The Value Of Data
If you spend some time online reading about data, you almost cannot avoid noticing that many (including publications in magazines and websites with excellent reputation) emphasize the value of data. Some examples are “The world’s most valuable resource is no longer oil, but data” (The Economist) and “Data: The CIO’s Most Underutilized Asset” (Information Week). Data science jobs are on the rise, and many success stories concern organizations that use data to create a competitive edge, to change the rules of the game, and ultimately to become successful, and even dominating in the marketplace. So why doesn’t everybody manage to replicate this success? And what does it take to generate value using data? How do you become a data-driven organization?
Data has become a “sexy” term. Success stories include, for example:
- Planning your logistics, based on the deepest insights about your clients (example: Amazon can plan its logistics ahead of orders being placed, based on its knowledge of its clients’ behavior; the so-called “anticipatory shipping”)
- Developing new revenue streams (example: Vodafone was able to offer traffic jam data to TomTom for its navigation systems)
- Stitch Fix uses data analysis and data-driven algorithms for recommender systems, merchandise buying, inventory management, relationship management, logistics, operations and even for designing clothes
Use Case Driven Value
In an environment where “everybody talks” about data leading to success, it is easy to succumb to the will to duplicate this success without really understanding what it takes.
One common pitfall is treating data as a technology project. Such an approach would entail that you develop a technology solution – often the term “data lake” is used – to capture “all our data”. Yet without considering what it is that you want to achieve with the data. What will you do with the data? For which purposes will you use it? Who will use it? How does it need to be used? For which use case will you use it?
If you give curious people with business acumen access to large datasets, they will come up with innovative use cases that you may not have thought of before. I am convinced about this (see some personal experiences). Thus there is no need to plan every usage of the data in advance. Yet the important message here is that data initiatives should be driven by specific use cases, having concrete value in mind.
Don’t Stall While Considering All Use Cases
This does not mean that you should not develop an information infrastructure before having all the concrete use cases in mind. You should have a clear strategy on what you want to achieve, and allow your data science team to experiment, together with business users, to further develop new ideas by using the existing information infrastructure.
Data Science: Skills
A senior manager at a Government agency said to me: “Two years ago we introduced a data-driven strategy; now we realize that we don’t have staff with data skills”. Similarly, while working for another Government agency who was implementing an ambitious data-driven law enforcement strategy, I provided guidance to the agency on the IT-project being a change project. In preparing a meeting with the CEO of the agency, I plotted the growth path that the agency was implementing on a timeline, and added the timelines of similar agencies that have already implemented similar initiatives. This agency was trying to implement a data-driven strategy much faster than any of its peers did, while treating it as merely an IT project. The implications of this observation were clear, as was the CEO’s response to my diagram: “I have to show this to my minister” (who was putting pressure on the project).
Becoming a Data-Driven Organization Entails Change
The conclusion from these two examples is that not for nothing does the term “digital transformation” include the word “transformation”. Every transformation requires new skills. In becoming a data-driven organization, you will invest in developing new skills in the organization, and new ways of working. Thus you will change the fabric of your organization.
Most organizations of a reasonable magnitude have or had to deal with a situation where data is distributed across multiple systems. There is no “golden record” and no “helicopter view” because the various systems are not connected. Moreover, in many cases, there is no uniformity (across systems) on how data is captured.
Once you have defined the scope of your data strategy, you need to deal with the data governance challenge, to allow you to…
- Connect various data sources; and
- Combine datasets from these sources; and
- Create insights from the combined dataset; while
- Managing which source of data provides “the single truth” (about various types of entities)
In my earlier blog “Turning Data into Business: Data Quality vs. Data Quantity” I discuss how to achieve data quality. If you just copy data to a central repository without dealing with data quality challenges, the project may be doomed to fail, because you will fail to create insights from the data, and thus to delivery value.
Disruption Is Coming!
Engaging with data is not just a small change that you can do to fix a small challenge. Are you ready to become a data-driven organization? You should be willing – and even eager – to let go of the old ways of doing things. Open up to new ideas, to challenging the status quo and to reinvent your business. Data-driven solutions may bring upon new… (examples):
- Pricing models (and giving up old ones)
- Delivery models (and giving up old ones)
- Revenue streams
- Cost structures
- Business partners
- Go-to-market strategy
Similarly, data-driven solutions may bring upon more changes, for example:
- Change in how staff do their work
- Changing organizational units
- Shifts in manpower, due to shift in required skills
- Change of management (especially those who cannot make the transition)
- Change in organization culture
A Data Culture
We’ve discussed the need for data science skills. Yet a data culture goes beyond the data science team. It extends to every part of your organization. As a leader, you want each of your employees to think data-driven. For example:
- If your supply chain planning is under pressure, your employees should engage the data science team to analyze the bottlenecks and define solutions.
- When your sales team is faced with a new need of a customer, you want then to consider which data is required for solving this challenge.
- On a continuous basis your Customer Success Champions should analyze the available data about your customer, to understand their needs, behavior and preferences.
Curiosity and a pro-active approach are key characteristics of employees who will embrace a data culture. And last but not least: management should become ambassadors of the data culture. They should lead by example. This is no different than any other change initiative.
Are you ready to reinvent your business model, and your operational model? Change is inevitable. Not because data has become an end. But because the changing environment requires you to use data as a means to an end. Who thinks that their organization can be successful in the future without embracing this change?
- Turning Data into Business: Data Quality vs. Data Quantity
- Why Referential Matching Is Superior To Standard Entity Analytics Techniques
Go back to the blog start page.