The problem of continuity in data analysis

Andres C.

In these years of experience in the area of data analysis, I have seen a recurring problem in all organizations, public or private: whenever an analyst leaves, they take with them a large part of the intelligence collected in their analysis.

With the fluidity of relationships and the natural turnover of employees over the years, the fate ends up being the same, much of what is arguably the most valuable asset in analytics departments is lost, whether in public safety or in competitive intelligence departments.

This does not mean that the professional is doing something illegal or unethical, but rather a simple natural consequence of the lack of processes and tools.

The process of building intelligence must be continuous, however, cases are not rare in which frauds are repeated over the years because the analysts who participated in the initial analysis no longer work in the company and the information has been lost. experience in that specific case.

I remember that a few years ago I participated in a case in which a few million reais had been embezzled in a bank receipt fraud. What would be expected in a case like this is for the company to learn its lesson, establish some processes and never fall for the same scam again.

In English, the expression for the difference between what is expected and what really happens is “Wishful Thinking”, something like “Hopefulthinking”, or as they say in the south of my country… but what hope! What really happened was that with a simple change of form, the fraud was successful again.

But what to do?

1. Identify the most important data types and knowledge each employee should have to leverage your processes most efficiently.

2. Establish an agile process such as Scrum and determine that the deliverable at the end of the process is a set of changes to be developed by the other team, whether compliance and processes in cases of fraud or marketing actions and presence in the case of intelligence. business.

3. Acquire tools that document the entire analysis process in detail and make it easy to find information. A tool focused on the area of intelligence, do not expect Word or Confluence to solve your problem.

4. Establish a well-defined workflow based on modern analysis tools and internally define roles for Analysts/Reviewers/Approvers so that more people participate in all analyzes and facilitate knowledge dissemination.

5. Innovate, try new ways of generating and storing intelligence.