top of page

Predictive algorithms.

VOL. 0035 – FLORIDA, MONDAY, OCT 9, 2023





Juan F. Arjona Harry

President & CEO Strategee


In the current marketing, it is inconceivable to generate strategies, tactics and campaigns that are not based on the mathematical analysis of the consumer buying behavior.


Gone are the days in which, with simple segmentation criteria, the company issued bulletins, sent emails, designed promotions, developed cross-selling tactics sent to its entire database.


For various reasons, among which are:


1. Increased effectiveness regarding redemption rates.


2. Better supply chain planning.


3. Greater precision in the allocation of marketing investment resources for each segment and customer.


4. Best accuracy in terms of metrics associated with each strategy, and campaign tactics.


5. Increased likelihood of building Brand Equity.


6. Best performance in alliances with suppliers and allied companies with which to work for cross-selling or promotions.


7. Increased marketing revenue.


It is vital that marketing consider today a strong link with a robust platform in terms of mathematical modeling, which allows the achievement of the above objectives in its execution. When working with the database, there are dozens of trails or traceable curves on the purchasing behavior of customers and / or consumers, which are shaping themselves iterating cycles, which when analyzed by formulating an interpretive algorithm, probably leads to the conclusion that buying behavior is repeated, in this case, the buying decision may be impacted until modified by an action of promotional marketing or tactics indicated as cross-selling with sufficient power to stimulate it to become a purchasing decision.


The possibility offered by the technological advancement of software development must be accompanied by the mathematical formulation that allows to read the transactional iteration of the DWH, until the formation or algorithmic validation, so that with the addition of features based on the rules of purchase decisions (compensatory and non-compensatory), plus an addition of external macroeconomic variables can be designed precisely a predictive algorithm that enables the extraction of a section of records from the data warehouse of the company and on it, deploy indicated pacing power of cross-selling tactic or a promotional tactic, which can noted a higher probability of success in implementation.


This is what is now required by the current context. Working with previous general methods, which could be used indiscriminately to implement tactical actions to stimulate the demand, is no longer feasible within a marketing model that increasingly must seek investment optimization and rule based its management.

0 visualizaciones0 comentarios
bottom of page