In this forum, we’ve recently explored different aspects of achieving Data Intelligence: why forward-thinking companies need to create a Data Office, what kinds of ‘products’ might come out of that initiative, how data can play a critical role in guiding business decisions in uncertain times.
But here’s the key: none of this works without concrete proof. If the Data Office is to help the organization move toward a state of Data Intelligence, there must be metrics to measure that progress. We all understand this can’t happen overnight — the constant influx of new data, in new formats and from new sources, all being collated and analyzed in different ways, make for a dynamic discipline that requires ongoing investments of time and resources. But this is also why we need metrics that are tangible, transparent and comprehensible; anything less can seem like a vanity exercise.
It’s the most recent twist on an old problem. There used to be stories of how companies saw marketing as an infinite sinkhole until CMOs were able to draw a straight line between budget outlays and revenue growth. IT executives had it easy for a while — technology innovation got everyone’s attention and resources to match. But eventually, the ever-expanding budget drew scrutiny, and metrics were established to determine which upgrades or new releases would actually boost productivity.
Like every discipline, a multi-dimensional Data Office has particular characteristics that must be considered. For example, different departments can make different uses of the same data. The line between data and analysis, or even data and technology, is sometimes blurred. Most importantly, this is a specialty that has developed with astonishing velocity—in less than a decade to go from almost no Chief Data Officers (CDOs) in corporate America to 10,000 this year, most of them appointed quite recently, all representing significant diversity in expertise and experience. Maybe we haven’t had time to develop function-specific or even industry-based markers for success.
Success metrics for the Data Office
It’s time to measure how successful the data profession is in turning data into value. And since there’s no universal methodology, here are some possible avenues.
Building on what’s already in place is one option. A Harvard Business Review piece describes a medical data center that was clearly faltering with one-off projects lacking alignment with business priorities. The CDO matched specific data initiatives with a list of business scenarios where data could deliver value. These so-called value modes focused on measuring progress with core business issues: ensuring compliance, cutting operational costs, enhanced provider data across multiple clinics, easier patient access and better patient care.
Another starting point might be the business case organizations had to make for investing in a Data Intelligence function. It typically includes:
- A financial template attached to each use case
- Identification of the processes and resources required
- Cost estimates for technology
- Professional services
- Training for employees
Any projected ROI in those calculations could become the foundation for a new set of metrics.
Some organizations have developed specific markers for success. LexisNexis Risk Solutions — which is actually in the business of data distribution — has identified four specific metrics for its data operations, all aligned with the overall data strategy:
- Enhance product development
- Offer intelligence on customers
- Competitors and markets
- Identify the potential for advanced analytics
- Focus on quality
Without going too deep into specifics and therefore too narrow, it’s possible to set some general guidelines. They can’t be too meta — while the metrics need to highlight the data and analytics programs themselves, Gartner recommends that the organization should also measure how they affect business processes, along with the impact on the different stakeholders.
Choose the right metrics for your data office
To develop quantifiable metrics that apply to business impact, start with three topline areas:
- Business Value: Every organization has key performance indicators (KPIs) with particular metrics — how does data move the needle on those?
- Stakeholder Value: Are data assets only accessed by specialists (data scientists, analysts, IT specialists.) or are they leveraged by professionals in different functions and used for collaboration with fellow employees, partners and customers?
- Information Value: What is the cost of specific data analysis initiatives, what is the quality (perhaps as judged through reviews, adoptions, etc.) and what is their speed relative to operational requirements?
How metrics evolve over time
Depending on the maturity of your data office or CDO role you should also consider evolving your metrics:
- Progress metrics: how many data assets did you identify, and how many are fully covered (e.g., in terms of definition, ownership, use, etc.), how many KPIs are fully aligned, what is the coverage of certified data sets, and how many are in process )e.g., bronze, silver, gold).
- Organizational metrics: how many domains do you have, and how many have responsibilities fully set up (e.g., domain owner, data steward, privacy steward), what is the degree of data literacy across parts of the organization
- Business metrics: eventually — and inevitably — you’ll need to connect the dots from data to the business: impact on growth, cost savings and risk mitigation.
Defense vs. Offense
One of the more intriguing aspects of the data discipline is that it can take only one of two distinct paths.
- It may be defensive: ensure that the data doesn’t get misused or abused, and fall out of compliance (a major concern in the era of GDPR and CCPA).
- It may be offensive: the data is widely used and shared by authorized business professionals to support collaboration and decision-making. It’s definitely not a binary choice
No CDO really has the luxury of focusing on only one. But it is true that they cover different areas and have different priorities, which mandate distinct but overlapping metrics.
Collaboration across the organization
Last, but not least — remember that the Data Office does not operate in isolation. As a matter of fact a data office should be anti silo in principle. Relationships to other areas (business functions, HR, finance) are crucial so consider shared metrics with your executive peers.
In effect, there’s no rule for developing metrics, only that metrics should be developed. The concept of Data Intelligence is vital and unimpeachable, and building a Data Office to focus company wide efforts and deliver valuable data products to diverse functions is at the heart of this strategy. But these initiatives don’t flourish in a vacuum — like every budget item, they’re subject to boardroom priorities, market shifts and personal dynamics. Solid metrics not only fortify defenses against these pressures, but also accelerate data-driven collaboration throughout the enterprise and elevate the state of Data Intelligence.