For more than 50 years, enterprise IT groups have argued that integrating data, putting data into central databases, single copies of data, integrating transaction data with operational activity, and more saves/makes dollars. These initiatives are a 1st generation in exploiting enterprise data.
As information technology spending moves to Sales&Marketing, Engineering, PLM and other spending centers exploiting data becomes more complex AND expensive. Data integration can be even more difficult in the 21st Century era of multiple ‘closed garden’ cloud services, multiple regulatory jurisdictions, and massive unstructured data feeds. That difficulty is a barrier to investment – and an opportunity for investment. Enterprises that act to exploit the data by making good investments can be recognized as they move for market share, margin, and market depth at the expense of their competitors.
Early ‘first mover’ success stories in AI and Machine Learning have often come from enterprise IT where the long-term systems engineering, attention to data quality, and ongoing curation and data modeling leads to successful exploitation. Investment in enterprise data is needed to successfully exploit data in your enterprise in the future.
To get started, creating a data inventory (along with the applications’ portfolio) and starting to gather documentation about data is a good start on exploiting data to get money. Sequencing projects and handling some mandatory tasks may be a good process to build the documentation for directly and enable the data focused project.
The sooner quality effort starts, then the sooner enterprise-level AI/ML/DS projects are likely to succeed at a lower risk, with faster turnarounds, and likely, better costs. In the current era re-planning some initiatives and recognizing real costs may actually be a good step while cutting overall budgets.
Kommentare