Data, the heart of a Digital Transformation strategy
The emergence of the COVID-19 pandemic had a strong impact on the Spanish economy, leading to a marked drop in activity, especially in those sectors most affected by isolation and reduced mobility.
Particularly, the Spanish productive fabric is dominated by SMEs, which also, by their very nature, achieve a lower degree of digitalization than large companies. This scenario meant that they were at a clear disadvantage in a context where the highest degree of digital penetration was key to competitiveness.
The need for a digital transformation is rapid and profound. The Plan to Promote the Digitalization of SMEs 2021-2025 involves a set of public initiatives that aim to promote the adoption of new technologies and the digitalization of companies. This project is in line with the Recovery, transformation and resilience plan which foresees that, in the next three years, Spain will receive 140,000 million euros from the European Union within the Next Generation EU stimulus package. According to the forecast, around 30% of the funds will be allocated to digital transformation.
The opportunity to promote and consolidate the great promise of digitalization is unique. However, digital transformation is on the agenda as an aspiration that requires a concrete anchor: How to start? How to approach a process of this nature?
To begin, I would like to talk about “data”, which is the heart of such a transformation. More specifically, I would like to refer to “Data Science” projects, because they get even closer to the business objective. This is because facing a Data Science process involves managing data in such a way as to be able to make dynamic decisions that benefit businesses. It begins to carve an interdisciplinary field that involves scientific methods, processes and systems that extract knowledge from data with the aim of analyzing the current situation, predicting the future and making the most opportune business decisions.
Every data science project follows a process that we can summarize in 6 steps:
- Defining a problem: Translate the business problem and identify data source. We must be very clear about the problem we want to solve, so we have to ask ourselves: What is my main objective? What business problem do I have? What do I want to explain using data?
- Data preparation: Select useful data and extract it from its sources. There is a central question here: How much customer history do I have saved? Who is the owner of the data?
- Data study: Cleaning and transformation. Analyze the variables to understand their behavior and relationship. A data-oriented culture requires systematic decision-making based on a “cult of data.” It is vitally important to have equipment aimed at this purpose.
- Model creation: Create the model and train it. Once the model is built, it is possible to predict reality from any available information. Machine Learning is the great revolution for these processes: the use of computer algorithms allows models to learn automatically through experience.
- Validation and testing: Adjust parameters and evaluate the model through trial and error.
- Visualization: Display the data using appropriate visual tools.
Without a doubt, there is no better time to drive a data science process than now.
The process described is only a first approach to a process that involves investment of time and dedication to understand its steps and implications.
I will leave for future installments the challenge of stopping at each of these phases, deepening their scope and demands.
For now, it is important to be clear about the core of this type of project. The biggest challenge is not in obtaining data but in how to extract meaning from it. A team up to the task and a systematized analysis are the key for this process to allow the business to achieve the best decisions and achieve higher and better levels of competitiveness.
Julio Cesar Blanco – March 22, 2022
