Recommendations
- Explore AI and IoT integration opportunities across business processes to yield wider benefits simultaneously
- Build PoCs and MVPs to test the solution with a “fail fast” mentality to explore more avenues for value creation
Myth 3: All data is good data
Data can come from a huge range of sources – business applications, website analytics, industrial equipment, wearables, social media – the list is endless. But quantity does not equal quality.
“To build a trusted AI system, companies need to have trusted data – data that is from a reliable source, is compliant, is accurate, is clean, is relevant and is transformed for intended purpose,” says Beatriz Sanz Saiz.
Poor data can stem from poor data-collection practices, which have to be stopped. “Traditional companies capture data from multiple applications such as CRM and ERP, from call centers and so on,” says Sanz Saiz. “People collecting this data don’t necessarily care about data quality. What companies need is to have a strong data governance structure in place to control and monitor the data they use to train AI models.”
Unfortunately, most companies still lack a strong data governance model. Our research shows that only 8% of corporate companies have a governance function for emerging technology that is well established and active. But AI-focused corporates are taking the lead in this regard, with 11% of them saying their governance function is well established. “AI could be part of the solution to fix data quality. For example, using algorithms to extract information from handwritten forms, fix data entry errors and match information across systems”, says Gavin Seewooruttun, EY Asia-Pacific Artificial Intelligence and Analytics Consulting Leader. Although only a small uplift, data governance is one of the most difficult and time-consuming elements of establishing a robust AI strategy, so even small gains can be significant.
Recommendations
- Implement a strong and robust data governance structure to ensure trusted data informs business decisions
- Define a clear strategy to capture, cleanse, stage and consume data so that it is secure and provides AI applications with the best quality sources
- Introduce validation checks to ensure AI applications leverage only trusted data
Don’t let myths get between AI and your success
AI's potential to benefit companies, customers and society is enormous. But with hype comes confusion, and businesses must work hard to question misconceptions and look at AI with clarity –and define how AI investment can best deliver long-term value.
They can start by investing in robust, well-governed AI solutions that are deeply integrated not only into the core of their business operations, but also with other emerging technologies.
Greater adoption in turn leads to new mindsets that benefit the core business: for example, a well-established governance function for emerging technologies could help mitigate some of the perceived risks of AI and encourage a transformation culture.
And there are already success stories: 85% of the surveyed leaders who have seen financial gains are leveraging data and analytics insights to increase speed to innovate. To start realizing the benefits of AI, the best approach is “future-back”: organizations should ask themselves whether their business will still be relevant in two, five or ten years’ time. Then by working through future-back scenarios that incorporate AI into the strategy, organizations can flex and adapt to ensure they’re following a path to maintain relevance today — and 15 years from now.
By adopting redefined value levers –putting humans@center, enabling technology@speed and driving innovation@scale – and using their purpose to guide them, organizations can build adaptability into their businesses and enable bigger, better transformations, with AI prioritized as a long-term driver of value.
About the research
The analysis in this report draws on an extensive program of quantitative and qualitative research.
A survey of 570 C-suite and senior business leaders was conducted across 12 countries (US, Canada, Brazil, UK, Germany, France, Italy, Spain, Australia, Japan, China and India) and nine sectors (consumer products and retail; energy; health and life sciences; tech, media and telco; industrial; financial services; education; transportation and logistics; and hospitality). The 570 companies were split into two categories: 500 corporates and 70 start-ups. The data in this report refers to the 500 corporates only, unless otherwise stated. In addition, a number of in-depth interviews were conducted with leading digital transformation leaders.