Common and serious errors in automatic profiling
During one of my eMarketing classes, we were doing a standard exercise of creating a “buyer persona”.
The idea behind it was to establish a profile that represents the average buyer, including basic demographic information such as age, and gender, and also define some behavior patterns that could be used for a marketing strategy.
A bright student came up with an interesting question: How can we develop an accurate profile if we don’t know the average customer? I explained to the class that is the same as having a physical store, where you can observe how the customer reacts, where does he/she goes first and where they go next, where he/she spends more time, etc. In Internet, the same behavior can be “observed” by following click patterns, by the use of cookies, and several other tools available. So, if we want to open an online version of our physical store, the e-customer profile should follow the same pattern as a physical customer profile.
The next question, however, kept me thinking a little bit longer, when she asked: But isn’t online behavior different from regular behavior? People act different when they are online…” Once again, the question from my student was far more interesting than my response. And so, I start making some additional research behind the psychology of customer profiling and the differences, if any, between Internet and the “real world”.
Consumer psychology
Psychology is not an exact science, and, as it happened to every other business study area, consumer behavior has been heavily impacted by the Internet and the use of new technology. If we can learn one thing about the Internet bubble in the year 2000, is that online business is not the same as regular business. Similar basics apply, but, as expected, the real difference is on the details.
How the Internet is affecting human psychology is a fascinating subject. Just think about the reasons behind how a shy and introvert person in real life devised Facebook to gain friends, even if mainly online. How the difficulty to find information in an ever-growing world wide web led to the creation of Google. Amazon “predicting technology” revolutionized the industry by selling you books based on other people’s buying patterns. On a darker side, think on how online anonymity enables trolls and grooming, not to mention other criminal activities.
Back in the dark Internet ages (read 1990s), online consumer behavior was analyzed and discussed by a few “experts” that primarily wanted to apply the same rules and processes form regular business to a new marketing arena. Unfortunately, then there were not too many tools and systems available so they had to rely mostly on humans to do all the work.
Marketing -and now eMarketing- are facing major changes and new ideas pop up every minute. While in the past it was the responsibility of marketing experts, now, profiles are managed by complex systems, “robots” guided by a strict programming. Automation is good as it allows for a faster process but it also leads to some unfortunate errors.
The Internet footprint
Some months ago I was reading an interesting article on the current gun control debate. I found a reference to an “AK47”, which I never heard before. So, following an old habit of researching unknown words or concepts, I found in some web pages that it referred to a military type assault weapon. Not being a gun fan, I probably spent only a few minutes browsing through some pages until the concept was clear enough to return to my original reading. To my surprise, not long after, I started receiving offers for “guns and ammo on sale”. I was concerned that soon an invitation to join a guerrilla or a visit from Homeland Security or the FBI would follow.
So, what happened? A robot followed my internet browsing history, and based on the sites visited, “decided” that my profile fits a gun buyer persona. Age, gender and 15 minutes of browsing aside, I differ completely from this profile. I have never owned a gun and had no plans for buying one either; nevertheless there is a new profile for me on the Internet. This is also probably why, after 15 years of happy marriage, I receive messages regarding my “upcoming wedding”. Or why, after checking some nice pictures of Bali, I am flooded with flight offers and I was even contacted by a real state agent specialized in “luxury islands for sale”.
Aware or not, with your consent or not, these robots look, among other things for:
- Cookies
- Browsing history
- Search queries
- News read
- Social media likes and follows
- Online purchases
- Website affiliations, registrations or even “anonymous” surveys filled
The list is endless, and each item mentioned above, provides a small piece of the puzzle that robots use to create your Internet profile.
Privacy matters
I have a beautiful house with large windows. I love the feeling of the sun rays coming through every morning. But I also have curtains. And I use them, for example when I am cooking breakfast in my pajamas. Some aspects of private life are meant be kept private.
Even though I have nothing to hide, I don’t want to tell the world every single thing that happens to me either. Social media addicts love to post every step they take, every meal they make… and someone is watching you. Putting creepy and horror stories aside, keep in mind that eMarketers are also watching. So, not only for security reasons we have to be smart on what we do or don’t do online.
Robots and Criminal Minds
I love the TV show where a group of psychology experts develop a “profile” that eventually leads to capturing the bad guy. They do it based on an extensive experience and knowledge of human psychology. And they also do it thanks to the vast amount of information available on all of us.
Are you Internet paranoid already? I am.
Businesses and professional eMarketers, however, do not rely on a panel of psychology experts. They rely in systems that have been created for that purpose. I like to call them robots but I know they are just software stored in state-of-the-art servers with super high Internet connections, and a humongous database that stores Godzillabytes of information.
These robots just do what they are programmed to do: crawl the web looking for data and provide valuable information to marketers. If that information could then transform into knowledge, business intelligence or wisdom is another story.
Privacy and Security… Less is more
While is not obvious to everyone, information sharing common sense dictates not to publish your holiday vacation photos until after you are back home. Yet, I even know people that publish sensitive personal information like phone numbers, social security numbers and credit card information.
While identity theft or online fraud are very serious issues, the amount and type of information you share, -online and offline-, can lead to all kinds of headaches.
From major security issues to annoying spam in your inbox or social media, you need to understand how to control it.
You also have to be aware of where to publish information. While it may be ok to bloat a CV in LinkedIn for example, is not a good idea to publish every picture ever taken on Instagram or every move you make on Facebook. You know your friends. You may even know some of the friends of your friends. But it is impossible to know all the “friends” or contacts of the friends of your friends.
All the major social media platforms are aware of these issues and offer various levels of privacy settings. It will not hurt to read all these boring instructions, and revisit your settings once again before liking, sharing or following the next post.
Be smart. Be safe. It’s up to you to feed the robot with the data you want to share and have the online profile you want.