Ethics in AI
July 11th, 2019
“What can we do to improve data quality?” We asked this question in our recent State of Data survey, and resoundingly, the answer was not “invest more money in it.” In fact, quite the opposite. Our respondents ranked investment as the least effective way to improve the data function.
Tighter controls, stewardship and governance, clear methodologies and benchmarks – the most common answers to the question revolved around these themes. The majority of responses offered valuable insight into process improvements that must be implemented in order to optimise data quality, such as improving input controls, ensuring cleaner data, implementing a data dictionary, and appointing a data stewardship function.
Quality control over the sourcing, input, processing and storage of data obviously is clearly essential. But what struck us was that the two most effective ways to improve, as suggested by our respondents, were neither process-oriented nor technological in nature, but instead called for the human element.
Engaging users is considered was the most critical to the success of the data function, followed by creating clear definitions of the business purposes of data.
Engaging users requires companies to leverage emotional intelligence in addition to business intelligence. Catching people’s attention, presenting data in an intuitive way, advancing their understanding, providing user-centric experiences and actively listening to user concerns – all of these can create a feedback loop that ultimately improves the data and how it is used.
Similarly, creating clear definitions of the strategy behind the data is crucial; building a shared understanding and consensus around what business purpose the data is meant for can go a long way toward improving its quality. Unless IM professionals are aware of and committed to the purpose, they are essentially operating in a silo with no sense of why they’re doing what they’re doing. As is the case with engaging users, creating clear definitions of strategy calls for emotional intelligence, influencing skills, and active listening.
We asked Nicola Askham, an expert in data governance, for her insight on the survey results. She replied:
“These results clearly show that organisations are becoming more mature in their approach to managing data quality, but data professionals shouldn’t rest on their laurels just yet, as there is clearly still some educating to be done. It is encouraging that people are realising that throwing money at the problem does not improve data quality and engaging the users of the data is now recognised as one of the most effective methods.”
As we’ve said before, data is not just about gathering as much information as possible – it’s about using good information to inform business decisions that will drive your success as an organisation. Enterprises must ensure that valuable resources are being allocated in such a way that data is given the proper framework and context to drive results.
This cannot happen without communication, consensus and collaboration from a team with a shared vision – in other words, the human element. When data is used well, it creates a momentum, one that will see more stakeholders on board with the concept that data quality should be the gold standard for the entire organisation, not just the IT and IM departments.
How do you think data quality might best be improved? Is increased expenditure ever the answer? Do enterprises need to look more at their processes, technologies or business strategies? Please share your thoughts below
To learn more about the state of data in 2015, download our whitepaper here