By Virginie O’Shea, post-trade fintech analyst and advisor
In my role as an industry analyst I have spent a lot of time over the last decade talking to data management professionals of all kinds – from data scientists working on short-term, frontline business projects to C-suite executives tasked with championing enterprise data strategy. The latter role has a relatively high turnover rate – individuals on average tend to stay in chief data officer roles for less than three years – but their experiences can provide a font of knowledge for anyone working in data management. I thought I’d share five key takeaways I’ve gleaned from my years of helping these teams build their business cases and benchmarking themselves against their peers.
1. It’s really challenging to embed a data culture in the business, but keeping things simple works best.
Make the outcomes and processes understandable at the business level. Trying to get portfolio managers to take responsibility for their data? Explain what using specific formats and technologies means to their day-to-day job. Don’t give them too much to chew on – unless they’re really interested – and keep their governance responsibilities light, so the load is shared across teams.
2. When establishing your plan of action, identify key data sets across functions.
The more functions you can involve, the bigger the task but the less likelihood you have of one pulling the plug and the whole project falling over. As anyone that has worked on or with a standing committee will tell you, sometimes it’s hard to get functional heads to cooperate (and even get along) as they have very different goals. Just think of compliance officers versus trading heads, for example – one set is keen to ensure the details of compliance requirements are met, the other is keen to take action as fast as market opportunities possible. Don’t underestimate the cultural problems caused by trying to force these people to work together and with a data team. Tackling gathering their priorities and inputs one by one is sometimes the easiest option.
3. Use technology to monitor your progress in cleaning up data.
There’s a lot of amazing tools out there to help with data quality monitoring (and I’m not trying to sell you any of them). Deploying technology to help assess the quality of your data against the benchmarks your business people and end users have set (along with the data governance team), means some of the painful manual legwork is reduced and you can focus your efforts on improving rather than tracking data. But don’t expect the technology to work miracles – it’s a support for your efforts not a cure for data quality problems.
4. Have a communication plan and stick to it.
Make sure the C-suite executives have regular and easily understood updates on progress – ie, we have achieved X by the date we expected to. Make sure there is a centrally accessible place for this information and that there are several articulate champions from the business to explain how progress has improved their day-to-day activities. In terms of informing the masses – keep these concise and regular with little to no data management or IT jargon.
5. Invest time in setting up and supporting a feedback loop.
Find your business champions and those that understand the data best – they may not necessarily be the same people – and talk to them regularly. A regular honest feedback loop from people on the business side and from those that live and breathe data will help keep your efforts on track.