Marketing, as a key part of the Business world and in the development of Management with clients, is seen today as the "activity" that can lead to better knowledge of the tastes and preferences of clients and that in the current process has led to propose various Marketing formulations, one of them the Technified Marketing, which implements different forms of Artificial Intelligence obtaining two approaches at present, the first, Humanation, supported by tests such as the Turing Test (machines think) and the Human Test (adaptation of the Turing Test, to prove that one person can think the same as another), and the second, Datamining.
The first approach is based on an internal context and is aimed at the reasons that motivate the purchase, while the second approach is based on an external context and is concerned with products that motivate the purchase, concluding that by demonstrating, through The tests described above, the functionality of Artificial Intelligence applied to marketing and communication, improves and helps sales.
Keywords: Marketing, artificial intelligence, intelligent agents, Datamining, Humanation, scientific rigor, repeatable facts, purchase guidelines, return on investment, ARPU.
CONTENT
- Presentation Agents Inspired Technologies SA (Agents) Artificial Intelligence Turing Test - When is a machine intelligent? Human Test - When is a human being intelligent? Application of the Human Test to Databases Technified Marketing Approach 1 - Example of Humanation Approach 2 (optional) - Data Mining and example Conclusions
1- Presentation
Agents is a team of people specialized in marketing technology with their own tools and technology for the analysis and prediction of customer behavior.
Catalan company founded in 2000, is the first mixed capital business experience of the University of Girona, it is also owned by BCN Emprèn Fonsinnocat and the French Najeti Capital. It has its own technology, on which innovative marketing application tools have been developed in different industries or sectors, the result of 15 years of international research in Artificial Intelligence and its services having been successfully tested by some large companies, it is emerging as a large company with a future projection.
AgentsSince its inception, it has worked with the applied computer research institute of said university, taking advantage of its research experience on the subject of intelligent agents, datamining, and artificial intelligence applied to a wide number of sectors ranging from the meat and electrical industrial sectors., biotechnological, robotic to service sectors such as virtual education, management computing, Internet and marketing. The research groups of this institute stand out for their relevant work in the field of intelligent agents, experienced in the first robotic soccer team in Spain that competes at the highest international level. The results derived from these investigations have shown that the harshness of soccer competition presents analogies to the challenge of understandingthe laws of supply and demand when decisions are made by human persons, which has made it possible to use the know-how of the world of artificial intelligence in robotics and thus apply it to the world of Customer Intelligence in a broader sense, and, in a sense more rigorous to Marketing.
Thus, in Agents, there is a team of doctors in artificial intelligence who work side by side with high business academic training and, above all, experience in this sector. All of them, obsessed with transferring to the business world the knowledge that continues to be generated in said research institute, giving one more guarantee of having the latest advances in Customer intelligence.
2- Artificial Intelligence
It is a term that has given much to talk about during the 90's, both for its context and for the meaning of its words, that is why this document refers to two descriptions, the first to establish the relationship of "intelligence" with " artificial ”, outlining the meaning that the dictionary gives to the word“ Intelligence ”and in turn to the complement“ artificial ”, and, the second is the description that Agents (AIT) makes due to their knowledge and experience, as will be explained to continuation:
Description of Intelligence
- Faculty of knowing, analyzing and understanding Skill, skill and experience. Artificial intelligence Set of techniques that, through the use of computing, allows the automatic performance of operations until now exclusive to human intelligence.
Description and characteristics according to AIT due to the experience acquired and knowledge of the subject:
- Artificial Intelligence (AI) is called the scientific discipline that aims to create autonomous systems capable of working in uncertain environments. AI is often inspired by human, biological or natural behavior. AI has not met expectationsAnd above all, AI IS NOT MOVIE ROBOTS.
Its origins go back thousands of years, because in almost all mythologies there was some kind of divine or almost divine "machine" of this nature. Defining its beginning in the Modern and Contemporary Age is very difficult since there are many inventors and geniuses who have contributed to create these machines, Leonardo Da Vinci, Blas Pascal, Charles Babbage or Alan Turing, who will be discussed later.
Artificial Intelligence or AI in Spanish and AI in English, is a science belonging to the branch of Cybernetics, which studies the mechanism of human intelligence in order to create intelligent machines, capable of performing calculations, "thinking", making judgments and make decisions.
And, as the new marketing systems try to conceive products that are more in line with consumer tastes and preferences, designing special services and personalizing dialogues with consumers, they encounter uncertain environments every day and very difficult to predict the most accurate behaviors about them.
That is why today, when finding out how and in what way these increasingly uncertain market environments can be simulated, this branch of cybernetics has been implemented, thus obtaining a more technical and more successful marketing.
The question at this time lies in answering questions, such as: are the new Marketing systems, such as Relational, being optimized more and more, or are they not giving the expected results? And above all, are they being Noticing the tacit but existing information in the databases? Or if, on the contrary, this valuable and relevant information is being lost, which may be that hidden treasure right next to those most interested in this information.
Agents, by experimenting with two kinds of Test, which are derived from Artificial Intelligence, has discovered the way to manage customers in an optimal and very successful way. To go into detail, first analyze these two Tests.
The first Test, by Turing, defined to prove that a machine can make decisions and think like a person and, the second, by Human, to prove that a person can make decisions and think like another person.
But, before giving conclusive judgments, it is necessary to know a little more about these issues, therefore, the above-named tests are described below:
- Turing test
The Turing Test (1950) serves to prove that a machine can make decisions and can “think ” like a person.
With the previous Test, since the fifties, the importance of Artificial Intelligence in a generalized context has been shown, so the question is, CAN MACHINES THINK? ARE ROBOTS INTELLIGENT?
Alan M. Turing (1912-1954) described the Test as "an imitation game" in his 1950 article, showing that if a person cannot distinguish an answer given by a human being and one given by a machine, it follows that the machine was "smart."
For a better analysis, as Turing describes it, one must take into account what a machine is and what it is to think, and thus make clear the dilemma of meanings.
The Imitation Game: << ”I PROPOSE to consider the question, can machines think? ' This should start with definitions of the meaning of the terms 'machine' and 'think'. Definitions can be framed to reflect normal use of words as much as possible, but this attitude can be dangerous. If the meaning of the words 'machine' and 'think' is to be found by examining how they are commonly used it is difficult to come to the conclusion that the meaning of the answer to the question, can machines think? " >>.
Meaning of Machine?:
- Set of mechanisms arranged to produce, take advantage of or regulate a motive energy: sewing machine; steam engine Locomotive: the train has one engine and six wagons Electronic or electrical device that works by introducing money and offering a service or product: there is a tobacco machine at the door of the bar Stage of the theater for transformations of the scene: the actors thanked the people in charge of the machine for their work. Adv. With the help of devices or mechanisms: this garment can be machine washed. Adv. Very fast, at full speed: we are working at full speed to finish on time.
Meaning of Think?:
- Form and relate ideas in the mind. It is built with the prep. In and on: I was thinking about how to do it. Examine something before making a decision or giving it a solution: I have to think about whether I accept or not. Conceive a plan, procedure or means for something: I was thinking about how to tell it. Have intention of what is expressed: today I am not going out. Have someone an opinion about something or express it: I think you are not right. No way, absolutely not, in any way. To think badly or well loc. Interpret negatively or positively the actions or words of a person, distrusting them or not: don't think badly, I'm not going to deceive you. Adv. Suddenly or unexpectedly: he did it without thinking. Irreg. It conj. How to get it right.
Turing in his "game", performs tests between a machine and a person, and, if the interlocutor does not realize who has given the answer, whether it is correct or not, it follows that machines can think, for example:
First, an individual X is asked what is 2 + 2 ?, Or, Does God exist ?, He answers that 6 and no, respectively; second, the interlocutor does not understand who or what has given him the answer, whether it is correct or not; third, by not knowing the source of the previous result, it follows that the machine and the person who are responding can think, in other words, how the robot cannot be distinguished from the human person, as a consequence, and fulfilling the Turing Test deduces that the robot is intelligent.
It must be seen that an important quality to consider that a robot is intelligent, is its ability to emulate the behavior of a person, lying or hiding behaviors to simulate said intelligence. This characteristic, as will be seen later in the Human Test, is typical of surveys, where there are severe problems in understanding people's decisions.
- Human test
The Human Test (2002) serves to prove that a person can make decisions and can “think ” like another person. For example, it is used to decide if said person is stingy, has free time, is sensitive to privacy, if they do not care about personal treatment, and if they care about the brand…
Josep Ll. de la Rosa describes the Test as "an imitation game" in his patent Personalized Robotic Game, known as the inverse adaptation of the Turing Test, and in turn, adapted to Customer Intelligence, showing that if based on questions and answers it is not possible distinguishing an answer given by an "intelligent" human being to one given by another that we do not know, it follows that the second human being is also "intelligent". Likewise, if an answer given by a human being "sensitive to privacy" cannot be distinguished from that of another that we do not know, it follows that the second is also "sensitive to privacy" and so on for an infinity of attributes.
For example:
- If a person is asked for their mobile phone number to order a product or service from a company and she agrees to give the number, it follows that this person may “not be privacy sensitive”, likewise, if Another person performs the same act, it follows that he has the same degree of sensitivity to this attribute as the previous person; That is why the Human Test is described, like the Turing test, as “An imitation game.” Another example is, at a table in a restaurant, the 5 people at the time of paying, usually one person pays “x ”, It can be assumed that it is not stingy. And by adding that she buys movie tickets online by paying for herself, then the Human Test hypothesis is progressively being corroborated. This person answers affirmatively to the question, “Do you pay?” Now, if said person “x”,calls a phone number that is subscribed to a discount plan to call loved ones, and does so at 11:00 in the morning, this may mean that apart from having strong family ties, he has time availability, and may be liable to accept the request of “Shall we go out?” If the same person withdraws money from an ATM during office hours, that is, the office is open, this may mean that said person is not sufficiently attracted to personal treatment as well as queuing to be attended personally. Thus, said person presumably is not too attracted to personal treatment, when asked “How do you get money?” Likewise, if person “x” when asked “What do you drink?” Responds that Coca Cola, possibly said person is very affects the brand,at least it acts the same as people who, faced with the possibility of deciding to consume "brand" or "non-brand", usually opt for the "brand".
- Application of the Human Test to Databases
Prior consideration: the surveys are not reliable for several reasons, which will be explained later, on the other hand, in the databases there are numerous decisions that people make on a day-to-day basis, which is as if there were a lode of pending responses. find the right questions from an imaginary survey… but effective!
In the modern world there is a progressive standardization of purchase and decision guidelines within a globalized world, paradigms of normal marketing theories, but different sensitivities can mark different reasons for people to behave the same or differently before a purchase decision.
Now, the interesting thing is to realize that traditional marketing tools have their shortcomings and disadvantages, then, you have to experiment, renew, act in the face of change and try to innovate as much as possible with different approaches, systems, ways of carrying out a process, try to optimize the results obtained, always with the aim of satisfying the customer as much as possible, for example if the Human Test is implemented in the Marketing databases… what can happen?…
Before looking at the answer to that question, the current shortcomings and disadvantages of classic Marketing tools should be analyzed, such as surveys, opinion polls, pilot tests and / or tests in representative groups of the market, which try to detect the most needs. approximate, but not that the vast majority can be inaccurate and others can contain information from the target audience in a biased way such as surveys and polls generating a high risk in the collection of that information.
- Market research based on surveys and focus groups has the following problems:
- They can obtain ambiguous and manipulable results. Their answers can vary in a relatively short period of time. They are carried out in a very long period of time (months). The results are not easily repeatable or verifiable.
Returning to the question of, if the Human Test is implemented in Marketing databases… what can happen?…
With this Test and by implementing it in the databases, it is possible to find out about each person their SENSITIVITY regarding MANY psychosocial attributes, for example, a client named A, is known through his acts registered in the databases that, at the time to pay in a restaurant he does not do it and lets his companions pay, it is known that through the "friend" telephone plan he speaks "X" minutes much more than the average per person, it is also known that he gives without any qualms his mobile number to the companies that have requested it, likewise, that it is faithful to Coca-Cola, and that it likes to carry out its transactions by ATM. (See figure 1).
- Once the attributes that define each person are known, it is necessary to discover which are the most important, their hierarchy, which is called sensitivity. The sensitivity is obtained after repeatedly applying the Human Test in the client database or by exchanges of sales experiences between sellers.
- TECHNIFIED MARKETING
The current world is an ocean of changes and developments, that is why the basic, technological and humanistic sciences seek to develop in the best possible way in that modern ocean, and Marketing as a fundamental piece of Business Sciences and being one of its branches that interact directly with people, must be a pioneer in adaptations to change.
Defined, by recognized authors, as the " set of techniques that through market studies try to achieve the maximum benefit in the sale of a product, a merely theoretical definition, for many others, today Marketing is seen as the " activity " that It can lead to better knowledge of the tastes and preferences of customers and that in the current process has led to the formulation of various forms of Marketing, one of them, the one in this document, Technified Marketing.
It's called Marketing Tecnificado to a new way of marketing by introducing scientific rigor in all decisions of marketing and communication, taking full advantage of the information stored customers Databases, relying on intelligent agent technology, a term that refers to a programming paradigm based on artificial intelligence, in some cases; Thus, the various approaches to Business Intelligence techniques support technical marketing, in all its strategic, operational and sales support dimensions.
The dominant sensitivities, (which is all that subjective information of the client reflected in their actions and immersed in the databases) usually appear as explicit preferences, but in reality they must be interpreted differently for each individual. Different Consumers will have different definitions for terms such as: long movie, romantic trip, safe investment, or delicious food; Categorizing these differences and understanding the personal interpretation of each consumer is beyond the capacity of most technologies, so the tremendous variety of interpretations is daunting, however, understanding these dominant sensitivities leads to increased loyalty and produces greater benefits.
The preferences of each consumer allow a company to offer each individual the product they are looking for at the right time, offer a customer exactly what they are looking for at that moment, makes companies exceed what the consumer expected of them, which is well known by all marketing professionals, and that in turn, builds consumer loyalty and allows greater benefits.
With technified marketing, the necessary and current approach is achieved , direct relationship with the client, in all our actions and marketing decisions (Customer Intelligence), protected by the scientific rigor that validates all the arguments that support them.
It is the measurement and breakdown of the sensitivity of customers to predict behaviors and highlight the COMPETITIVE advantages of a company's offer, the fundamentals of the decision are investigated in all sectors, the daily life of a "typical" customer is full of decisions, stored in databases (DB), through records, whose interpretations are not always explicit, some decisions are influenced by sellers, others not.
Agents Proposes a Technified and simple Marketing, taking advantage of customer databases and learning from the success stories of vendors.
In its development, Technified Marketing finds two approaches, which are: Data Mining and Humanation, both approaches support the current trend of proximity marketing, see Fig. 3
- Approach 1, Humanation
His focus goes to the reasons that motivate the purchase (INTERNAL) and tries to solve How to sell said product and to whom?
It is called Humanation, an innovation in marketing that uses Artificial Intelligence techniques to create decision models of people and predict their behavior of purchase, use, abandonment, response or opinion.
This approach, Humanation, has the advantage of making business rules profitable
for example:
A consumer close to the consumption of beers is highlighted its characteristics of freshness and comfort, avoiding talking about health, so:
Beers = freshness, comfort and non-health.
Methodology:
In summary, the methodology used in Humanation is the following:
- The sensitivity of prospective customers is discovered thanks to in-depth knowledge of the most active customers, then the target quality is defined, which is then cross-referenced with the customer's data, thus generating a new database, which contains the sensitivity of each The next step is to use this dominant sensitivity and to define the arguments to convince good customers and winning arguments to attract new customers. So the sensitivity is used to improve the scoring in the generation of campaigns.
The Humanation uses artificial intelligence, with its new technology, thus it integrates the software existence, giving companies a "empowerment" that will reveal them closer to their customers, and above all, do it faster, this way, they will be able to offer the consumer everything they need and above all, at the right time, and thus, increase their benefits significantly by treating their customers in a more accurate and personalized way, and, as they deserve it, see figure 2, a case of success in Euskatel, in the adaptation of telemarketing arguments during the first 30 seconds of communication.
Humanation uses intelligent agents as software or cutting-edge technology, which is a term that refers to a programming paradigm based on artificial intelligence, they are numerous dynamic elements that interact and learn from each other or from their habitat, databases and / or information systems, which allows them to emulate the behaviors of people or machines, and make decisions according to the mission for which they have been programmed.
Approach 2, Data Mining
Its approach is aimed at products that motivate the purchase (EXTERNAL), it tries to solve What product to sell and to whom?
Datamining is called the statistical or Artificial Intelligence techniques used to create a black box model that correlates the effect between 2 or more variables.
This approach, Data Mining, has the advantage of implementing business rules.
For example:
If a consumer over 25 years of age buys diapers, then he will buy beers, that is, people over 25 years of age, plus, diapers, equal to beers.
- Conclusions
AI helps to sell more, as described above, since by applying the Human Test to databases, it is possible to know each person in a truthful and correct way, to then be able to use that knowledge and better argue the sales.
Thus, by applying Technified Marketing techniques on internal customer databases, much more can be sold thanks to a better selection of the target audience, a better argumentation, and a better customer focus, achieving the necessary and current approach "towards the customer ”in all marketing actions and decisions (Customer Intelligence), protected by scientific rigor that validates all the arguments that support them.
Likewise, in this document two approaches to technified marketing have been introduced and, Why prefer any approach? Because like every reasonable and critical human being who looks for ways or forms to optimize decision-making, call it life decisions or business decisions, in this case it is the latter that are of interest, preference should be given to the approach that at this time can better address the changing ocean of the business world, so the preference that at first glance stands out is the first approach, Humanation, which has the following advantages:
Better anticipate people's decisions, so that their decisions can be more effectively influenced, strongly supports the sales force and Relationship Marketing, selects the most effective arguments, supports creativity in Marketing, since, find values and reasons for purchase different from price, brand, quality, geographical proximity, etc.
It works with all clients and not only with samples, with the advantage of giving more solid results, less interpolated, and additionally, it can be better adapted to the Public Administration.
Apart from these reasons, it should be taken into account that it also provides the following benefits:
Higher ARPU (Average Revenue Per User), since it desensitizes the price to consumers, it is cheaper because better average results are achieved with 20% of the effort required with the second approach, also, it is more efficient because the resulting work is more easy to maintain and is much more focused on Marketing, requiring less scientific support, therefore, cost savings and reduction of culture shock.
And, the drawback, which in turn is a great characteristic of this approach, is that, by not using external information gathering tools, such as surveys, polls, focus groups, etc., therefore, it does not use sociodemographic data.
What should not be forgotten is that the objective of these two approaches is to carry out Marketing each in its own way.
Recommendations:
- Combined approaches with emphasis on improving the rigor and result of Marketing and Communication Extensive use of approach 1, Humanation, combined with business rules Use VERY SIMPLE versions of Data Mining, approach 2, to generate simple, powerful business rules, with rigor à Advantages: fast and safe obtaining of results.
BIBLIOGRAPHY:
BOTTI, JULIAN; INTELLIGENT AGENTS, pdf format. V., Department of Computer Systems and Computing, Polytechnic University of Valencia. (2003).
TORTAJADA, IMMA, conclusions of the ix innovacompres professional breakfast, Economic Dossier of Catalonia, article of July 8, 2003
DE LA ROSA, JOSEP LLUÍS, Conference of the Marketing Club in Barcelona Spain, Nov 13, 2003.
HAGUE, PAUL N. AND PETER JACKSON, Market Research, Harvard Library, Planeta D'Agostini, pp 90-92; Consumer behavior and Marketing, José Luis león and Elena Olabarría, Ediciones Deusto SA, Page 345.
HAGUE, PAUL N. AND PETER JACKSON, Market Research, Harvard Library, Planeta D'Agostini, p. 93
KOTLER, P. Marketing management: analysis, planning and control. Prentice Hall. nineteen ninety five.
KOTLER, P; introduction to marketing. (1997). McGraw Hill (3rd edition).
LAMBIN, JJ, strategic marketing. (1992):. McGraw Hill (3rd edition).
LEÓN, JOSÉ LUIS AND ELENA OLAVARIA, Consumer Conduct and Marketing,, Ediciones Deusto SA, pages 231 - 234.
MUÑIZ GONZÁLES RAFAEL, Marketing in the XXI Century, Volume I, 2001
SUEIRO VILLAFRANCA, ENRIQUE. Article Ten keys to reconcile scientific rigor and journalistic dissemination, November 15, 2002, Diario de Noticias (Pamplona).
WEBGRAPHY:
answermath.com/mineria_de_datos.htm
cogprints.ecs.soton.ac.uk/archive/00000499/00/turing.html
Custom Robotic Game patent P-200101720, from 2002, Spain.
Download the original file
