Page 11 - AeM_January2021

RESEARCH, ANALYSIS & TRENDS

recommendation system as a commercial marketing
tool and is not alone in the market. Several large and
small vendors are working on solutions to personalize
every interaction on a website. There is a demand for
that since most decision-makers assume that the
personalization of customer approach will be decisive https://www.youtube.com/watch?v=xFZ3HMleYYE
for being competitive in the future, according to surveys (Amazon Personalize)
conducted. The fact is that an AI-controlled
personalization enables an individual customer journey,
not to mention that machine learning draws its
knowledge, among other things, from user behavior
regarding what customers want to buy.

The advantage of personalization does not only lie in
the possibility to enlarge the shopping cart or to
discover new things but helps to make the shopping
process faster where it is not just about rummaging.
For many everyday products, shopping is more of a
routine than an experience, but you still must do it.
Previously, customers had to search for each product
individually in the online shop, select it and place it in
the shopping cart. Studies confirm that it often takes
longer to fill an online shopping cart with groceries than
to do the shopping in the supermarket, even if you
include the trip there. AI-based predictive baskets
provide a remedy here by analyzing individual
purchasing behavior and automatically detecting trends
in the behavior of all customers, so that forecasts can
be made about what the respective customer likely for specific applications and just look to see which
wants to buy today. answer in each list best fits the question. In contrast to
such rule-based bots, semantic AI bots can understand
Even offline, the intelligent personalization will advance the meaning of the question entered and derive
since with the help of sensors billboards will be able to answers from larger amounts of data. Semantic AI bots
recognize whether a viewer is sad or happy, young or do not work with question-answer assignments, but
old, and whether a person is in a hurry or has a lot of
time. And with every reaction of a viewer to an rather try to analyze the context of a voice input. So, we
can expect that instead of comprehensively trained
advertising message (“continues”, “stops”, “scans QR support staff, highly specialized staff will be needed in
code”, “smiles”) the system learns to display the right the future, to whom the bot only refers to if it cannot
advertising. Retail displays could recognize which solve the problem on its own. However, although
specific item of clothing on a stand attracted the Artificial intelligence is still in the early stages, and the
customer, which products were viewed afterwards, and big changes are yet to come, it will obviously bring
show corresponding recommendations.
enormous growth opportunities for companies.
AI-supported dialogue marketing
Besides foresight and know-how, integration into
corporate marketing requires the willingness to gain
In addition to personalization, automation is currently valuable new experiences and to learn steadily.
the most important use of AI in marketing. AI-supported
dialogue marketing, for instance, is not only becoming Therefore, employees should be directly involved in the
more effective by constantly learning new things using implementation right from the start, as the change
previously unused data. In the field of applications process through self-learning systems usually creates
existential fears. It should become clear, however, that
known today, chatbots are probably the instrument that
comes closest to the science fiction scenario of a AI will not replace and lay off employees, rather give
service robot. Important to know is that chatbots do not them more time for new tasks. ◊
work with "real" AI per se: most bots are programmed By Daniela La Marca

January 2021: AI & Automation in Marketing 11