The DALAI Framework
A framework for CDOs on how to achieve Data Zen
As readers will know, CDO (Chief Data Officer) is one of the most rapidly evolving executive roles in organizations today. It’s grown from nothing, to both Gartner and Forrester now reporting that more than half of the organizations have someone with that title[1].
The role is sprung from very different parts of the organization, with different mandates (more on that in a later blog). But at the very core of it lies the need to harness value out of data at an enterprise/transformational level. According to a survey of CDOs conducted by Gartner, “cultural challenges to accept change” is the number one roadblock, followed by “poor data literacy”[2] – both of which are people-oriented aspects. That’s why I’m encouraged that when I had the privilege to stand in front of some of the most senior data executives in all of Spain last week, the topic they had chosen to discuss was “data literacy”. Naturally, the question of how it ties to broader data driven transformation quickly came up. When it does, I tend to bring up the “DALAI” framework. For those of you that listened to my trends webinar earlier in the year, this will look familiar. 5 crucial components are needed; Data, Activists, Literacy, Applications, and Ideas.
Data: The idea that data will ever be on one place is not realistic. Especially in a world where new and fragmented data sets like social data, open data, and IoT come in at an ever-increasing pace. Hence, an equal level of governance can’t and shouldn’t be applied. Instead, focus should be on a) mapping out the entire data estate as much as possible, and b) applying different levels of governance as close to the data as possible. Central governance efforts need to be applied primarily to the most sensitive data for compliance reasons, but also to the data that is the most mission-critical or sees the most use and re-use.
Activists: Now we’ve veered into people territory. The activists are the people in your organization that have the highest data-skills. These are the data architects, data scientists, application developers, and business analysts. Create a “culture of de-silofication”, where they have a license to grab data from different parts of the organization, and combine it in novel ways. They should have data access, and tolerance to veer outside of the norm, but under certain stipulations. For example, taking a code-of-conduct course, or having to have a high enough certification-grade. Success should be put in the limelight. If activists are challenged–yet celebrated, your chance of maintaining these highly sought-after individuals increase, while providing a path for others to become like them.
Literacy: As discussed, this is a major roadblock, but can be the bridge to activism and bringing data transformation further out. According to Qlik’s own survey at www.dataliteracy.info, only about a third of business decision makers are fully confident in their ability to read, work with, analyze and argue with data; which incidentally is the level of BI adoption in most companies. 82% of respondents said they would be willing to invest more time and energy into improving their data skillet. So, this is low-hanging fruit. We heard at the recent Qonnections conference from thought leaders like our own Jordan Morrow and Gartner’s Valerie Logan[3] that it’s imperative to do an audit of the different skill levels in your organization, and categorize them.
Applications: It’s a cliché, but data without applications, is like oil that hasn’t been refined, i.e. less useful. Here it’s important to right-size apps to the different users. Some will only need key KPIs once a week, while others, like sales-people, will need it in “moments” on mobile phones. Some will want to explore in pre-built apps, while others will want to create their own apps. Everyone wants data contextualized. So, leverage multiple interfaces like custom apps, mashups, bots and embedded. Ideally in a hub where data models can be re-used. Just like the data and people, the applications benefit from their own certification system. I’m in the camp that thinks that app creation should be encouraged. Even if it causes proliferation, it comes closer to decision making and gets people started. The useful apps will live on and become certified apps (with higher governance) ready for enterprise-class usage, while others will answer more sporadic questions to then wither away or be retired. But without experimentation further out in the edges of organizations, data driven innovation can’t be fully captured.
Ideas: While a CDO might be the sponsor that acts as an arbiter and enabler of data-driven ideas, these ideas often happen in the fringes of organizations, from people who sit closer to the problem. I once heard from a wise man, that “none of us is as smart as all of us”. The best way to leverage ideas, apart from doing the previous four points, is to have an infrastructure and culture that supports gamification and collaboration, to capture the data-driven ideas; and then have an infrastructure that enables a promotion path from sandboxing to work-group, to departmental, and finally enterprise-grade. That way you provide a virtuous loop from top-down best practices and governance, and bottom-up ideation and agility.
While theoretical, parts and pieces of this framework have been present in the most successful data-driven programs I’ve observed. Yet, of course, it needs to be applied differently, depending on industry, regulatory environment etc. CDOs and other analytical leaders can and should adopt some of these principles to increase impact.
[1] Survey Analysis: Third Gartner CDO Survey — How Chief Data Officers Are Driving Business Impact & Forrester: “Chief Data Officers Play A Leading Role In Business Transformation”
[2] Survey Analysis: Third Gartner CDO Survey — How Chief Data Officers Are Driving Business Impact
[3] Gartner: Information as a Second Language: Enabling Data Literacy for Digital Society
