Migrating data and apps to cloud can be a game changer for financial institutions. It can also be more expensive than expected and raise issues around data quality, governance, security, access and integration. A clear data strategy with a defined goal, and an understanding of what can and cannot benefit from a move to cloud, are a good starting point for successful migration, but where is the return on investment?
This question was debated at A-Team Group’s recent Data Management Summit London, with Oli Bage, founder and co-chair of the Cloud Data Management Capabilities (CDMC) initiative at the EDM Council, and head of data and analytics architecture at LSEG, moderating a lively panel session. Joining the panel were Hugh Davidson, director, enterprise data architect at Marex; Indhira Mani, group head of data transformation at Danske Bank; Alpesh Doshi, managing partner at Redcliffe Capital; and Chris Brook, co-founder and head of architecture at FINBOURNE Technology.
ROI from cloud deployment
An early poll of the summit audience asked, ‘how does your organisation view the ROI it is getting from cloud deployment?’ Some 54% of respondents ticked the box ‘ROI? We haven’t seen any yet’, 26% said they are struggling to achieve ROI, and 14% that ROI is below expectations. Just 6% said ROI is in line with expectations.
“Moving to cloud and getting ROI is difficult, you need to be clear on what you are doing and start by discovering the benefits of cloud as well as the costs,” said one panel member. Noting the role of cloud vendors in costs, another said: “People say cloud will be cheaper, that is up to the vendors. I have seen a lot of firms just pick up data and move it to cloud to cut costs. Overall, costs will be higher.”
On migrating apps to cloud: “Migrating any apps to cloud is hard. If you try to, say, unpack cash management and all the 400 systems it uses and come up with a migration path to move all of that to the cloud, that’s a massively risky undertaking, the business could stop running. A misalignment of expectations is one of the biggest challenges.”
On a more positive note, and describing cloud as a ‘game changer’, one panel member noted the benefits of faster insights from data in the cloud, better decision making, and improved customer experience. “With disparate data sources, moving into cloud is a game changer because it gives you transaction history, customer interactions, market data, everything coming together. It can give the customer a very personalised offering.”
Underlying challenges
The panel acknowledged these benefits before moving on to discuss underlining requirements of data migration. Data integration with existing IT infrastructure was the biggest tech challenge noted by the audience in a second poll question. Solutions proposed by the panel included a use case approach, and a learning approach of putting a new product line in the cloud first to prove success and then using the same operating model to migrate additional data and apps.
Also considering integration: “When deciding which cloud provider to use – they all offer similar services – you need to think about what works with technology you already have and what will be the best fit.” On issues of data governance, lineage, and controls, a speaker commented: “Important things don’t change, for example you still need data quality and lineage, so 100% migration is not viable in the short term.”
Business value and the next frontier
Coming back to the thorny issue of ROI, Bage asked the panel how business value could be optimised in the cloud. First, said one speaker, ROI and business value should be split to provide a better understanding of value, and value should then be measured across various areas of the business that have migrated to cloud. FinOps was also mentioned, with a speaker saying: “Cost management needs to be future looking, not retrospective. It needs to be upfront so everyone can see the cost of cloud.”
Looking at the next frontier of cloud, generative AI was a top talking point, but may not make it into the cloud at some organisations. Again, cost was the challenge. “It is too expensive for big firms to run generative AI in the cloud, it will probably run on clusters on premise.” That said, all agreed that it would be a step change in data management innovation.
View our full agenda and more details here.
Subscribe to our newsletter
