‘Garbage in, garbage out’ still plagues IT in an age of digitalisation and machine learning,Cloud and Infrastructure,Big Data and Analytics,Software ,Cloud,sopra steria,data governance,Data warehouse,data quality,Andy Whitehurst,Machine Learning<\/b>
Organisations are still plagued by poor-quality data, despite decades of data quality and data cleansing initiatives — and could undermine digital transformation projects in the future.
That was one of the key messages of the Computing IT Leaders’ Dining Club this week, which was sponsored by services company Sopra Steria.
«Data quality is a prerequisite for so many projects these days,» said the CTO at a major communications company. «There’s some really interesting technology you can apply to get some valuable insights. But you need an environment that’s viable. »
According to Andy Whitehurst, a principle consultant with services firm Sopra Steria who has spent the best part of two decades working in data warehousing projects, the old problem of ‘garbage in, garbage out’ remains. «It’s not necessarily about perfect data, but data that is good enough to base business decisions on. »
The IT Leaders’ Dining Club was held at Searcys at the Gherkin, at 30 St Mary Axe in Central London. Because IT Leaders’ Dining Club events are held under Chatham House rules, the attendees — with the exception of Sopra Steria — cannot be directly quoted.
The issue of data quality will only grow as organisations consider predictive analytics projects, and find that their data is not good enough. «Everyone wants to try predictive analytics, but can you trust the data,» asked the CIO at a major financial services organisation.
«The data is coming from many different, complex systems so you can’t be 100 per cent sure about the data,» he added. And that’s before the issues such as context — certain data may only be relevant in particular contexts — and data protection: does the organisation even have the right to use data from some systems for the purpose of predictive analytics?
«So, what are the controls or processes to make sure that the data is checked regularly? » he added.
That, said Whitehurst, brought-in complicated issues of data governance. «The data quality and in-built governance [you have at the moment] will be about whether it’s fit for purpose for a particular task, not for combining with other data for gaining new insights,» warned Whitehurst.
The next Computing IT Leaders’ Dining Club will be on Tuesday 14 March at the Shangri-La Hotel , The Shard. Membership is free to all qualifying IT leaders (CIOs, IT managers, IT directors and equivalent).<\/p>\n