The Gist

  • Sourcing data. Complexity increased with abundance of sources, but few organizations leverage existing data for transformation.
  • Defining problems. Involve business leaders, start with goals, and build data strategy backward. 
  • Articulating strategy. Communicate effectively, establish ownership, governance and connect to cloud.

Without question, data is not only the life blood of the modern enterprise, but also the key building block for digital customer experience and digital transformation. Given this, the question is do CIOs and other technology leaders have a clear strategy for how to harvest the data needed to drive self-service, customer experience and business transformation?

The Complexities of Sourcing Data

In the old days, the data an organization had was largely determined by what was in a source system. My question for CIOs and other data leaders was how have things changed? CIOs answered based upon their organization’s data maturity. ClOs were also clear that the complexity of data has increased.

Deb Gildersleeve, First CIO said, “There are many more source systems today as well as many sources of outside data that can be purchased or created.” New Zealand CIO Anthony McMahon added, “What is considered source is no longer a singularity. This is partly due to the integration between multiple systems, including external. This has allowed companies to access a wider range of data.”

Unfortunately, “Too few organizations are working from the business opportunity or problem backwards to leverage existing data, create new data, or add external 3rd party. Hence, data science teams are always asking for all data to locate value within the existing source system data,” said former TD Bank Enterprise Architect Craig Milroy. Mevotech CIO Martin Davis added, “The value of context is often undervalued. For example, manufacturing line data may be very different depending on outside temperature and humidity. But yes, bias as well, comes into the context of data. The abundance of data sources both inside and outside the company.”

Meanwhile, Miami University CIO David Seidl claimed, “sadly, our vendors don't all seem to have realized this and presume they're the sun that everything revolves around instead of planets in an ecosystem. Fortunately, there are plenty of APIs that you can use to interconnect them. This helps. Many also ensure their APIs have reasonable lifespans. The question, for me, is what is the middle and how do you manage, define and govern data from that ecosystem? I think we're doing better about thinking about data lifespans. But this doesn't mean we're doing a lot better with it, but we're at least trying to think this way. This means data must survive past product and vendor lifespans.”

Without question, data governance matters. According to University of Michigan CIO, Carrie Shumaker, “In the old days, IT governance and data governance were often handled by the same groups, because there was 1:1 correspondence between the system and the data. Today we separate them because the data is much bigger than the system.”

At the same time as business ecosystems have grown, data leaders are no longer just talking about one division or one company's data. Pedro Martinez Puig, head of Americas technology at Edenred said, “First focus on what is at hand, if you cannot manage your own data, you won’t be able to manage external data. Today’s reality said that the power of combining external and internal data is where the treasure lives. This includes data from multiple sources inside, IoT, external contextual data, to social media data. All of this can be put together to form a richer picture.”

Additionally, Constellation Research’s Dion Hinchcliffe said there is “data privacy, ownership, control, and ethics, as well as how it relates to AI are becoming a top issue. Especially how public large language models find, ingest, and surface, and re-use personal data. The amount of ready-available external data has exploded, from the Web, social media, and sensors/M2M/IoT/cameras, etc. There is, also, now a sea of derived data in analytics, BI, AI/ML systems.”

Related Article: Challenges or Opportunities? Maximizing Customer Data to Thrive in 2023

Defining Business Data Problems

How should data leaders involve their business in defining the problems data should solve? This clearly starts with business strategy and moves to initiatives and actions. Business leaders need to ask what data we have and how it should be used to solve  business needs. This should then define the data strategy. Shumaker said, “Data leaders should start with their overall business goals and work back from there. If your goal is 'X,' how can data help solve that problem.” Similarly, Pitt suggested, “Start with the KPIs that the business wants to measure and existing reporting that can be done more quickly. Build onto this to increase complexity as the business begins to understand the value of data strategy.”

CIOs believe a great starting point for the discussion is a data workshop. As a goal, the workshop should get key stakeholders to share outcomes of use cases that would drive the data strategy framework. This includes tomorrow’s business challenges more so than past solutions. McMahon argues for “workshops and deep conversations. They should help the team understand where they are today and where they want to be in the next period. The question is what holds the organization back, and what will move the dial. This will usually be one or more of data, people, platform, or process.”

As a goal, data strategy should include improving data maturity. Maturation clearly happens at widely varying paces. So, ideally data leaders should engage with partners to understand their needs, what data they have, and look for ways to provide expertise and tooling to help them. It is important to look for business pains including customer churn, fraud, lead generation and waste. Some of these clearly will convey a sense of urgency. Given this Hinchcliffe suggests data leaders have frequent and clear communications out of IT to find data stakeholders not known in the organization. The CIO likely knows the usual suspects that care and can draw them in as participants and sponsors.

Finally, Capgemini executive Steve Jones said, “Start with the basic questions. Do you want to make decisions based on reality? Then understand where having an accurate view on reality helps deliver real business value. All data is meant to be reflections of the current and historical business reality. The question is then the value in how accurate that reflection should be.”

Related Article: Creating an Agile Customer Data Strategy

Learning Opportunities

How Do You Articulate Your Data Strategy?

Being an effective communicator is part of being an effective leader. For Seidl, this is all about CIOs “figuring out how to think about data context. It's no longer just I have the data, it's this data's context is, and it was gathered in the following way and locations, and has bias, gaps, or insights because.” This means recognizing the importance of adding external ecosystem data, McMahon said. “A data strategy isn't a walled garden; it is an enabler of business goals and outcomes. The key is the ability to tie data to strategies where practical. I'm working on building a strategy where data and technology are layers in the strategy, not stand-alones.”

Davis adds everything “starts with ownership and governance. Who owns different aspects and how is it controlled? Then how is that data used to satisfy business needs? This then needs to be widely communicated and agreed to, so a lot of change management is required.” For this reason, Gildersleeve said, “Tie the data strategy to the company strategy or group strategy where you can. Examples usually help.” Puig concluded by saying, “I am giving data a prominent space in our external communication from annual reports to investors days. I am committing from the top and making it as key dialogue from CEO to all the leadership lines and making it a transversal priority. It is a lever for me to keep a promise.”

How Do You Connect Data Strategy to the Cloud?

Clearly, a lot of amazing data work takes place in the cloud. Pitt said, “from where the data is located to the increased scale then to its movement across systems, complexity has increased ... as well as our community's understanding of who is the trustee for that data and who allows access to that data.” Pitt continues by saying, “Start with data governance, assuming you have good data governance. Your partners in this process should have the best understanding of data strategy. Clearly, you only do this if it makes sense. So, it's only connected if it's a part of your data design.”

Seidl added, “The cloud is worth discussing at many levels. Can you get your data out? What happens if your vendor goes away quickly? Are your eggs all in one basket? Do you have a lifecycle plan? Of course, those aren't all unique to the cloud, but they're something that you really must think about. Similarly ask, will a feature be there in six months? A year? What happens if you do something wrong at scale and pay for it. Are you seeing this get better for spaces you can, see? I've seen latency generally get better as more regions and direct connect points become available. At the scale of data instead of automation, it seems viable in many cases from my side. I'm pondering building and facility lifetime length support for building automation, fire alarms, and similar things for campuses. Technology moves far faster than embedded infrastructure! I'm not sure it's unique value from cloud databases. I've seen cloud infrastructure and service designs allow scalability and flexibility that is hard to provide in-house for organizations our scale and smaller. But it's not because it's cloud.”

Hinchcliffe concluded by saying, “There are two major issues with enterprise data and the cloud. These include:

  1. All our business data is spreading out to the cloud.
  2. Our data is better when it's with everyone's else's data. Fusing third-party cloud datasets to enrich ours.

In most organizations, data strategy is fragmented among data owners or IT systems owners or a combo of both. This is why the chief data officer role or data center of excellence has needed to be established to break parochial data out of its tightly controlled villages and make it part of business strategy.”

Who is the Primary Person Responsible for Defining Data Strategy?

CIO answers varied by organizations, their data maturity and organizational structure. Common answers included:

  • Chief Data Officer
  • Head of the Business Intelligence Team
  • Enterprise Architect/Solutions Architect
  • Enterprise Architecture
  • CDAO
  • VP/Director of Analytics
  • Lead of the Data Management
  • Someone in the Office of the CIO
  • Lead Data Strategist

Jones is right in saying, "Most companies have lots of pockets of this, then someone in IT pretending they’re unifying it all.” Puig said, “It is our Chief Data Officer, who is part of the technology leadership who champions the process. She is responsible across all our business lines, as the magic comes both from focusing on use cases related to each market and then finding correlations among them.”

Parting Words on Data Strategy

Shumaker said during the conversation, “I'll be the annoying person and ask if data strategy is different than business strategy? You look at your business strategy and say, what are the ways we need data to support these ends? That doesn't solve your data architecture and tool strategy, but it does inform it.” Clearly this perspective is correct — data strategy needs to be driven by business strategy. It is also correct that CIOs need more direct involvement in data strategy than even in the past. And having a CDO, does not mean an abdication of this responsibility.

fa-solid fa-hand-paper Learn how you can join our contributor community.