Note: this post is by no means meant to be exhaustive. Return on Investment for data can fill a book. What I try to do here, like in most of my posts, is to stimulate thought and conversations.
I often tell clients that there is no magic when it comes to data. Yes, it is hard. It is not complicated, but it gets complex very, very fast. So consultants come up with frameworks to help us think through some of the complexity so we can come up with an actionable roadmap that will deliver business value. Vendors come up with products that speed up time-to-value. So far, this blog post is a waste of time. Bare with me.
I like simplicity because it makes it easy to explain and get people to buy-in, and we need a lot of that. So let's start with breaking things down. ROI is Return divided by Investment. We have two metrics to go after to have an impact. Since we are all customer-focused, let's start with Return, which we can also call value, usefulness, etc.
If we think of data as a product and want to raise its value, we have to think about ways to use it MORE. Here are some key ideas to do this:
Have more people use it
Use it to inform more decisions
Apply innovation to come with new insights
Now, for the Investment part. As I mentioned, data is difficult. So how can we reduce how much we spend to be able to use this data? Focus. Here are some ideas:
Reduce needless work
Know why you are investing, with a clear idea of how it will benefit
Invest not only in the technology and the data work but in people, so they can better collaborate and more often use the data
You will hear a lot about Data Literacy in the coming years. If you want people to use data, train them. Not just in the technology but especially understanding what they are looking at and think critically about data. Two technologies that can have a huge payback in this area are Data Catalogs and Search-Based Analytics. AI comes in many forms, and many of the technologies today embed AI to drive a better impact. That is another way to have AI payoff.
Inform More Decisions
To achieve volume and economies of scale, we need to automate. Automation at every step must become part of the culture and Data Strategy. Applying AI and machine learning is great, but it must be with the intent of automating decisions. Portions of the development cycle with data can also be automated. Include R&D in your strategy to continue improving at every step.
Innovate for New Insights
Scale data innovation. The same data can be used in many different ways. For example, you may have used a certain data set, created a model, put it in production, and reaped millions of dollars of benefits. Now, some tools will try out dozens of models on the same data, automatically tune the models, train them, and test them out for performance - all automatically! If that is available, why wouldn't every organization want to use it? This could save months of research.
Reduce Needless Work
Some data work is very hard. So we have to have a Data Strategy that identifies these areas, and how they will contribute to value so that we know whether they are worth doing. I wrote a blog post on what three questions to ask when deciding whether to embark on an AI project. I am not advocating analysis-paralysis, just a little critical thinking.
Some new technologies, like Snowflake, can completely change the game in terms of where your data warehouse team (we all still have those) spends its time - 80% of their time tuning or 10%?
Also, start with what you have to generate quick benefits.
Focus on Where You're Going
Define a Data Strategy. Even if it's a few Powerpoint slides, the benefits are overwhelming. Would you consider starting a business without a business plan? As they say, if you don't know where you're going, you will never get there. Apply asset management disciplines to select information assets that will pay off.
"Data literacy is the ability to read, write and communicate data in context. This includes an understanding of data sources and constructs, analytical methods and techniques applied, and the ability to describe the use case, the application and the resulting value." (Gartner)
What does Gartner say?
In a recent survey, here is what Gartner found to be the activities critical to data and analytics success:
And then, the essential data-driven competencies:
There are smart people everywhere in organizations. Once they get the basics, they will be able to formulate what they need from data. They will work with the data professionals and guide them on what they need. But there is a minimum they need to know. They need to be able to recognize the situations where data can help them, and they need to know where to look. This is about cultural change and transformation. For an organization to become data-driven, it starts with its people. Don't forget about the people!