Don’t Be Data-Driven, Be Information-Driven
Have you heard of LEO? The Lyons Electronic Office (LEO) is known as the world’s first business computer. It kept track of the costs of all ingredients used to make the baked goods for the J. Lyons & Co. tea shops (among other tasks).
LEO began computing in 1951, and it helped J. Lyons streamline their inventory management and improve daily distribution of goods to their various tea shops. Because of its ability to improve business outcomes through data, some say that LEO was the impetus of the data-driven movement.
In the 70 years since LEO was introduced, technology and the data that feeds it have made astounding progress. Which is why we argue that now and moving forward, organizations need to shift from being data-driven, to being information-driven.
Though they sound similar, there is a crucial difference between the two. Let’s dive into what each means, and how you can make the transition from data-driven to information-driven.
Data-driven decision making is at the heart of the data-driven organization. It means that the company collects, analyzes, and uses data to inform key decisions.
In essence, it’s about taking guidance from facts rather than “going with your gut”. It sounds obvious. Everyone does this, right? In reality, everyone tries to do this, but it’s harder than it seems.
Being data-driven is an ever-evolving quest. Through the decades, the amount of data we can access and the tools we have to analyze that data keep expanding and improving. This is both a blessing and a curse.
Organizations now have truckloads of data. And today’s advanced AI and machine learning make it easier than ever to analyze it. But data is scattered and siloed, and most of it is unstructured, adding more complexity.
In just the marketing field alone, Gartner’s recent study found that only half of marketing decisions are influenced by analytics. There are many reasons why, from poor data quality to unclear recommendations. Reasons aside, the takeaway is that being a data-driven organization, to put it simply, is really hard.
The challenge is worth it. According to a 2019 Deloitte survey, among respondents with the strongest analytics culture, 48% significantly exceeded business goals in the last 12 months. This is why most organizations are focused on building data-driven strategies despite its difficulty.
To be data-driven, companies need to set clear objectives and have specific business questions that data can help answer. Then comes the arduous task of locating all the applicable data, and cleaning and organizing it. An IBM study showed that 80% of analyst time is spent on this stage, leaving only 20% for actually analyzing the data. After the data has been collected and analyzed, recommendations must then be drawn from the findings to guide decision-making.
Source: Northeastern University
Fostering a data-driven culture is also key to succeeding as a data-driven organization. This means having the right talent and a leadership team that reinforces the value of data to all levels of employees. It also means making the data and analysis tools widely accessible and creating an environment that supports a test-and-learn approach.
According to McKinsey, companies that encourage employees to consistently use data to inform decisions are “nearly 1.5 times more likely to report revenue growth of at least 10 percent in the past three years.”.
– Warren Buffett
Mr. Buffett concisely explains why making only data-based decisions is problematic. Data is good for validation. But it’s only as good as its interpretation, and on the completeness and accuracy of the data you found. Without the right context, your data-based decision can lead to the wrong outcome.
The untapped potential of data is in the context that is added to it. Think: data and content that is connected along topical lines and enriched with the intricacies of language. It’s not just spreadsheets and reports, but relevant presentations, Slack conversations, emails, PDFs, and more. With the power of natural language processing and machine learning, we can now add rich layers of context, so that data becomes actionable information.
The more actionable information we’re given up front, the less work we have to do to evaluate it and make optimal decisions.
TLDR: Information > data.
Being information-driven means that employees spend less time searching for what they need, or making sense of the data, and more time making the decisions that will help the business grow. It’s all about reducing cognitive burden.
What’s that, you say? Cognitive burden is the extra thought and effort that we require as humans to evaluate all the available options and make optimal decisions. We experience it on an individual level, and organizations experience cognitive burden in aggregate, as the sum of each employee’s burden.
To be information-driven, organizations need actionable information presented in context to surface insights, inform decisions, and elevate productivity. The key to making this work is to automate as much of the process as possible. In doing so, organizations can reduce the burden on its employees, and the amount of interpretation that they need to do.
Information-driven organizations need to be able to:
- Access all of their data and content – both structured and unstructured, across applications, environments, file types, languages, and more.
- Decipher meaning from each piece of content – automate the interpretation through natural language processing, concept extraction, part of speech tagging, etc.
- Use machine learning to scale – AI and machine learning can develop models based on the curated content and monitor user behavior to adjust and scale.
- Give users a simple, intuitive interface – a single place where users can access all of their relevant information and insights.
To get these capabilities, organizations should look for a solution that is:
- Cognitive – It should be able to process natural language and learn from curated content.
- Complete – It should be able to grow and work for multiple use cases and different business ideas over time. You don’t want to have to find a point solution every time a new business requirement, use case, or new idea comes up. The solution should be fully integrated, adaptable to any size project, and leverage your existing systems so that configuration and not coding is all you need.
- Proven – It should have experience with large scale deployments in complex, heterogeneous, and data-sensitive environments. It should be agile and able to take learnings from edge and corner cases and build them back into the platform.
Related blog posts.
Recently, we announced that Sinequa Search Cloud is now available through Microsoft Azure Marketplace. This blog post explains how acquiring our solution directly from the marketplace brings additional benefits to your ...