How to Galvanize Analytics Digital Transformation

4 key ways analytics can effect changes that deliver real value

Analytics digital transformation is the shift from technology helping the business to technology being the business. In the last two decades, we’ve seen our fair share of winners that have become completely different companies and losers that have failed spectacularly. Consider Netflix, the company founded as a DVD delivery service in 1997 and transformed into a streaming giant in 2007. Then consider Netflix’s former rival: Blockbuster, which tried to adapt but couldn’t move beyond its brick and mortar heritage.

Why do some initiatives fail while others succeed?

It’s difficult to pin success or failure on any one reason. It’s usually a constellation of reasons. Ultimately, however, it’s all about how an organization deals with change.

Digital transformation is the deliberate attempt by an organization to adapt to change. At the heart of all transformation is data. Because, after all, what is digital transformation except for the use of technology to find better ways of leveraging and processing data to do things?

I would argue that analytics—which is the use of data and mathematical algorithms to find patterns to understand why something happened or predict potential future outcomes—is a core component of digital transformation. It’s not as prominent as other digital transformations, but it’s just as impactful.

Analytics digital transformation is not just “how technology becomes the business” but “how data in the hands of the right people becomes a competitive advantage.” It’s a crucial component of transforming raw data across the information supply chain into actionable information.

In this article, I share four practical ways organizations can succeed in their analytics digital transformations.

How to transform

Digital transformation is pursued by organizations to create new value, either in the form of internal efficiency or by creating new streams of revenue. So how can it be done? Oftentimes, executives who rightly understand the value of data look to their analytics infrastructure to begin their journey. Let’s look closer at analytics digital transformation. From a practical perspective, how can an organization effect meaningful digital transformation in their analytics?

Let’s consider four ways:

1. Understand and acknowledge the importance of trust

The word “trust” is often used when people talk about data and technology. It generally implies the need for security or protection. However, there’s a negative association at play: when you say words like trust and security, the conversation is invariably about preventing unwanted things from happening. Ultimately, though, trust is a human emotion. For us, trust is not about preventing things, it’s about enabling things. Trust opens pathways: to new relationships, to new opportunities, to new ideas. It’s a bridge. And humans like to humanize inanimate objects—like data. So let’s use this to identify four core aspects of trust as it relates to data and analytics:

1. Reliable

Whatever analytics solution you put in place needs to be reliable. It is the cornerstone of trust. Nothing destroys trust more than when a system or platform goes down. Keeping analytics systems and data pipelines up and running should be IT’s number one priority. Simply put: your analytics need to “be there” when called upon. Yet time and time again, I've seen organizations slip-up on availability. I’ve heard it said before that if IT had run NASA, we would have never put a man on the moon.

2. Believable

Your analytics need to be believable. Belief comes when you know the source of your analytics and are confident in its accuracy. One of the most common statements you will hear when presenting a new dashboard to the executive suite is: “I don't believe that number.” Ultimately, trust requires more than a number: there needs to be an understanding of the number’s source and its context.

3. Credible

Trust thrives when systems are being used. Humans are social creatures, and when we see evidence of use it validates the credibility of the systems themselves. Think about how social media has transformed “credibility” in the digital age, and apply those some social media concepts of usage, likes, and comments in the analytics solution you deploy. It’s crucial to able to measure and communicate how many people used or consumed a particular piece of data. When it’s clear that lots of people are interacting with data, that adds validity to the data analytics. I’m not saying that we measure credibility exclusively on usage, but it’s a crucial signal. And credibility is further strengthened when you pair it with believability (see above)—and your analysis is supplied by sanctioned sources and backed by proven analytical and mathematical-based techniques. Those are important ingredients for true transformation.

4. Autonomy

Lastly, trust is built on autonomy: your solution needs to deliver and promote self-service capabilities. You need to trust people to do what they need with the data and avoid micromanaging how they interact with it. If you constrain it, you lose the opportunity to maximize creativity and innovation, and your data will have limited use and impact. Trust your valued employees to be responsible with the data and let insights flow. In conversations with dozens of customers, I frequently hear concern about data security, or that employees will “draw their own conclusions” with data. I would argue that’s exactly what you want to happen. If you stifle ingenuity, you will never see the innovation that results from insights—which fundamentally depend on access.

2. Make analytics a “real” part of the business

Too many important initiatives are considered strategic imperatives but aren’t backed by any meaningful resources or support from the board and executive team. They’re considered extra credit or “nice to haves”—and never bear fruit. If you don’t give someone ownership of the analytics digital transformation role–-if you don’t pay someone to do it and put measures in place on their performance—it will be yet another failed technology initiative.

You need to put actual weight behind the initiative: put the proverbial cape on your champion’s back and give them a mandate. Then give them the resources they need to succeed and watch them fly. That’s the only way to realize the true value of an analytics data initiative. I’m encouraged to see this happening in more and more organizations. They’ll appoint a Chief Data Officer (and in some cases a Chief Digital Officer) and put them in the executive suite. In some cases, these individuals will come up through the IT organization and be given a business focus.

The other thing I’ve seen work is making analytics a shared service. If you already have other shared services initiatives in place, you already have a framework in place to understand how it works and you can track where and how the analytics are being used. Ultimately, this provides a way to measure performance and a model to recoup costs.

The combination of these things—a leader who owns the initiative and the deployment and management of a shared analytics service—provides an organizational pathway for analytics digital transformation to occur.

3. Make data (and analytics) broadly accessible.

To make transformation possible, you must make data broadly accessible and give users an environment to interact with it on their terms.

First, help people get to the data they need without having to figure out where the data is, the field names, who owns it, permissions, etc. This frees them up to focus on the big picture (outcomes) without getting wrapped up in the technical details.

It’s common to see organizations make a verbal commitment to data accessibility but still struggle to provide it because it’s locked up in numerous silos. And even when access is finally granted, more challenges arise: the tables may come with technical or unfamiliar fields, or it’s the wrong format. This mismatch of intention and actualization is very common.

That’s why your transformation depends on a Rosetta Stone to interpret the data and guide its use. With it, everyone in the organization gets what they need, knows where the data comes from, can see how others use it, etc. This acts as a semantic layer (or metadata layer) on top of the data itself so that can help others understand it. For example, if I’m on the Finance team, I see familiar names and fields when I’m looking at my data, and there’s a shorter path to insight.

Another crucial piece of data accessibility is inside the analytics environment itself. The UI needs to be specially tuned to individuals based on their specific needs and analytics maturity.

Most people within an organization have simple analytics needs. It’s not so much they lack sophistication or intelligence; it’s more that they’re focused more on live-fire business situations and need ready access to analytics at a high level. For them, their needs are simple and straightforward (and crucially: trustworthy) so they can make the best decisions possible.

Others have slightly more nuanced needs. Maybe they need to explore the data: filter, dice, swap, add code, all without writing code or having to learn a particular skill. Still, others require robust analytics capabilities: they need to access the granular raw data, want to write or inject code, or need a sandbox to play with the data.

As you consider analytics digital solutions, you need systems that can be tailored to individual users so that skill and familiarity build progressively. It can’t be a constant game of swapping out new tools for each user—it has to be one adaptive tool that grows with them.

4. Don't let perfect be the enemy of good or good enough

Sometimes the perfect is preventative. One of my favorite concepts is “don’t let perfect be the enemy of the good.” It’s common to misinterpret this to mean “oh, I did my best.” I think of it differently. Sometimes 60 percent of what you need gets the job done better and more effectively than being complete. Consider traveling within London using underground public transit. I can use an actual map of London to identify way-points to get my destination, but it’s much easier to use a map that’s suited to the job: all the extraneous information is removed so that every detail matters. If all I need is to get to a particular destination, I don’t need a literal map to get me there.

As you think about analytics digital transformation, don't let finding or having the perfect solution get in the way of getting things started or getting things in place. The reality is you will never have consumers of your analytics who need the perfect: they will be much better off with good enough.

What’s next?

Analytics digital transformation is hard, and not all organizations will be successful. At the heart of all transformation is data—and how well or poorly organizations can harness it. My advice is to focus on trust, make analytics a “business” driver, make data accessible, and keep it simple. Consider it practical advice from a guy who’s seen his fair share of failed IT initiatives (and sincerely hopes your organization finds a way to win).