This article originally appeared in the April 2015 issue of Smart Business Philadelphia.
Businesses today have data on every touch point of a transaction, whether it is among their employees internally, or with suppliers, vendors or clients. As a result, organizations are collecting a wealth of information and putting it to use.
"The challenge that management faces is to harvest big data and then leverage it as knowledge to make more informed decisions," says Sassan Hejazi, Ph.D., director of Technology Solutions at Kreischer Miller.
"There is a lot of data available, so much so that it can be overwhelming. If it can’t be managed well, an organization is not really leveraging that asset. It becomes a matter of effectively managing the information overload."
Smart Business spoke with Hejazi about harnessing the power of data to create knowledge.
What is the initial step in creating knowledge about a customer base?
The first step is to collect the data. A lot is already being collected, but you have to make sure it is good, clean data, with no mistakes or errors in it. For example, at a supermarket checkout, if the cashier can't work with the UPC code on the product, the cashier just types in a miscellaneous code. That is an example of an error that has been entered in the data. While there is no perfect data, the degree of data quality is very important.
After collecting the data, how does an organization process it?
The data then needs to be organized. If it is not organized in a proper fashion, it is going to be difficult to massage that data to achieve higher levels of business intelligence.
The data comes from different systems, and every system has its own format.
To overcome that issue, organizations can implement a process called data normalization, which involves data scrubbing so management can use the data assets to support knowledge creation. The data needs to be normalized and stored in an environment called a data warehouse so it can be easily retrieved and analyzed.
There also are smaller warehouses that are referred to as data marts, and these can be useful for departments like sales, financial and production.
How can management then analyze the data?
One of the most popular trends is building a management software dashboard as a tool to present the information to support decision-making. Dashboards are highly customizable, based on each manager's preference and depending upon the function and nature of the data the manager is reviewing.
A good dashboard tool will allow flexibility. As managers become more proficient in using the dashboard and their needs become more sophisticated, they will want to adapt it to address their needs. The ability to modify the dashboard with minimal help from IT is important because it will allow managers to experiment.
What are the key advantages of these tools?
When managers make a decision, whether individually or in a collective fashion, dashboard tools can help ensure that everyone is on the same page. Stakeholders can see an up-to-date picture of issues, performance levels and key performance indicators within their organization. They will become more knowledgeable about trends and challenges, what's working and what's not working. The organization then will reach a higher level of knowledge, leading to improved performance, better decision-making and the ability to develop new products and services.
Are management dashboards difficult to implement?
Dashboard tools are becoming more powerful and easier to use. There are a lot of new data visualization and analysis tools, and companies can test drive them via cloud-based applications.
Previously, these technologies were very expensive and required extensive resources. It is becoming more affordable for middle market organizations to experiment with business intelligence-type solutions without investing in major hardware and software platforms. It allows them to gradually build their business intelligence capabilities using a crawl, walk and run approach. ●
Sassan S. Hejazi can be reached at Email or 215.441.4600.
You may also like: