Bergson Lopes, CEO and founder of BLR Data, a pioneering Data Governance consultancy in Brazil, says organisations launching AI initiatives without first assessing their data maturity risk misdirected investments, failed projects and strategic blind spots.
Topics related to data management and governance have become a constant presence on corporate agendas. Still, organisations constantly initiate ambitious programs, acquire sophisticated platforms and announce Artificial Intelligence initiatives without answering the most uncomfortable and essential question of all: what is, in fact, the company’s current situation in relation to the management of its data?
The data management assessment should be the natural starting point of this process. It exists to replace guesswork with evidence and build an objective reading of how the organisation really treats its data on a daily basis and not about how it believes it treats it. However, in many companies it continues to be treated as an accessory step or worse as something dispensable.
The result often leads to misaligned expectations, misdirected investments, initiatives that do not scale, constant rework and AI projects that fail before they even generate real impact for the business.
In addition, the developments of corporate strategy which depend on data are now guided by fragmented perceptions, politically convenient discourses or generic benchmarks, almost always detached from internal reality.
For this reason, the data management assessment should preferably be carried out at the beginning of the Data Management Program, serving as a basis for any structural decision, relevant investment or prioritisation of initiatives. The same reasoning applies to organisations that intend to advance in the use of Artificial Intelligence in a structured and sustainable way. There is no point in discussing AI, scaling models or automating decisions without a clear understanding of real maturity in data.
In addition, the assessment is not limited to the initial moment of the journey. Whenever the organisation needs an objective reading of its current situation, whether to review directions, correct deviations, redefine priorities or plan the evolution of the AI program and initiatives for the coming years, a new evaluation becomes not only recommended but necessary.
Data and AI programs are dynamic and the absence of periodic diagnoses turns strategic decisions into bets, especially in contexts of growth, Digital Transformation, regulatory changes or increased analytical complexity.
A serious assessment in data management is a structured process which confronts the reality of the organisation with consolidated practices widely used by experienced professionals in the area. It objectively evaluates the data management and governance practices that are or are not executed in the company.
It goes far beyond technology, encompassing processes, roles and responsibilities, policies, architecture, quality, engineering, infrastructure, security, privacy, intelligent use of data and other data functions widely recommended by the market.
This diagnosis cannot be constructed from generic forms or standardised questionnaires. A consistent assessment requires qualified interviews with technology and business professionals conducted by experts capable of dialoguing at the same level as these interlocutors.
The questions need to reflect the concrete reality of the performance of these people, their processes, responsibilities, restrictions, goals and daily dilemmas. Generic questions produce generic answers and useless diagnoses.
More than that the diagnosis needs to capture, in a structured way, the real pains and expectations of the business public. Operational bottlenecks, difficulty in accessing information, low data reliability, delays in critical decisions and frustrations with analytical initiatives need to be made explicit.
The true value of assessment lies precisely in connecting these pains and expectations to the level of maturity identified in the best practices, highlighting points in which fragility directly impacts business results.
Consistent diagnosis is essential
In this context a central and often ignored question arises: how could someone with no real experience in assessments craft the right questions and properly interpret the answers? Developing a consistent diagnosis requires repertoire, experience, practice and deep knowledge in identifying contradictions, silences, defensive responses and natural biases in interviews.
Without this baggage work becomes an educated checklist, unable to reveal risks, weaknesses and structural inconsistencies.
At a minimum an assessment needs to deliver three fundamental elements. The first is a clear and unequivocal diagnosis based on evidence and connected to the pains and expectations of the business, with the evaluation of existing or non-existent data management and governance practices in the organisation.
The second is the strategic direction of the Data Management Program explicitly aligned with the company’s corporate strategy. The third is a structured, prioritised and executable action plan capable of evolving data maturity progressively and sustainably.
There is no AI maturity without data maturity
Here is a point that needs to be faced directly by the C-level body: the assessment is not a technical exercise nor an operational initiative. It is an instrument of corporate governance and support for strategic direction.
It allows executives to clearly understand where the organisation consciously takes risks and where it simply operates in the dark. Without this diagnosis relevant decisions about data, analytics and AI will never be strategic and will continue to be ill-informed bets.
When internal views arise usually at the tactical or operational level arguing that the assessment is not necessary or that the company already knows its own reality the alert to top management must be immediate.
This kind of argument rarely reflects maturity. Most of the time it reveals a limited vision excessively operational and restricted to the context of the area itself. Experienced executives know that organisations do not fail because of a lack of initiatives but because of too many poorly grounded decisions. Ignoring an independent diagnosis is a classic sign of strategic myopia.
This negligence becomes even more serious when the corporate discourse involves Artificial Intelligence. There is no maturity in AI without maturity in data. No advanced model, modern architecture or robust budget compensates for data that is poorly governed, inconsistent or lacks clearly defined accountants.
The Data Management assessment is a determining factor for the success or failure of AI initiatives. Companies that ignore this step build solutions on fragile foundations and when the results do not appear attribute the failure to technology when the source of the problem is structural.
The impact on regulatory risks
There is also one point that executives cannot neglect: regulatory, reputational and accountability risks. Without clarity on how data is governed the organisation exposes itself to increasing risks especially in regulated and data-intensive environments. The assessment brings visibility into these risks before they materialise.
Another critical factor is in the way many assessments are conducted. The market is full of generic diagnoses based on superficial questionnaires or visually attractive tools but fragile from a methodological point of view.
Most of them exist more to justify the sale of solutions than to guide strategic decisions creating a false sense of maturity and postponing structural problems when the cost of correction becomes much higher.
Therefore, a Data Management assessment must be conducted by an external, specialised and independent company. Evaluations made exclusively by internal teams are hardly exempt as they are directly influenced by political pressures, organisational disputes and local interests.
Without independence there is no reliable diagnosis. And without reliable diagnosis the board decides in the dark.
Assessment is not a luxury, it is not a formality and it is definitely not an expense. It is a mechanism for generating value, reducing risk and sustaining data and Artificial Intelligence initiatives. Ignoring it or replacing it with convenient versions is not boldness. For responsible executives this has another name: strategic recklessness.
