Accenture: Prioritise data to pave the way for AI models
Prioritisation over perfection is the message from consultants Accenture to companies stuck in data paralysis as they prepare for Artificial Technology (AI) technology.
Companies can be overwhelmed by the volume, velocity and variety of their data and find it difficult to access its fourth, and probably most important point, v: value, delaying the transition to AI models claims the report from Accenture.
“My advice is to assess what needs to be done with your data across business functions, but then isolate a small area that is a priority for the organisation,” said the author Jean-Luc Chatelain , Managing Director – CTO Accenture AI (Applied Intelligence).
“This is where you can make a focused, valuable start and gain some momentum, with a view to gradually expanding or replicating the approach over time.”
Prioritisation over perfection
AI projects need data and according to Accenture it doesn’t have to be perfect, but it needs to have enough quality and consistency for useful patterns to emerge.
“The better the data, the better the AI,” is the message in Accenture’s report, Overcoming data paralysis when preparing for AI.
But Chatelain points out that for many companies, there’s a problem: 85% of their data is either dark (whereby its value is unknown), redundant, obsolete or trivial.
“It’s not always easy to determine where you will find value, but to understand the landscape, the data needs to be cleaned up and integrated into your business. It’s all about making sure that the data has a structure and format that will enable you to develop it into the training data you need for your AI models,” he said.
How do you clean up data?
Restructuring the data is an onerous challenge due to format inconsistency. For example, an address could be provided in a variety of ways from New York, New-York or NYC. AI models will represent these as three separate entities unless trained to associate them.
However, Chatelain doesn’t advocate putting data scientists on the task as this isn’t always the best solution as they could feel trapped in “the data dungeon”.
“To make the most impact from data and pursue valuable data-driven transformation, companies will want to avoid this data paralysis and uncover ways to move the AI agenda forward,” he said.
Chatelain’s advice to businesses stuck in data paralysis is to choose 10 pain points where it needs to improve, then rank them and focus on the top two which will achieve a substantial improvement on a key metric.
“Find an example like that in your business and zero in on it before moving to the next pain points. In this way, you secure tangible AI successes, win the confidence of stakeholders and establish methods you can replicate and scale up,” said Chatelain.
The report also highlights the risk of over-ambitious strategies and cites a case study of a global financial firm with large-scale disorder resulting in a failure to change when it came to their AI agenda.
“When we began to help, rather than trying to revolutionise all the organisation’s data in one go, we found the better strategy to be looking at the firm’s most critical pain points and its most valuable business units and geographies,” commented Chatelain pointing out the firm should have achievable ambitions.
“This helped us determine where data improvements and AI would have the greatest impact. Focusing on smaller-scale, but highly valuable, transformation has increased the speed of returns and will accelerate enterprise-wide transformation in the long run. Now, we have established a repeatable process which helps the firm rapidly replicate and scale digital transformation in other areas of their organisation,” he said.
