Why Mystery Shopping Beats Artificial Intelligence and Big Data
(Mystery Shopping Adds the Human Element of Unpredictability)
Mystery shopping beating AI and Big Data? Artificial Intelligence (AI) and Big Data have excited the market research profession with their capability of providing insights that were previously unattainable. This writer does not dispute the value that can be derived from AI and Big Data. There exists, however, the potential of being overly dependent upon these two technologies which automate market research analysis.
I will compare Artificial Intelligence and Big Data against Mystery Shopping: two technology-driven methodologies against a people-driven methodology.
‘Artificial’ implies that there is some ‘artifice’ in AI technology. While ‘artifice’ suggests creativity and intelligence there is also an undertone of deception or falseness. The deception or falseness may not be deliberate but may be the result of pre-existing biases or assumptions put into the initial creation of an AI program.
AI should stand for ‘Automatic Intelligence’ because what is really being sought is a way to automate analysis and predicting human behaviour. If this is the case, then the use of AI is risky. People behave the way they do for complex reasons which are not easily understood. An element of unpredictability and change exists in human behaviour: AI may not be able to adequately handle this complexity.
AI is as much a chimera as the mock turtle of Alice in Wonderland.
People do not behave in predictable patterns like machines or robots.
Big Data’s main selling point is that it can provide predictability based on past and current behaviour. The implication is that past predicts the future, or the past repeats itself. The users of Big Data have control over their customers in that they can supposedly predict future behaviour. Control is the operative word.
Like AI, Big Data does not consider the unpredictability of human behaviour. The past does not really repeat itself, although it sometimes rhymes. Accordingly, one must be cautious about the claims of Big Data as a game changer.
Mystery shopping is a data gathering methodology that has many advantages over AI and Big Data. Here is a short list of its advantages:
• Mystery shopping makes no assumptions about human/consumer behaviour: the past cannot be depended upon to predict the future;
• Mystery shopping is conducted directly at the consumer level at the time of transaction, while AI and Big Data depend on data is gathered in the past and may be outdated;
• Unpredictability and change are the assumptions behind mystery shopping, while AI and Big Data likely work best with data that is static and does not change appreciably. Accordingly, mystery shopping is better suited to uncover changes in consumer buying habits; and
• Mystery shopping reduces the risk of bias.
The conclusion is that while AI and Big Data provide the capability to valuable insights, dependence on these technological innovations may lead to inaccurate conclusions based on past behaviour or programming biases.
Mystery shopping provides data that is current and is less likely to be biased.