Data contamination is a natural result of reliance on decentralized point solutions. It happens when users introduce unverified data into an environment and share it with others. Without a governance framework in place, there’s no way of verifying the accuracy of that new data. This model might work between a few individuals within a small team; but imagine the consequence of relying on this method across hundreds of users and dozens of lines of business. It’s going to be chaos.
Ultimately, everyone is working towards the same goal: they want their business to be successful. But let’s face it—different teams have different priorities. IT is primarily focused on managing process and ensuring trusted data. Meanwhile, business teams generally view security and governance as obstacles to timely decision-making. With these conflicting priorities, it takes a strong commitment to make analytics work for everyone.
If organizations want trusted, governed self-service analytics, they need to put mechanisms in place to make it happen. I like to break these down into seven key considerations that can help teams collaborate effectively:
As with any initiative, getting buy-in is crucial. The entire organization needs to recognize the value of having an enterprise-wide data governance initiative—whether it’s through standardized technology, systematic reviews, or appointing data stewards. This needs to go beyond executive sponsorship—the heads of all business teams need to commit by defining goals and guidelines and making sure team members are held accountable.
Each business unit needs representatives—let’s call them “report jedis”. Ideally, inter-team dialogues will be a set routine aimed at getting everyone on the same page. It’s the stakeholders’ responsibility to ensure their teams adhere to established processes in order to build a more collaborative organization.
Open communication and collaboration between IT and business teams is the best way to ensure everyone’s needs are met. Business teams need to be transparent with their needs and not treat IT like the dark side. IT needs to loosen up a bit and be more flexible. But neither should be forced to extremes. Your analytics platform may be able to provide monitoring capabilities to help bridge that gap, but the technology itself can only take you so far. To make it work, governance processes need to be fluid and open, and IT needs to continuously monitor and prioritize how they promote essential KPIs into a governed framework—and be prepared for recurrent change.
It’s naive to think that all critical data assets exist within the enterprise data warehouse. Essential KPIs are often located across the enterprise on personal spreadsheets and in cloud applications, and someone needs to curate it all. That’s where the Chief Data Officer comes in. According to Gartner, in the next three years, this position will be filled at 90 percent of large companies. That’s good news. The CDO exists at the crossroads of IT and business, and is a critical player in helping curate data and foster communication between teams.
Standardize, standardize, standardize. I’m in marketing, so I build a lot of presentations. I’m also a Mac user, and given the choice, I’d stick to Keynote over PowerPoint to build my slides. However, there are a good number of folks on my team who don’t have a Mac, and so for every Keynote slide I build, I end up replicating a PowerPoint slide. The time spent doing all that extra work adds up. If there was a standardization of technology, even though it might mean giving up my pledge to Keynote, I could get work done twice as fast. With analytics, as with all technology, it helps to standardize across teams.
The truth is that you can’t fix data governance overnight. With business needs changing so frequently, data governance is going to be an iterative and cyclical process of identifying gaps, prioritizing applications, and promoting assets. My suggestion here is to start small. Pick the most important application and start there. Even if you are only able to certify or promote a single application a month, that’s twelve critical applications every year.
To truly achieve governance, you need to invest in a long-term strategy that includes a vision for growth and change. Quick, one-off fixes work against this. By focusing on short-term goals and adopting the first free analytics tool they come across, teams hamper collaboration and make governance even harder. So when you’re selecting a tool, be sure to ask these questions:
7 Key Considerations for Data Governance: Data contamination is as bad as it sounds. Unverified data sources can poison a company’s well of information, negate the hard work of employees, and set up countless problems down the road. If the thought of contaminated data makes you cringe, you’re not alone. Every company looking to scale up the…