Data-driven Decisions: Is Data Driving You in the Right Direction?
“In God we trust, all others must bring data.” W. Edwards Deming
Leveraging data driven insights across the IT organization to drive better, faster business decisions and outcomes will likely be the deciding factor in who thrives in the app economy.
To prevent data-driven disasters, it’s crucial to continually examine data quality and analytic processes, and to pay attention to common sense and even intuition
What we don’t lack these days is data. We have big data, small data, data warehouses and data lakes. Is more, better? How much data do you need to make a better decision? Well, that depends on what you are looking for. Deriving true insight from all that data is where things get tricky. Understanding the core of the problem and asking the right questions are pre-requisites for making a good data driven decision.
How Data can Mislead
“Facts are stubborn, but statistics are more pliable.” Mark Twain
While well-analyzed data can lead to solid decisions, it is not necessarily true that every data-based decision is a good one.
Business Intelligence tools make it easy for employees to mine their own data and draw their own conclusions but easy to use does not necessarily result in correct analysis. There are many ways data-based decisions can mislead, such as asking the wrong question or confusing correlation with cause-and-effect.
In addition, bias toward a particular outcome can lead to poor decisions, either by basing analysis on limited data, or data predisposed to a desired result, or jumping to a conclusion as soon as you get a result you like.
To prevent data-driven disasters, it’s crucial to continually examine data quality and analytic processes, and to pay attention to common sense and even intuition.
Asking the Right Question
“Computers are useless. They can only give you answers.” Pablo Picasso
In getting to the core of the problem the human factor is important. Instead of starting with what data is available to you, start by asking what you’re trying to achieve for your business. We tend to rely on a hunch that comes from our experience, so having data to qualify and quantify our hunches helps with our question-defining process—learning to ask the right questions is an iterative, fluid process.
Keep the lean start-up mentality in mind. Successful, lean start-ups develop a vision for a product and validate that vision with customers until it reflects a consensus. Then they develop a Minimum Viable Product (MVP) to satisfy that vision. During development, there is ongoing testing by the customers and revision, as needed, until the result solves a real customer problem. For startups, it is a matter of survival.
Lean start-ups don’t collect massive amounts of data, instead they collect and measure relevant data to answer a specific question. But how do they know what data is relevant? They combine data analysis and customer input and look at three key points.
• Is the customer problem reflected by their budget? Spending reveals the gap between nice to have and essential.
• Are they talking to different personas within an organization or only to a select group? A select group’s opinion can be valid but only if the group reflects a big enough market to make it viable.
• Do the data and the customer tell the same story? If not, they know when to pivot.
Focus on the Road ahead
Analytics have always been with us but have typically focused on answering the question, “What happened?” We learned a lot from figuring out what happened and refined our actions going forward but the faster you are moving the more important it is to keep your eyes away from the rearview mirror and focused on the road in front of you.
Gartner’s analytics maturity model shows how far we have come. Descriptive and diagnostic analytics gave us the “what happened” and “why it happened.” Predictive and prescriptive analytics give us the “what will happen” and “what should I do?” Some prescriptive analytics can go as far as automating corrective action.
As the app economy advances, IT needs the ability to quickly predict issues before they occur and prescribe/automate corrective actions to stay ahead of the game and keep the transaction secure. The good news is that today’s analytics make that possible. The ability to measure the flow of data to your customer, see customers’ reactions to your data and protect their transactions with context-based and risk-based analytics transforms the experience.
“If you can’t measure it, you can’t improve it” Peter Drucker
For every app there is performance and user experience data that can be measured and analyzed. Look for apps with embedded analytics that eliminate the need for a separate analytics platform.
Analytics solutions that directly measure the flow of data to your customers by your apps gives a real view of their experience. This view can answer questions such as, how is my app performing to this request? What did the user do at a particular point in the app process? If a new feature isn’t being used is it because customers can’t find it, or it doesn’t work well on certain devices, or the performance is poor overall?
Believe what you see. Seeing the entire customer experience—combining user behavior data with app and underlying infrastructure performance data—can reveal issues you didn’t anticipate, or were hidden in siloed, point monitoring tools. Don’t ignore them, even if they uncover the weakness in your carefully crafted question. Consider them as insights and take advantage of the data to make improvements.
Your customer experience will be fluid so measure it on an ongoing basis and respond as you gain insight. Established and agile Dev and Ops methodologies enables quick and continuous improvement cycles.
Advanced, embedded analytics do a good job of identifying anomalies that may be linked to online fraud. Analytic solutions that embed context-based analytics to calculate risk, evaluate each transaction against a wide range of historical and behavioral data by answering questions such as, is this transaction originating from a location and a device typically associated with this customer? Is this a typical purchase for this customer? Risk analytics look for subtle anomalies that may indicate fraud and can offer automated actions such as requesting additional authentication to prevent illicit access.
By keeping identity access front and center, embedded risk analytics can keep identity secure without negatively impacting the customer’s experience.
How IT can Leverage Data-driven Insights
The focus of analytics has evolved from using data to explain the past to using it to predict the future. It’s no longer enough to locate and mitigate system failures as quickly as possible after they happen. Now the job of IT is to anticipate issues that impact user experience and take the necessary actions to prevent them from ever happening in the first place.
To do this IT must look for data-driven apps that embed data science, machine learning, and powerful algorithms in virtually every app, automatically capturing and analyzing data to generate insights and empower better, faster decision-making. These solutions can give your business a competitive advantage by delivering near real-time predictive and prescriptive analytic capabilities to satisfy the expectations of even the most demanding customers in the app economy.