will come in the guise of a stranger armed with learning algorithms.  That’s quite a bit different and not without its problems.  As big data opens up a new world of possibilities, we’re going to have to come to terms with what we really want.

Are You An Outlier?

In the industrial age, mass production led to mass marketing.  Rather than cater to individual clients, corporations learned to create products that would suit most people.  Marketing became a game of statistics and firms learned to see consumers like this:

Normal distribution

They would shoot for the middle, knowing that their product would suit two-thirds of the market very well and 95% of the market to some extent.  It wasn’t perfect, because that left quite a few people underserved, but if you wanted to ring up big sales, you shot for the meaty part of the curve.

The problem is that none of us are average.  We’re all unique in our own way.  This hit home to me when I was living in Eastern Europe and every time I went on a weekend trip, my credit card would get blocked.  Even fairly ordinary activity for me looked strange to my American bank.

Marketers tried to cope by identifying niche markets with their own statistical curves of probability, but it was a blunt tool at best.

Personal Patterns

With technology getting cheaper and more efficient at exponential rates, firms are able to store and process an amazing array of information about us and, as we are increasingly sending them a continuous stream of information through the Web of Things, marketers are learning to see us more like this:


This is obviously more complicated, so I’ll explain.  Imagine you do three things: work, sleep and eat.  By monitoring you, it could calculated that if you just woke up, there are discrete probabilities that you would then work, eat or go back to sleep.  By adding just a few more variables, such as date and time, this model could be made fairly accurate.

Of course, the model I just described is an exceedingly simple one, but today’s computers have the ability to work with hundreds of variables and work in real time.  So, in the future, my bank will be able to look at my own habits to determine suspicious behavior and also know that if I just stepped off a plane, those habits are likely to change.

In effect, marketers are seeing us less as inert data points on a distribution curve and more as unique sets of patterns that can be identified, analyzed and catered to.

Sharing More Information Than We Think

We are all constantly giving out signals to the world.  If we go to a fancy restaurant, we will pay a lot of attention to how we will dress.  Some of us will dress up in order to look like we belong, while others will dress down to that show we don’t care.  If we’re angry and want to complain, we’ll be sure to alter our tone of voice in order to display our discontent.

Yet technology can decipher signals that we aren’t even aware of.  Mattersight is a company that has developed software that can analyze your personality during a routine customer service call.  In his book, Honest Signals, MIT’s Sandy Pentland describes a machine that can predict behavior from subtle physical cues.

So we can imagine that in the future when we walk into a store, facial recognition software will recognize whether we’ve been there before, analyze our buying history, predict whether we are there to buy or browse and alert a salesperson who has a compatible personality to wait on us.  All of this technology already exists in one form or another.

Clearly, this would improve service.  It’s hard to see a store with that kind of capability refusing to show Oprah Winfrey an expensive bag.  Yet how much do we want strangers to know about us?  How much control should we have over our data.  Do we have the right to see our own financial and psychological profiles?

The New Global Village

In the 1960’s, Marshall McLuhan argued that the inevitable consequence of information technology would be a global village where our consciousness extends out and the world looks in.  He considered the shift transformative, but value neutral, where we all become closer as a human race, but also more tribal.

There is a quaintness to villages, but they also chafe, with petty gossip and rivalries that can feel restricting.  While we take comfort in other people, we take shelter in anonymity. It’s true that “we all want to go to a place where everybody knows our name,” but we also like to leave.  In a world of mass personalization, that may no longer be possible.

Joelle Kaufman, Head of Marketing at BloomReach, a company developing personalization solutions for e-commerce, believes that the key to managing the conflict is through value and transparency and that it is incumbent upon merchants to give consumers control and ensure that they collect the least amount of information to give us what we want.

She may be right.  Credit ratings agencies have been collecting information on us for decades and we accept it without much complaint.  But I’m not so sure.  Clearly, there is a dark side to technology.  Along with mass personalization we’re also building massively intelligent systems that can know us in a way that even those closest to us don’t.

One thing is for sure.  Our technology is getting to know us to a degree that was unthinkable even a few short years ago and we’re all going to have to decide how we feel about it.

– Greg

6 Responses leave one →
  1. Ajoy Vakil permalink
    February 2, 2014

    Thanks for a wonderful article Greg!

    I am sure you will agree that for the customer – personlisation is welcome – intrusion is not.

    Marketers need to distinguish between the two – and if we can build so much Marketing intelligence into our systems – we should also be able to build intelligence to distinguish between the two.

  2. February 2, 2014

    I think the problem is that those things are very hard to separate, especially since I don’t think anybody has been able to define exactly where good service ends and intrusion begins.


  3. Kenny permalink
    February 3, 2014

    Or the even further problem is that intrusion begins at a different point for different people. Meaning of course that the service / intrusion line for personalization must be, well uh, personalized.

  4. February 3, 2014

    Very true and a very good point.

    Thanks Kenny!

    – Greg

  5. February 3, 2014

    Successful companies already know how to provide excellent customer service: they listen. Corporations by and large find customers annoying and only wish to separate them from their money. While all this fancy technology could be useful to the few that implement it wisely, it is simply the latest attempt to figure out how to generate income without having to actually deal with humans.

    In a way I can’t blame them as humans can be very difficult. But how you treat them does determine your success – in the long run. And therein lies another problem. Corporations don’t care about long term success. They foolishly pay their executives based on stock price instead of profitability which led them to cook the books. convert real pensions to cash pensions, sell off valuable assets, and down-size to the point where they are in breach of contract, their employees are burnt out, and if they do still build anything it is of such poor quality it isn’t worth having.

    The future will be quite interesting as more and more formerly ‘middle class’ are pushed out of the formal economy. Most do not want to leave their conditioned comfort – but as inflation rises they will have no choice. I wonder what percentage will end up living outside the world the poor never lived in?

    What percentage do you suppose will reject all this technology, refuse to carry a tracking device that i-beams their every movement, stop buying from corporations, and live a simpler life outside technology?

  6. April 5, 2014

    A whole new level of personalization is coming in digital experiences as well. Ecommerce as of now is very much a self-service experience. The shopper browsers to the what he/she wants, applies filters where available and browses through a maze of pages and products to hopefully get to the the product(s) that are relevant. The so-called personalization solutions that exist today try to analyze historical browsing and buying behavior, which is incomplete more often than not. It is a known fact that over 95% of etailers dont have historical browsing / sales data on over 95% of shoppers visiting their sites. If you put aside giants like Amazon, eBay, Overstock, etc., the rest of etailers dont have enough data to understand their shoppers while they are new; and if the shoppers don’t get what they want in the first visit, they are very likely to not return to that etailer.

    However, the trends of the last few years will change this soon. Shoppers make their buyig decisions based on needs and preferences, which can now be inferred from data sources other than historical browsing / buying. Social data is one such massive source of information on consumers. Companies are us are developing ground-breaking technologies to infer shopping references from social interactions to build the next-gen personalization solution. Web shopping experience will soon be like the neighborhood restaurant where the servers and chef know you personally and know exactly how you like your food. Gone are the days of segmentation.

Leave a Reply

Note: You can use basic XHTML in your comments. Your email address will never be published.

Subscribe to this comment feed via RSS