Friday morning. I receive a text message about the special offer of jackets that ends this weekend. Then, before noon, my phone reminds me about the email received from one of the airlines with the last-minute offer. In the afternoon, I spot a sponsored article on Facebook about the electronics sale. This sounds like everyday life of all of us that we have been accustomed to since quite a few years. We have easily learnt dozens of habits to check notifications. Now I find it odd when I recall the beginnings of the Internet when first emails triggered so many emotions. The personalization in the world of digital marketing absorb more and more areas of people’s life with the use of new technologies. However, this was not always the case. Interestingly, its first forms had come into existence long before the Internet was even invented.
The Beginning of Personalization
The beginnings of a worldwide personalization (or: the individualization of products, services and offers) can be traced back to the transformations of loyalty cards, which had been rather anonymous before. Even in the automotive world, there had been “personalized” cars available but a long time had to pass before such offer became accessible for the masses. It is also worth noticing that first loyalty programs were introduced in the air transport industry due to the Air Miles program in 1988 and, soon after that, also the retail branch spotted its potential.
At this time, Tesco initiates the Clubcard program, which has become a true revolution. It is thanks to Clubcard that the first “real” loyalty cards started to appear to gather information about clients.
The problem of clients identification was somehow resolved, but another one appeared: if we do already (at least partially) know our clients, how can we select the best ones? To be more exact: how do we reward those who generate more revenue for the company and what can we offer to the rest?
Then, the concept of so-called “baskets” came into being. Every basket was a client segment divided according to demographic criteria. Simultaneously, the term RFM (Recency, Frequency, Monetary value) start to be used.
What exactly does this concept mean? Simply put, each basket combined its demographic data with its RFM value, for instance:
Basket #6 :
- a women, age: 40-50, city size: 50-150k (demography)
- last shopping done not earlier than 30 days ago (Recency – period since the last shopping)
- she averagely buys not less frequently than 2 times a month (Frequency of shopping)
- annually, she spends the amount of 2k EUR for shopping (Monetary value – the yearly sum
On the basis of such data, in 1990s Tesco printed personalized brochures and sent vouchers to their clients quarterly. Every client segment (basket) received a different one. At that time the number of segments could be counted on the fingers of both hands. Yet, it was still perceived as a huge marketing achievement.
The baskets went through another transformation in 1998, when based on multiple studies and surveys (and plenty of presumptions) marketers decided that not every basket could suit a client’s need. A single student of medicine, dependent financially on his parents and a young mother who lives with her husband could have been selected to the same basket. Despite many similarities, these groups had different needs. This was the origin of the segmentation based on a lifestyle.
Today this type of segmentation is often called Personas (especially in the Design Thinking patterns). The division of clients according to their lifestyles created dozens of new segments and these, in turn, allowed to get to know clients better than before. Individual needs of every one of them could finally be satisfied.
Along with the development of information technology, there was a sharp increase in client data. However, the computing power could hardly manage to deal with it. Then, there was a major technological breakthrough, which laid foundations of nowadays’ digital personalization. Not long before the new millennium came, client information used to be archived on discs, which were transported to data centers by trucks. As soon as the Internet became available for anyone and the transmission speed increased, everything changed.
The development in the IT area has brought new technologies. Those have allowed to gather data concerning a client’s activities during every step of the shopping path and to aggregate them in data centers. Advanced cash registers, credit cards, mobile applications, high-resolution cameras and the miniaturization of every electronic device has led to significant changes. Initially, a client’s behavior was assessed instinctively and the sale speculations were estimated based on little available information. However, later the sale was subject to the optimization process during every stage. Printed advertising leaflets have been replaced by emails, receipts are more and more often electronic and cash is being less and less used in most of the developed countries of the world.
Algorithm that predicts future
It can be predicted – with the use of the customer behavior analysis – which goods would be bought by each individual during every stage of their life. For instance, we can foresee that a person buying particular products will become a mother within the next year. The biggest benefit for a retailer is to be able to prepare personalized offers for her much earlier.
It is worth to mention a situation that happened in one of the hypermarkets
A father of a young girl came to the store demanding strongly to see a manager. During a heated discussion, a father clearly stated that by sending a leaflet full of products for young mothers, his daughter may feel encouraged to premature maternity. A manager, of course, apologized for this misunderstanding and deleted the address from the database. Imagine how surprised he must have been, when a few weeks after the whole disagreement, a father came to him once again to apologize since it turned out that his daughter had been pregnant before he came to the shop.
Mathematical models don’t lie and the algorithm had known earlier. This example clearly shows that despite we all have individual behavior, sooner or later we will be categorized to one of the marketing behavioral patterns.
Soon after, retailers and corporations started to suffer from the problem of a thorough sales estimation with respect to marketing budgets. Thanks to this, companies receive information about which client type is profitable enough to invest in and how much personalized a message must be to become well established in a receiver’s mind.
One of the main worldwide promoters of personalization the Nike company which enabled its clients to create individual patterns on sports shoes. What seems like a virtual fun from a client’s point of view, is in fact an immense undertaking for a business. Of course, Nike is not the only company which has tried this solution. Bivolino, Hoffman, Boss, Timberland or Shoes of Prey are a few other examples in the fashion branch.
Nowadays personalized marketing message, personalized goods and services are just part of everyday’s life. Formal polite phrases in emails or internal communicators (e.g. Help Desks) have been successfully replaced with expressions such as “How are you Peter?” or “We know that you have been passing by our shop today and that you have a credit card. Maybe it is high time to spend your first salary on a new mobile phone?”. The last message sounds scary but is it really different from the reality? Now it is not about remarketing anymore; information about us is used to create a personal message which is difficult to ignore. Is it possible that such actions are legal? Yes, they are…
What data do we exactly share?
How does it work and where does the information about users come from? Mostly, we hand them over by ourselves and the most advanced corporations make use of it. We have a range of information which are managed by a company. They can be divided into 4 categories:
- Search queries
- Page views
- Cart actions
- Past purchases
- Browser sessions
- Product details
- Visits and views
- Inventory information
- Social media
- Context information
- Time and date
EDW, ETL and DMPs – what does it mean?
All this information is exchanged between different platforms. A user is identified by cookies, pixels and IP address or a session. Then, corporations create data warehouses (EDW – Enterprise Data Warehouse) where the information on every client retrieved from every available source with the use of ETL tools (Extract, transform, load) is stored. Later, such data is processed and analyzed by Data Management Platforms (DMP). Such prepared gigabytes of information are sent and combined with hundreds of tools within different departments of a particular company. We are talking about thoroughly gathered information retrieved from diverse sources, thanks to which a company can be almost 100% sure that a sent message or an offer will reach you, and that you are really you.
How is this information used?
A good example of highly personalized service is a customer service. From a client’s perspective, it could be a regular phone call to check service costs, but what does really happen on the company side? The system identifies our phone number even before a consultant picks up a call. Simultaneously it prepares a personalized offer based on our telephone number. On the consultant’s screen the whole set of information is displayed (e.g. how old I am, if I am insured, where I am calling from, where I live, if I am married or have any kids, how many phones I have or if I have a TV etc.). Having such information at their disposal, corporations such as T-Mobile can suddenly make a personalized offer to you for a car lease during a conversation.
Another example are recommended products on websites even though we have just visited this page for the first time.
Marketing Automation, on the other hand, enables us to create, among others, scripts for particular clients’ segments. It means that with the use of such tools I can personalize the way how a particular group of recipients will receive my brand’s message. If it is someone using a smartphone, he or she will be redirected to a different website than someone who uses a laptop. If this smartphone is iPhone and its language is English, then a person will receive a different offer than a person who visits this website on a PC.
Additionally, with the use of Big Data we can configure algorithms in such a way that they can independently decide how to personalize the messages in order to achieve the best ROI or KPI.
Something of a novelty that started to appear on the market around the year of 2014 was the mass products personalization which has been picked up quickly by many producers. The leading example are Coca-Cola’s cans with names. A product that looks almost exactly the same as over the past decades, after a slight marketing exercise takes on a new meaning. After all, isn’t it nicer to buy a can of coke for a friend when his name is printed on it?
Personalization: Side effects
Every marketer is aware of personalization advantages; yet, in broader – social terms – the personalized content may cause changes in the way we think. The biggest problem in a global scale is the polarization of social perception which locks the users in hermetic information clusters. What was shown before on, for instance, Facebook, was a set of randomly chosen information posted by our friends. Now, users do not have anymore “access” to information outside the range of their (predicted) interests.
The polarization of social perception
This leads to situations where we are surrounded only by information which we perceive without giving to it any deeper thought. Our “online life” is limited to the groups of our friends which share similar content. It makes it seem that if everyone has similar views and everyone probably thinks the same. We miss information concerning disquieting topics and different opinions, we are happily locked in a bubble of common interests. Even when a friend shares some extremely distinct view, Facebook algorithms successfully filters it out and gives to it a low priority. In such conditions, extremism is easy to achieve. Everything that we receive matches our taste and there is no need for any deeper reflection.
Another side effect is an ordinary lack of inspiration. In our social media channels, we receive the same repeatable content over and over again from the same people. Everything is very much alike and instead of inspiring us with a different view, the same schemes of “similar” people are copied.
The future of personalization may result in some threats. The examples of not so much distanced visions are presented in the Black Mirror series. If a system assigns us ones to a specific category, it will be difficult to become “unlabeled”. New technologies will help the companies to know more about us than we may imagine. Netflix already advertises the movies based on our taste. Maybe sooner than we think screenplays will be written by algorithms?
Of course, there are also many positive examples. RFiD chips (or similar ones) implanted under skin, despite much controversy, may make our life easier. The development of bioengineering will let us create personalized therapies based on the information gathered from the patient remotely in real-time. All house appliances during the time of IoT (Internet of things) will adjust to our mood and prepare our favorite meal for us or advise us to watch a film we will love. Today we can personalize almost any kind of service, map behaviors and adapt the service to our liking without having to conduct a survey first. The personalization in the marketing area should be treated not as a weapon but as a tool which helps us to get what we want without much effort.
 Clive Humby, Terry Hunt, Scoring Points: How Tesco Continues to Win Customer Loyalty
 Charles Duhigg, The Power of Habit: Why We Do What We Do in Life and Business
 Dana Mattioli, On Orbitz, Mac Users Steered to Pricier Hotels, “The Wall Street Journal”