As we move our enterprises closer to a self-service environment, our business users may achieve some enterprise benefits. But are those benefits worth the cost? Some business-funded projects can cause more issues and expenses than expected. Here are some thoughts.
Out of sight, out of mind
If we in IT ignore how the business users prepare, integrate, transform and use data, we could be causing more issues. And those issues could involve more work and more resources. Partnering with business is the most important skill set. If we are not partnering with the business users in this business-funded endeavor, the data may be used and/or reported incorrectly. It's so easy these days to put our heads in the sand and ignore what the business is doing with the data, thinking that it's not our problem. Does this come from being burned in the past, or from lack of resources and authority? Interesting questions. Clearly, it is our problem, too.
Budgeting and project funding move to the business
I see more and more companies where project funding flows from the business side without much regard for IT. While this is not altogether a bad thing, it can create dysfunction in the continuity of knowledge across the enterprise. That said, when project funding is completed by the business, we have to make sure we have a seat at the table and that all aspects of implementation and production are taken into consideration (i.e., data usage, performance, security and governance). One of the good things about project funding via the business is that it lets the business know exactly what it costs to maintain production environments (not only in terms of money but also in terms of people).
Some of the Issues IT faces as we try to partner with the business include:
- Software upgrades – yes they happen. There just always seems to be an upgrade to a code, an ETL tool or a data modeling tool.
- Data changes over time – you have to look for other instances of the same data throughout the enterprise. I find it amazing that data can proliferate redundantly across the enterprise at the speed of thought. I like to do an exercise where we identify all the places the same data resides. Maybe they get the data from the data lake, data warehouse or an operational data store. Look there first for pulls or extracts.
- Server upgrades. The server upgrades take time, and should not affect our data – but consider running a full backup first, as a precaution.
- IT Involvement – it's crucial for IT to be involved up front. This entails not just people, but also computing resources. For instance, the project may need another server, processor or more storage.
- Data governance, security and performance. Our enterprise DBAs usually understand performance of data based on usage. We also have to consider where to build an index, table design, etc. IT has to be involved in the implementation of a project that needs to perform efficiently.
IT and business groups need to be partners in business-funded projects. If they're not, resources may be consumed with no way to project future growth and requirements. Data may be released to an entity outside of the organization, or used inappropriately. There is no out of sight and out of mind when it comes to enterprise data assets. Data usage must be understood and governance must be conveyed across the enterprise.
Download a TDWI best practices report about data preparation