CTO of Pipedrive with expertise in scaling technology and organizations. Experienced as an innovator, founder and C-level manager.
It doesn’t matter the role or the business; data is essential. Whether the captain of a squad, a chief people officer for a multinational business, the CTO of a unicorn technology firm or the help desk worker, if you don’t know what’s happening, you can’t make meaningful progress.
Over the past decade, data has become the North Star of business. Technologists, the enablers of business transformation, must deeply appreciate data and analytics. Many industries have been revolutionized and new sectors have been created out of whole cloth. It all emerged from the wonder of abundant and useful data that is often created as a mere by-product of user interaction. But as with most activities in life, results fit along a bell curve. Just as if overwatering your houseplants, too much does not always improve the result. Just the right amount of data is better than too much or too little. And data alone can be isolated and hard to use. Connecting data and layering on metadata offers businesses the leverage to make a big difference in performance.
Connections Are More Important Than Data
For example, a set of random words like Kamala, Joe, Harris and/or Biden, is a set of data but with no real utility. If we understand that the first two are first names and the other two are family names, meaning emerges, and we know we are talking about U.S. politicians. With more metadata, we can know that the first and third names make a full name, just like the second and fourth ones. This enables us to identify specific people and make decisions with the data we have.
Intelligence comes from data that’s been put to work by linking it. The capability to make connections between data to form intelligence that can be applied has formed the basis of the business intelligence and analytics boom that’s brought us tools like machine learning-driven ChatGPT and AI assistance. It’s how front-line staff can access enterprise data by asking natural language questions instead of needing to learn to code to create their own data models.
Signal, Noise & Managing Data
Ask your cybersecurity or software engineering teams if there is such a thing as too much data. They should vehemently agree. Most systems, devices and accounts pour out a continuous stream of data and metadata that can be stored, processed and used. But not all data is useful—let alone useful to everyone. Culling the “signal” from the “noise” is what we call taking only the most useful data to base a coherent and useful process on.
There is certainly no shortage of data that can be brought in or tapped from internal sources. Going on a self-imposed data diet to ration yourself to the data that matters is hard. Different internal teams will require different regimens. Ensure that all models and solutions can scale and flex to different needs. One size may not fit all. Your sales team and your engineering teams will have unique needs.
To do this requires careful planning and data curation, but also automation. Using machines to perform the sift, flag relevant intelligence or even take initial action is essential. It’s how incident managers are able to rapidly manage IT failures for massive companies with complicated and interdependent technology systems. It’s how cybersecurity software can rapidly stop malware from spreading and causing a data breach once it’s entered a system. It’s how electronic trading firms ingest geopolitical news and algorithmically turn profits based on rapidly trading stocks that will be impacted by this news later.
All this is to say, automation and decision making are the other side of the data “coin.” Without augmentation, human colleagues will not be able to perform their best. There was an estimated 2.5 quintillion bytes of data created every day globally last year. Humanity cannot analyze it all, nor should we. Machines can help bring intelligence to people so that we can make decisions based on the best data, simply and easily, in the right timeframe.
Compliance, Security & The Toxic Half-Life Of Data
Caring for employee, customer and all stakeholder data is important for ethical, trustworthy and profit-seeking businesses. That’s why compliance isn’t a have-to; it’s a must-do. Ensuring compliance with national and international data protection is simply part of smart data management. What’s more, shedding old and excess data helps the business focus on the data that makes better intelligence and reduces the risk of keeping outdated, irrelevant or sensitive data that may be leaked or lost via mishap or cybercriminal activity.
All business leaders have seen a variety of metaphors used to describe data like the “new oil” (so valuable) to the “new solar” energy (an ever-renewing resource). However, some have also called it a radioactive material. Data becomes less useful as it ages, often becoming not just outdated but completely misleading and unfit for purpose. Moreover, storing it comes with a cost, and protecting it from loss comes with the ever-present risk of failure.
Being clear-sighted with the reasons, purposes and timeframes that data is useful and sunsetting and removing it when it’s time helps protect stakeholders from past-prime data.
Purposes, Processes & Outcomes
As with every element of stellar business success, knowing the purposes and setting the right processes to help achieve predetermined goals is key with data projects. Data supports business leaders in making better business decisions across the whole spectrum: hiring, compensation, suppliers, customer acquisition, etc.
Clarity of purpose and setting SMART goals are especially important in data projects because of the importance of aligning data to the real world. Creating the best product for a customer who can no longer afford it is an error. Testing, validating and ensuring the best fit between models and the world is an ongoing dance.
It’s exciting to be a technologist and an essential part of what makes a modern organization tick. It also comes with responsibilities. Understanding how and when to use and retire your data is key.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?