Marketing technologies: the main mistakes in their use
Marketing technology is a powerful tool for business development, but it does not always work the way the company expects. In this case the problem is not in the effectiveness of the algorithm or technology, but in the wrong formulation of the problem.
Unfortunately, this often happens in companies which use systems based on artificial intelligence. The Sloan Management Review surveyed 2,500 executives who use AI technology in marketing. About 90% of them admitted that the effectiveness of these systems within 3 years was less than 40%. Experts believe that the reason for such a failure is the wrong approach, and in this regard they identified the most common mistakes.
Incorrect formulation of the problem. For example, in the case of customer churn it is not necessary to look for those who are potentially ready to leave, but ways to retain them. For its part, the algorithm worked correctly, according to the original goal, but this data did not bring the company a solution for the problem.
It should be understood that AI forecasts have a margin of error, but in one case they can be detrimental and in another they can be beneficial. A prime example: a consumer goods manufacturer has improved its sales forecasting system, reducing the probability of error from 25% to 17%. However, marketers achieved this result through changes for low-margin products, while the reliability of the results became worse for high-margin products. As a result, the manufacturer incurred losses, as the benefit of high-margin products is much higher than the accurate prediction for the low-margin category.
One should take into account the fact that the results of algorithms based on artificial intelligence can be incorrect due to various reasons. At the same time, the marketer should understand the consequences of the error and assess the risks associated with it.
Modern technology makes it possible to process a huge amount of data and use it to generate predictions. However, marketers continue to apply template solutions without using all the capabilities of AI. As a result, the efficiency of implementing new technologies is reduced, and the business is limited in its development.
An important point is the communication between the different divisions of the company. Marketers and analysts must work together to achieve a common goal. For technologies to produce results, it is necessary to have a clear understanding of how to interact with them. You should not try everything in the hope that in the future, specialists will find the right option. Such an approach can be very expensive for business, and mistakes in constructing the forecasts will not only cost financial expenses, but also the time. In this case the algorithm will be given impossible tasks, and as a consequence, the result will not be applicable to the real situation.
