Over the last 20 years through my involvement in the world of data and analytics, a few key challenges continue to repeat themselves over and over.
As drastically as our business needs and supporting technologies have evolved over these years, the way data leaders, consultants, and analysts approach these recurring problems largely has not. As a result, we fail to make progress toward our goals — and more importantly, we often fail to legitimize the transformational role that data can, but often doesn’t, play in companies today.
How do I know this is true? In the spirit of being data-driven, I’ll let the numbers speak for themselves: only 24% of companies characterize themselves as being ‘data driven’; only 30% of CDOs say they are meeting objectives on the ROI from data and analytics; and only 44% of data and analytics leaders report that their team is effective in providing value to the organization.
An Evolution of the Modern Data Leader
This prolonged inability of CDOs to deliver business value has led us to what I believe is a turning point in the evolution of the role. Will data leaders finally figure out how to overcome their challenges, or will the role of CDO slowly fade into corporate oblivion?
When you analyze these numbers to understand the core drivers (or inhibitors) of CDO success, you see a recurring lack of traction on some of the same core deliverables, over and over, survey after survey.
These include things like a failure to:
- provide tangible business value through investments in data management and governance
- build and promote a data culture
- and define and implement a data governance framework – to name a few.
For most, these very failures are the core responsibilities of the CDO. So it should be no surprise that CDO tenures remain half of their CIO counterparts – even as the CDO role is now present in over 80% of all companies.
While a Gartner analyst, it was my job to help CDOs succeed and overcome their challenges. In this role, I spoke with multiple data leaders daily and advised over a thousand of them on what we’ve come to accept as the demonstrated best practices in the discipline. As much as I sincerely wanted my clients to succeed – many did not – which caused me to spend the better portion of the last two years trying to understand why.
Leadership Takeaways from Thousands of Consultations
When I reflect on my thousands of data leader, consultant and analyst conversations over the last four years – many of which happened over prolonged periods with the same people and companies – a few notable things stand out:
- A significant portion of data leaders largely ignored many of the steps that we know could help improve data management. There are many possible reasons as to why, but this happens often. This is particularly the case with building business cases and quantifying business outcomes.
- Many data leaders acknowledged a need to improve their capabilities, but even after repeated engagements with analysts and expensive consultants (most of whom all advocate for established ‘best practices’), they were largely unable to operationalize what was being recommended.
- There’s a palpable sense of frustration with data leaders (especially those trying to do the ‘right’ things). In turn, there’s a growing trend to point fingers at forces or people outside the data organization as the root cause for their difficulties. The ire of data leaders, analysts and consultants tends towards the business stakeholders who both create (through their interactions in business applications) enterprise data or consume the data products of the CDO organization in the form of analytical insights.
- Data leaders’ frustration is being echoed by analysts, consultants and other ‘thought leaders’ in the industry – most of whom are equally perplexed as to why only marginal progress is being made to CDO goals when utilizing their recommended best practices. This leads many – some with extremely influential platforms, such as Gartner analysts – to validate the idea that forces outside the data organization are playing a significant role in blocking CDO success.
- The combination of finger-pointing, frustration and ongoing lack of traction toward goals is helping to exacerbate the already problematic divide between producers and consumers of data that creates a negative feedback loop that’s only making the problem worse. As an analyst, I often heard what could best be expressed as animosity (if not outright contempt) – from data leaders, consultants, and even my peer analysts – towards the business stakeholders outside the data organization constantly.
- The animosity and frustration were often expressed in phrases like ‘the users just don’t get it’, or ‘you can lead a horse to water’ (when trying to understand why data products weren’t being used), but perhaps the best example of this frustration is around data quality. In many organizations, business stakeholders are repeatedly reminded on how poor a job they’re doing in cultivating data assets in the eyes of the data organization – even while they may be meeting every conceivable business performance metric. Over and over, published soundbites like ‘80% of data scientist time is wasted due to poor data quality or ‘only 3% of enterprise data meets minimum quality standards reinforce the narrative that CDO struggles to deliver value are partially, if not completely, a function of problems emanating from outside data organizations.
What has become abundantly obvious to many, is that when measured by three of the biggest yardsticks of job success (fulfillment of job responsibilities, value delivery, and adherence to best practices), excellence in data leadership is sadly lacking at a time when it’s arguably needed the most.
This naturally leads to the question of why?
Modern Data Leaders Need to Shift Their Mindset
The answer, I believe, is one of mindset. In the world of data and analytics, many of our leaders have adopted a toxic mindset that creates a culture within data teams that is completely antithetical to the stated mission of most Chief Data Officers.
This mindset blames others outside the data organization for a failure to become ‘data– driven’, while at the same time endorsing the idea it’s impossible (or impractical) to measure how better data improves business performance.
This mindset blames a lack of user skills as a primary reason for low product adoption, while at the same time overlooking the very basics of good product design. This mindset clings to old and outdated approaches to data management and governance that clearly fail to promote meaningful change, at a time in business when transformation matters the most.
And on, and on.
Do all data leaders endorse this mindset? Of course not, but if this mindset is the reason for low data leader performance, this suggests aspects of this highly toxic mindset are widespread – and have been for a long, long time.
So where do we go from here? One thing is for certain: – What data leaders, and those who advise them, have been doing is not working, and it’s time for some major changes.
The Playbook for Modern Data Leadership
This leads me to what I’m calling Modern Data Leadership, which is a set of guiding principles designed to break outdated, ineffective, and toxic approaches to the role of the CDO. These principles are meant to foster the development of a more productive, innovative, and adaptable culture within data organizations that will ultimately lead to the desired end goal – CDOs delivering prolonged, quantifiable value to their organizations.
Modern Data Leaders are committed to building a culture within the data and analytics function, where they:
✔ Look Inward Before Looking Outward
Building a data culture must start *within* the data team and spread outwards from there.
The same is true with our operational maturity. Data leaders cannot question the maturity of others until they’ve modeled the maturity they wish others to adopt.
✔ See The Refinement Of Data As An Opportunity, Not A Burden
Yes, data capture errors and laziness certainly affect data quality. However, most issues reported as ‘quality’ problems are more related to the natural variability that occurs in data stored in disparate systems for use in equally disparate business processes.
Data in a CRM will intentionally ‘look’ different than data in an ERP, and it’s the job of people in the data organization to resolve those differences. These differences are a function of the fact companies operate in functional silos (sales, marketing, finance, procurement, etc.), which is by design.
This means that data in these silos is optimized for operational uses, not analytical uses – yet we continue to label data any data that needs to be transformed before it can be consumed in analytical processes as ‘low quality’. This is highly problematic.
A modern data leader would acknowledge that disparate data exists for a reason, and that resolving those differences is an opportunity, not a burden.
✔ Believes Their Customers Have Positive Intentions
A modern data leader would constantly remind their teams that the customers of data products are doing their best and mean well for the company.
They are not intentionally trying to make the data team’s life more difficult.
Modern data leaders would actively promote job exchange programs to ensure everyone in the data team has a chance to ‘walk a mile’ in the shoes of the people who depend on data to do their jobs.
✔ Puts Their Customers at the Center of Everything They Do
Modern leaders need to embrace product and design-centric approaches to building solutions.
Modern data leaders acknowledge that customers – and their success – come first, and that technologies only exist to enable customer success.
Modern data leaders strive to build solutions that customers *want* to use and would otherwise happily pay for. They execute against a product strategy that’s fully aligned to a business strategy.
✔ Sees the Inability of Customers to Realize Value from Data Solutions as a Product Failure
If customers distrust data products, struggle to use them, or fail to realize value from them – then a modern data leader realizes the likely cause of the problem is a poor product. Requirements were missed, the design was suboptimal, or the data team failed to adequately understand the customer’s need.
If your product is suboptimal, then nobody will want to use it and no amount of training will change that. Modern data leaders realize that training customers on products they dislike or distrust does more harm than good.
✔ Measures the Value They Provide and Operate as a Profit Center
Modern data leaders have a mindset that embraces the necessity of modeling and measuring the impact of data on business performance. All projects and programs in the portfolio of a modern data leader start with the evaluation of the financial viability of the initiative, measured through the lens of both data and business KPIs.
Modern data leaders aspire to operate the data and analytics function as a profit center – even if the CFO doesn’t share the same perspective.
✔ Manage Data Governance as a Customer Enablement Function
Modern data leaders have a mindset that promotes data governance as foundational element of business enablement and operational excellence. If customers of a data team see no value in data governance, then the data team has failed to meaningfully articulate its value.
In a high-functioning data organization, business stakeholders *want* to be the beneficiaries of data governance, because they have tangible proof that it provides value unattainable through any other means.
✔ Questions the StatusQuo and Legacy Mindsets Focused on ‘Data First’
Modern data leaders see the notion of being ‘data–driven’ for what it is – a distraction that helps to promote the misguided idea within data teams that data is more important than business process excellence.
Modern data leaders know that being ‘data–driven’ was never about the data itself – it was about how data is used for fact-based decision-making. Modern data leaders promote a culture that eschews the belief that data is the most important thing and instead embraces customer success the primary focus.
Change Your Mindset – and Start on the Path to Modern Data Leadership
These perspectives on modern data leadership are certainly not entirely novel – nor are many of them even that provocative. Many, like focusing on customers, are business basics.
But the idea these basics must be included in these guiding principles speaks to the degree of the problem faced by many today. It may sting to admit it, but based on a repeated failure to drive substantial change, it is clear many data organizations have completely lost their way and need a full reset.
If data leaders embrace these alternative mindsets, they can begin to slowly embrace the new perspectives needed to drive meaningful change.
Else, we can keep doing what we’ve always been doing – and suffer the consequences.