Solving Cleanup & Integration Challenges for Data Driven Success

In an emerging digital economy, data is celebrated as the foundation of success. The transfer of knowledge from experienced individuals to algorithms has made data models more reliable. Experts are now contemplating how data cleanup, standardization, and integration efforts can enhance data-driven decision-making and actionable insights for business leaders.

“Knowledge that was once held by individuals with decades of experience has been transferred to algorithms in the form of data inputs, making business models more reliable,” said Salvador Hernández, CDO, Frisa Forjados.

A Denodo study, as previously reported by MBN, uncovered a striking reality: only 54% of Mexican companies effectively harness their data resources, emphasizing the urgent need for streamlined data solutions. Surprisingly, while 90% of companies are in the midst of their data-driven transformation, just half are actively using data for informed decision-making, and 12% have not yet embarked on this journey. The study also reveals diverse objectives among surveyed organizations, including improving operational efficiency, gaining a competitive edge, reducing errors, and minimizing operational costs.

However, before companies can hope to realize these objectives, companies need to understand that the quality of data is an important qualifier. Whether for advanced analytics or predictive analytics, data quality is a fundamental component to an effective data strategy. According to Luis Pintado, CTO, Interprotección, "the insurance industry is undergoing a significant shift from relying on instinct to engaging directly with consumers through digital channels. This transition has prompted a need for more data-driven strategies.”

Despite ambitions to become more data-driven, use cases often remain isolated, according to Ana Coronel, Data Science VP, BanBajio. To expand them, it is essential to not only focus on data quality but also consider when and how data will be used to avoid wasting time cleaning data. She also considered comprehensive data governance, from inception to exploitation, to be vital to strategy formation. Without procuring these efforts from the beginning of strategy formation, companies’ risk undermining their  ability to monetize their data.

"Data quality is paramount to building any data model, be it for advanced analytics or predictive insights. We all have a role in the lifecycle of data, and even a minor error, like an incorrectly written tax identification number (RFC), can have far-reaching consequences," said Coronel.

Experts also shared how hyper-segmentation of data is vital for enhanced personalization, a capacity that has applications across industry segments. However, disorganized and poor data integration could lead to skewed investments in digital advertising, ineffective conversion rates, and challenges in targeting the desired audience. Herein lies productive capacity offered by disruptive technologies like machine learning, which could potentially expedite data clean-up and standardization, says Pintado. Nevertheless, no process should be left to automated models in its entirety. Human supervision will continue to be an integral part of data analysis, emphasized Hernández.

The application of emerging technologies in data analysis becomes relevant before a global shortage of qualified talent to clean, analyze and oversee data strategies. Moreover, continuous personnel turn-over can undermine the continuity and the viability of data initiatives. These issues pose a significant challenge to the medium and long-term data maturity of companies.