The challenge of the newly appointed CDO: how to build a data governance policy from scratch

Articles, cdo, data governance, 3 October 2019 |


The other day, September 26, at the CDO Day event, I had the great luck to participate in the panel of experts entitled “The Challenge of the Newly Appointed CDO: How to Build a Data Governance Policy from Zero” together with Julio Valero from Banco Santander, Manuel Ferro from Abanca and Jesús Armand Calejero from Funidelia.

The debate that emerged was very enriching and the experiences that were shared will surely serve more than one of the attendees who have the difficult pathway ahead of them to implement a data strategy and governance model in their organization. For this reason, and also at the express request of several people who were unable to attend, I do not want to miss the opportunity to share the main keys given during the session, expanding on them and including my point of view in each of them.

I hope these lines work as ideas and guidelines for those who may be looking for answers to their questions within the data governance field.


Data culture and communication

With no doubt, the main point to take into account is the data culture that exists in the organization at all levels…

  • “Are there specific departments or roles focused on data?”
  • “Are data-related goals included in corporate and employee objectives?”
  • “Are there policies and procedures focused on the use and treatment of data?”
  • “Do senior management, middle management, and other employees consider data as a strategic asset of the organization?

The maturity of the organisation in terms of data understanding will be the starting point for our governance model definition and both the culture and profile of the people who make up the organisation will indicate how quickly we will be able to go on with change management or where we need to put effort when undertaking initiatives.

Like any initiative that starts from the culture of the organization and its employees, it requires a very important communication strategy, focusing on reaching all levels and giving visibility and importance to the changes that are undertaken, as well as involving every stakeholder.

Data management has 3 axes: technology, processes and people; of which the most important and complex element is undoubtedly people, since they are the ones who must be impregnated with the data culture we want to establish so that they do things the way the organization expects them to do on a daily basis to help achieve its goals.

Involvement of stakeholders

In line with what is said above, the people involved have to feel empowered to make decisions and see themselves as part of the process of change, perceiving the value of it while assuming a series of responsibilities around the data that perhaps they did not had until now.

It is very important to make stakeholders see that data governance is not about “putting sticks in the wheel” or “playing bad cop” but about providing the organization with the necessary capabilities to generate business value through better use of data. In order to do this, it is also essential to provide them with the necessary tools and resources because without them, their workload will increase and they will not see a quick return.

Formulas that may work to achieve this involvement can be the inclusion of objectives or bonuses in relation to data governance initiatives, participation in committees or meetings of importance, visibility to colleagues and managers or the granting of space for decision making.

Likewise, it is key to create a collaborative and interactive environment where data governance is built with the sum of all and not only with the involvement of a few roles and/or areas since this causes frustration in some and misunderstanding in the others.


Implication of the Top Management

As it cannot be any other way in any cultural change for an organization, the Top Management must be the first one involved in the initiative because at the end it is the one that is going to make the strategic decisions and priorities that are going to appoint where the resources are allocated to face them with guarantees.

In this case, it may be understood as one more stakeholder that has to have its presence within the whole governance model with its role and functions but also its support is essential since it has to provide the necessary power to those who require it and they are the first ones that should follow steps to expand that data culture to the whole organization.

It is not easy to measure the ROI of a data governance initiative in the short term but it is necessary to sell it properly to achieve the required involvement from stakeholders in case that Top Management does not decide of its own to launch it with all its effects. This participation is really necessary because otherwise we will not have the corresponding resources at our disposal to achieve our goals.


First steps: the assessment

How can I be able to come up with metrics that will make the internal sale of this kind of initiative if I don’t even know where I’m coming from?

The first step after we are clear that we want to implement data governance at all levels in an organization is to know where we are starting from.

Just as we must know the existing data culture or the resistance to change that we may find, it is very important to know what we have, what we want to achieve and which tools we have to reach it.

A good initial analysis will not only provide us with a detailed roadmap and with a greater probability of meeting our goals, but will also make us capable of measuring our KPIs before we start. Then, we may compare them with those measured at the end of the exercise, thus obtain improvement results.


The importance of the use case

Once we have landed an initial assessment and have the As-Is vs To-Be approach at a high level, the next thing we need to do is to select a use case and set a time limit for measuring results. This step is very important since it will allow us to get a fast time-to-market with a limited scope and we will be able to measure what we have achieved in order to sell it internally.

One of the keys to making our use case successful is to choose an appropriate and striking one. The choice of this use case is not trivial and depends on many variables that are conditioned to a greater extent by the culture and degree of maturity of the organization.

For example, we can propose the following scenarios for the choice of a use case:

  • Select some of the reports that are most interesting to the Top Management and define an end-to-end governance model of the generation process and the data that is represented in it. This has the advantage that we can achieve a good internal sale since it is an important use case by definition and it is surely very extrapolated to other areas because of its cross-wise nature. However, we can also fall into the trap of having chosen a complex use case, with a too wide scope and that has too many participants, who are also going to be very exposed to the organization’s management.
  • Select some area of a particular department, so that we reduce the scope and number of participants but we may also be falling into the trap of doing something very ad-hoc and that we cannot extrapolate later to other areas as well as choosing a use case that does not have much impact for the Top Management.
  • Finally, we can look for something more innovative and jump on the bandwagon of a new strategic data initiative, be it the implementation of new technologies, the development of algorithms with advanced analytics or the capture and processing of new sources and types of data. Like all of the above, this one also has its pros and cons. The good part would be that we would have the support of the Top Management and we would get the involvement of the participants quickly as it is something strategic, besides that starting with the initiative from zero we could go hand in hand and achieve milestones quickly. The negative side would come from the difficulty of measuring the results obtained from the implementation of the governance model in the initiative, since we could not measure the status before and after and we would have to find another way to get these metrics.

Whatever the case of use, what is clear is that we need what we do to be as extrapolated as possible to other areas and cases of use in order to be able to extend this governance model to the whole organisation, impregnating it with that data culture that we talk so much about.


Metrics to measure results

We have commented it above and throughout the article but it is simply because of the importance of defining and measuring the metrics that will allow us to achieve an internal sale of our initiative.

The problem with metrics in data governance initiatives is that it is not easy to define them since in most cases we cannot obtain a direct ROI in monetary terms.

Only in those cases where data have an economic value for our business or directly impact our P&L, we will be able to get that ROI directly but in the rest of the situations the most common thing will be that we have to infer that economic impact from time reduction, errors decrease and costs savings and also taking into account the level of satisfaction of those involved with the new way of doing things.

In addition to the metrics that help us with the internal sale, it will also be very important to define the metrics corresponding to the follow-up of the initiative and goals achievement. These metrics have to cover mainly the following aspects:

  • Degree of coverage of the governance model (completeness of implementation)
  • Stakeholders involvement (in quantity and quality)
  • Data quality (basic controls)
  • Time invested in tasks performance (dedication in their day to day of each team)
  • Identification of “bottlenecks”


Technological tools and solution

They are not something essential from the beginning and it is important to know when to incorporate them into our model but they will certainly help us to achieve our objectives and above all to achieve the involvement of every stakeholder as soon as we include them in the picture.

To achieve this participation it is essential that tools include a high degree of automation of manual tasks and that they offer a neat and intuitive UI & UX where the learning curve for their use is not an added problem.

In this sense, technological solutions have to be understood as accelerators and facilitators of change.

It is also important to understand that the implementation of a tool is not the solution by itself and that we will need to integrate it into our ecosystem, so it is also essential to choose it based on our variables and make a detailed study of the options available.

Variables such as configuration and adaptation capabilities, scalability and interoperability are essential in a data governance solution in the current era, where it is very important that we can opt for solutions that are totally agnostic to the technology we have for data processing and that we can adapt to our governance needs as they evolve over time.

With this scenario, it is reasonable to consider whether the best strategy is to opt for a specific market solution offered by a software vendor, to carry out in-house developments with our own or external personnel or, finally, to reuse tools available in the organisation with this approach. Normally, the choice of one of the options is not exclusive of the others and what is usually considered is an environment where different pieces of different types fit together to form a puzzle that integrates with our vision, which is ultimately what is important.



To sum up, one of the most remarkable aspects we could extract from the session was that there is no whitepaper or instruction manual fully applicable as the ones that can be followed by someone who wants to comply with a specific regulation or install a software. Rather, it is a collection of good practices, implementation guides and experiences that must be transferred to the specific scenario of each organization, trying to align what is applied to the corporate strategy.

In line with the above, it is also clear that there is a difference between companies in different sectors and of different sizes, mainly due to:

  • The regulations and standards they have to face
  • The amount of data they have at disposal and the ease of obtaining new data
  • Investment in technologies capable of handling large volumes of data
  • The importance of data in their strategy and the competitive advantage they can achieve by using it
  • The direct margins obtained through the correct use of data within their particular business activities.

For a new CDO, implementing data culture and data governance in your organization is a major challenge and not easy, but there are more and more mechanisms and information that can help you achieve the proposed goals. I hope that this little article will serve to shed some light and guide the walkers lost in the forest as the anjanas do…

About the Author: Mario De Francisco

CEO de Anjana Data