Informatica: Chief Data Officers: Measuring the Value of Data
Those of you who are in the chief data officer (CDO) chair know that expectations for the CDO role are changing. The one common theme is that CDOs are experiencing a job that varies by the day. It began with compliance and minimising the risk associated with data. Now CDOs are being asked to measure and increase the value of data. To do this they need to democratise data, making it more trusted, relevant, and accessible – but not make it a wild west.
Businesses are starting to grasp the importance of valuing data as an asset, but how do you determine what that value is? How do you build that language and framework and make them available within your organisations?
To help answer these questions, we were joined at our board meeting by special guest Douglas Laney, Innovation Fellow, Data & Analytics Strategy, at West Monroe Partners, former VP and distinguished analyst with Gartner’s CDO research and advisory group, and author of the book Infonomics. Doug’s presentation of the ideas underlying “infonomics” sparked a lively exchange with the board. Here are some takeaways from our discussion:
Define data as an asset. The definition of an asset is something owned or controlled, exchangeable for cash, and that generates a benefit. Data meets that definition, but many pundits believe that data doesn’t have value unless you use it. This is contrary to how other assets are valued – they have economic benefits whether they’re used as a revenue stream or for expense savings.
Manage, monetise, and measure. What do we do to assets if they have value? We manage them with discipline, standards, and processes and figure out how to monetise them externally and internally to generate value. So, we need to manage, monetise, and measure data as an asset. But you can’t manage well what you don’t measure well.
Apply economic concepts to data. You need to think of supply and demand and pricing elasticity in the context of data. And data has many forms, ranging from raw to enriched. For example, it’s easier to monetise enriched analytical data that underpins data science models.
Focus on data literacy. Data literacy represents a major cultural aspect to doing more with data, around politics and change management. By moving toward self-service analytics and providing support for the foundational steps you need to take related to delivering value from data, you can guide the training and inject data literacy concepts into business culture, communication, and collaboration.
Don’t fixate on any one type of data. You can share, publish, and monetise raw data, which has great general utility, but typically consumers want data more specific to a particular use case, which becomes more expensive to build and deploy. Because data is nondepleting, you can package it up in a variety of ways.
Look at intrinsic values for data valuation models. Foundational measures such as accuracy, completeness, and scarcity have typically greater potential value. Also consider whether data is relevant and timely and look at how having or not having data impacts various processes. Take financial measures as well, such as cost value, market value, and economic value.
Flipping the Conversation on Risk
Many organisations are focused on risk mitigation. Laney advises to flip the conversation around: What’s the risk of not capitalising on your data? Board members agreed that the conversation around risk is important. For example, if you lose data, what would the impact on customers be and how much trouble would you be in?
As well, CDOs are often seen as an enabler of commercial value, as opposed to data having value in its own right. Laney added that infonomics on data monetisation is a way to have conversations with business stakeholders to inspire executives to do more with data. He downplays asking, “What is our data worth today?” and encourages people to think about the trajectory – how are you improving real or potential data over time? To participate fully in the data economy, you need to know what you have and what it’s potential or actual value is.
Also, you should make the case that data is not an IT asset but a business asset. The real value of data is often hidden behind the customer. Data is seen as a background element, so it’s hard to see what data part led to an outcome – look at the value chain of the data. Think about and explore data and what we know about customers at the start of project.
Expanding on Data Literacy
For the final portion of the meeting, the board segued into a discussion on how they foster data literacy in their organisations, led by Susan Wilson, Informatica VP Data Governance and Privacy Leader. Here are some of the ideas that were shared:
- Data folks bridge the literacy gap between business and IT. Implement new solutions and get people at the table to make sure we do the right thing.
- Build a relationship between a data steward and a data owner in each division. Have a consistent understanding of data practices and share the same language across the organisation. Don’t assume a level of understanding. Do a needs analysis, starting with executives – meet them where they are.
- Start a book discussion on data literacy – mark a passage and pass it on and ask the recipients to do the same. And, have data scientists have one-on-one sessions with executives and bring outcome stories to life.
- Create an organisation with a charismatic leader who can promote what data science is. And set up a formal program with certification or badges that can help generate enthusiasm and encourage employees to share when they complete the program.
- Put an embargo on literacy tools (which organisations sometimes see as a panacea) until the governance group can get together and define how they’re going to measure success and objectives.
- Look for a quick win – value that the business wouldn’t realise on their own. It can be as simple as recognising that business units are measuring the same thing but doing so inconsistently. Connect the dots by working together with the business.
- Don’t boil the ocean when getting started with a data literacy program. The temptation is to mushroom and expand it but what really works is to put a ringed fence around a set of topics.
About this Author
This Guest blog was written by Jitesh Ghai, Chief Product Officer, Informatica. Jitesh S. Ghai, is senior vice president and general manager for Informatica’s overall data management portfolio including Data Integration, iPaaS, Data Quality and Governance, Data Privacy, and Metadata Management offerings. He is responsible for the portfolio vision, cross-functional strategy, and roadmap. Previously, Jitesh led Informatica’s strategy and operations group where he was responsible for the overall definition and governance of Informatica’s multi-year organic and in-organic portfolio strategy. Read more...
To read more from #techUKSmarterState Week check out our landing page here.