Let’s deal with the big issue right away. It’s intimidating to think about becoming a data-driven organization.
It’s impossible not to be overwhelmed by the challenge of collecting all relevant data about your business. Each customer interaction, each production process, can create dozens of pieces of information.
So why put yourself, and your business, through it? Three reasons.
- Data is necessary to measure performance and drive decisions.
- Data serves as the foundation for making confident predictions about the impacts of decisions.
- And finally, it’s not as difficult as you think. Help is available.
Aunalytics and Data Realty, two Indiana businesses, partner to help clients build an infrastructure to collect and maintain data that can be used to develop insights about where the client is and what the future may hold.
“It’s common for businesses, when they have a problem, to look at their data to try to understand why,” says David Cieslak, Vice President of Predictive Modeling at Aunalytics. In fact, organizations may begin to look at their existing data only when they have a problem. Often, once they begin to assess what they do know, they discover they are missing crucial information.
Each Journey Begins with a First Step
This discovery of the issues regarding current data and data collection practices is exactly where most organizations start on the journey to becoming a data-driven organization. And remember, each significant journey begins with an initial step.
Phillip Whelan, Data Scientist and Deployment Engineer, and Nick Weiler, a Software Engineer, are two members of the Aunalytics team. They help clients get over the intimidation of data by suggesting they begin with two basic first steps. First, identify existing data sources and then ask if data that is supposed to be collected is, in fact, being captured.
“The client begins to understand what they have and, more importantly, what they do not have. Right away, they see something important to fix and relatively easy to improve,” Whelan says.
Next, address any holes in existing data. Is it really not being collected at all or has it been collected and stored elsewhere? Decide if missing data is essential. If so, establish a process and begin collecting missing information.
Then, confirm that the data makes sense. If it’s not being accurately collected, then it will at best waste the analyst’s time and at worst, lead to incorrect conclusions.
Data Stewardship is Critical
It is crucial to document the business practices involving data. For example, if one event happens, what specific action(s) should be taken? Business should document the steps to be taken, including who is responsible for collecting the data and who is responsible for taking the actions.
Incorporating data collection into the larger business process is a key element of successful data-driven businesses. So work toward establishing explicit rules for data stewardship. The data steward is not only responsible for ensuring data is collected, but also for making sure the data continues to be collected properly if there is a change in software or other business processes. While it is good that people throughout the organization are involved, management drives the organization, sets the example and makes data collection an essential priority.
Finally, identify data both inside and outside of the business that isn’t being tracked. For example, power outages may be obvious, but any unusual event should be recorded because analysis can often discover unexpected relationships. Information about weather may help explain logistics issues. Perhaps trucks couldn’t get through or snow kept workers at home. Even something like the air conditioning being down for an afternoon could explain an uptick in customer dissatisfaction if overheated agents were less than patient with customers.
Aunalytics’ data science team helps guide clients through this process, aiding to determine what is important and what is necessary. Further, the team can help find ways to collect and integrate outside data into predictive modeling systems.
As a client moves through these steps, new possibilities emerge. “Rather than using data to report on the past, data science lets us look into the future and predict what will be,” says Cieslak.
Expect Results from Data Analytics
One Aunalytics client keeps track of interactions between customer service representatives and subscribers. While the company has a good understanding of its customers and of the reasons they give for canceling service, Aunalytics was able to use that information to help the company anticipate which subscribers were more likely to cancel so proactive intervention could be directed toward the most vulnerable.
Sometimes data can confirm what is suspected; but with the right analytical tools, data can offer a new interpretation and create new business opportunities. A financial services organization did not want to make loans to borrowers generally thought to be of higher risk. But analysis of the company’s data showed that not only was the organization making more high risk loans than initially thought, but also that those borrowers were no more likely to default than customers the organization traditionally thought of as more desirable. Data analytics helped target new customers and grow that area of the business.
These are long term gains, improvements made after a company has experience at collecting data.
But even the first steps can have measurable benefits. The client’s first advances come with plugging holes in existing data. One client found that more than half of the information it thought was being collected about existing customers was either missing, misfiled or inaccurate. If a business wants to analyze potential customers, it first has to know about current ones, so repairing these types of holes starts the client down the road toward really understanding its clients.
It is a long-term and iterative process. Each new layer of data builds upon a previous layer and as time goes on, the organization benefits from ever-deeper levels of understanding about its business.
Data Analysis Can Offer a Competitive Advantage
Traditionally, organizations have thought of the Information Technology department as a cost centers. However, when they get to be good at data collection, IT may become a center to save costs, even a revenue center.
Every organization is using data, whether for marketing, risk management or sales, but few are using data as effectively as they could, says Kyle Davis, a business analyst at Data Realty. Working with Data Realty and Aunalytics, a business analyst can create a system that lets the client do analysis when needed instead of asking IT to run a report that may take days to create. The objective is to create a platform that gives the client a timely knowledge of what is happening and present it in a visually compelling way that allows the client to see relationships and make informed and immediate decisions.
Whelan doesn’t mean to discourage any potential clients, and he has simple advice for organizations interested in using data to look into the future. Simple, but intimidating. “Collect everything. Document everything.” he says.
“If you aren’t doing it well, your competitors are. They’re getting better information about your customers and your potential customers than you are, and that information will help competitors take business away from you.”
And that’s what is really intimidating.