Consolidating, Integrating, and Using Data to Drive Business
I’ve heard them referred to as Rabbits and Vampires… things that reproduce rapidly and others that stare at you in the dark with gleaming red eyes. To us- in the Chemical manufacturing industry- they are sensors, mobile devices, and components of machines, copiers, and computers—really any device connecting to the internet with an on/off switch. They all collect information and are sources of yet more data, waiting to be consumed.
"Chemical companies have lots of information but can’t use it effectively to drive business. Data is both difficult to access and needed by more applications"
Consuming data sounds like the newest sparkling, shiny coin, providing a simple path to being able to provide insights with the press of a button. Unfortunately, nothing is that simple.
As many Chemical Manufacturers, Financial Institutions and Oil & Gas Exploration companies will attest to, the first step of using data to drive business decisions is to figure out what the data means or “data mining” which is really hard to do well. It also requires vast amounts of data, patience, persistence and most of all, minds creative enough to see the hidden story that lies within.
In my experience, the keys to success can be summed up simply as:
● Don’t let the sheer volume be overwhelming
● A little side trip isn’t a bad thing
● Get Business leader engagement early and often
● Don’t try to figure out the solution before wading into the data
Many companies want to start with large volumes of data, encompassing multiple years, all products and customers. Instead, start with smaller, manageable sets, such as a segment of the business that is a good representation of the entire business.
Don’t worry if all the data doesn’t reside in the same database or system. Many of the big data tools are made to consolidate data from multiple sources, which saves valuable time and money by not having to worry about consolidating all the data in the same place and managing yet another database.
Speaking from experience, the time and money savings can be crucial, both in obtaining the business value from the insights delivered, but to also prevent mining fatigue – the condition caused by trying to develop business centric reporting and dashboards with too much data and very little clarity on the overall objective.
Research and Analysis
Next, ignore conventional wisdom to “keep on message”. In order to be able to use the data, we have to test and follow some leads, no matter that it’s a little divergent. This is also where the unintended benefits and “ah ha” discoveries are made. For example, we were trying to design reporting dashboards for production exceptions and ended up also creating a real-time visual production report, by line with changes in color depicting product changes by line. It also provided a visual indication of the number of hours each line or asset ran in each 24 hour period, showing massive amounts of change-over that no one anticipated or really understood.
By following the discovery thread, we started to uncover root causes of these significant changes over, which were results of decisions that were made in good faith, but without any real facts. Couple this with an incomplete view of a business process and you have a significant issue. The good news, is with more clear information, many of these types of challenges are simple to fix and measure the results.
A critical component at this stage is also to engage the owners of the information or process. There is always a person or two who have the in-depth information about the data and how the data is collected, i.e. what is being done with it today, where the gaps are, and often, also has a business perspective of how it could be used if only they could get one more thing.
These Uber consumers are able to find the needle in the haystack for themselves, but typically struggle to explain it to anyone else and ultimately are unable to convey the business value. When we have paired these data experts with business leaders, we are able to identify the trends and insight in the data, and quickly turn it into business reporting, which can be consumed by operating leaders.
Once understandable data is in the hands of these leaders, they are able to insert it into their daily business cycle, providing real world feedback of what part of the data is useful and what changes need to be made to improve the usefulness. This will require a few rounds of validation and changes, so it’s important to get data in their hands quickly as well as leveraging tools that allow for minor changes, such as a formatting, to be done easily so that on-going maintenance doesn’t become an impediment.
Develop the blueprints before building the house
Lastly, keep an open mind and don’t try to rush in to solve the business needs, with the tool kit of existing tools and solutions. It’s tempting to have ‘a one size fits all’ approach, keeping the mindset of a small foot print. Sometimes, a true fit for purpose solution may be a better approach, especially if it’s a tool intended for business users.
The areas of digitization and Internet of Things aren’t entirely new; however, being able to manage them effectively is. Many of the standard, off the shelf reporting tool sets are great for business and financial reporting where decision support topics are about profitability, accounts receivable and sales. Statistical analysis and modeling can share the same tools, but it may make more sense to invest in a small, but specialized set of software solutions and services, that are built for purpose. When the need extends to machine learning and neural networks, there are tools that are not only built for purpose, but often come with significant content that dramatically shortens the development time.
Providing tools that let users drive can be the difference between good information and useful information driving business decisions.
The world we live in, with increasing digitization from the Internet of Things, is exciting and the perfect environment for Information Technology professionals to stand out and demonstrate their ability to address real world opportunities. It’s little wonder that progressive, data rich, information starved industries are looking to their CIO and their teams to lead the revolution to use the data to drive decisions and run our businesses on a fact based foundation.