Business transformation has emerged as one of the most critical endeavors in today’s enterprise, a key undertaking that, done properly, can ensure the livelihood of a business for the foreseeable future. Done wrong, business transformation can leave an enterprise in even worse shape than before, facing the prospect of having to spend big to fix it.
Today, many business transformations are based on cloud technologies and a host of other technologies such as robotic process automation, machine learning and artificial intelligence, industrial IoT sensors, and more. The common thread that links all of these technologies, though, is data. None of these technologies is useful unless it has the ability to access and produce high-quality, consistent data which can, in turn, be used to make business decisions.
When starting a business transformation project, it’s critical to consider the quality of your data as early as possible. “The successful approach is really to start in the early phases of the implementation, sometimes as soon as during requirements gathering, but more often during design,” says Dan Sorenson, Digital Risk Solutions partner at PwC. “Key design decisions often rely upon assumptions gathered during requirements, but rarely do they include a full range of considerations for data.”
Too many business transformation processes begin by taking small samples of data to use as a reference, says Sorenson, and that sample ends up being not representative of the enterprise’s entire data lake. “Maybe it’s 70% accurate, maybe it’s 90% accurate,” he adds, “but you don’t really know. It’s really scary for companies to not know how much of this data is going to be inconsistent with the requirements of the business.”
The bottom line: As you embark on business transformation, a leading practice is to complete an analysis of your enterprise’s data upfront. This assures that design decisions take the overall quality of data into account when it’s needed most. If data needs to be remediated or cleansed, this process should also be front-loaded, giving digital transformation tools a more consistent data store with which to work. “Doing this as soon as possible in the process is very helpful,” says Sorenson. “If you wait until you start working with mock loads, it’s often too late – or it leads to overruns and becomes more costly to make changes.”
For more information on this topic, please listen to this podcast from PwC.
For further reading on how to manage data at the enterprise level, check out the Enterprise Data Strategy series on the Digitalist.
PwC is an SAP global partner.