Turning Yesterday’s Dark Data into Tomorrow’s Data-Driven Strategy
Is “data-driven” just a slogan in your organization, or a real approach to decision-making?
In a new Splunk report, 56 percent of respondents said they felt that their organizations were merely paying lip-service to the idea of a data-driven business strategy. Meanwhile, 90 percent agreed that, to be successful, organizations will have to extract value from as much data as possible.
However, there is a lot of data that is unknown, untapped, or even not understood to exist. Lack of access to this “dark data,” which is generated by systems, devices, and interactions, is a problem that many organizations are just waking up to.
In the report, which is based on a global survey of 1,300 business and IT leaders, those surveyed reported that more than half their data is dark. Yet 81 percent rate data as very or extremely valuable to their organization’s success.
In addition, only 57 percent of those polled are enthusiastic about learning new data skills (although a higher percentage, 92 percent, is “willing”) and 71 percent expect data to become more valuable over the next 10 years. Nearly all expect data to become influential to their decision-making.
Obstacles to recovering and using dark data
Respondents cited a long list of obstacles in accessing data, including:
- the sheer volume of it
- a lack of necessary skill sets to mine it
- a lack of resources
- difficulty in coordinating across departments
- lack of dedicated data managers and finders
- difficulty in coordinating with data-generating third parties
- lack of control over data-generating devices and apps
- lack of interest from organization leaders
- lack of creativity
“The key is to use the human component plus data plus AI to truly get a foot up on your competition,” said one U.S. retail CIO.
However, many organizations admitted they weren’t ready for AI. They cited a lack of trained AI experts, a lack of understanding of AI, not knowing what can be automated, and difficulty in successfully wrangling the data.
Interestingly, one problem in adopting AI and machine learning in the first place is a lack of data. As Antonio Piraino writes at Forbes.com, “To keep up with the rapid pace and scale of today’s digital environments, enterprises are turning to AIOps [artificial intelligence for IT operations], which is powered by machine learning (ML) and artificial intelligence (AI).”
Piraino points out that unfortunately, ML-based algorithms and AI-based automation, key elements of unlocking digital transformation, are easier said than done. “ML needs to evolve into AI, and to do that, it needs cleaner actionable data to automate processes.”
Adopting a data-driven strategy
As data accumulates at a tremendous rate, organizations are going to be looking for more storage and protection solutions, like Bell’s Q9 Data Centres.
But simply having data isn’t enough. Business analytics skills to use the data in a way that leads to profitable action is crucial. In the coming decades, analytics training will be an increasingly valuable skill, and data skills in general are going to become more important for workers across the board, not just IT employees.
Data experts are going to be the new business strategists. Becoming a senior leader and decision-maker in an organization will require data literacy and the skills to translate data into business solutions or actions. The combination of technical data skills and business acumen will be most in demand over the decade, though these workers will also be the most difficult to hire.
In the end, the Splunk report makes four key recommendations to defining a data-first strategy.
1. Embrace AI and machine learning. These technologies are potentially transformative if you can find the use cases that make sense for your industry and organization.
2. Build an infrastructure and culture of data. Make sure “data-driven” isn’t just a slogan, but a real approach that ensures you’re making full use of its potential. AI is fundamental to tapping and analyzing this resource.
3. Recruit wisely. The intersection of data management and business analytics training is where the intersection of future decision-making lies. Find employees who have a head for business and data analysis. Curiosity, self-motivation and collaborative skills a bonus.
4. Provide opportunities for training. Make it easy for your existing workers to learn the new technologies and skills that will help your organization mine data, dark and otherwise.
In addition to these for recommendations, Google Customer Analytics Global Program Manager Neil Hoyne has found examples of how successful companies look at data. One important finding: they look at their metrics as part of a story, not the whole picture.
According to Hoyne, successful data managers ask: Do I know what this metric truly means? What could influence that metric, and how? Am I limiting what I can learn from my metrics?
When it comes to developing data strategies for machine learning and AI, Forbes’ Piraino says that it’s a matter of realizing the need for two strategies: one for historical data, the other for real-time data or continuous learning.
Data, as one Splunk survey respondent said, “Has all the secrets, if you know where to look and how to study it. It’s just a matter of getting the right people and the right tools to leverage it.”