How do you become an AI-fuelled company?

1/ What’s the current state of AI in companies?

Overall, some 50-60% of large companies around the world are doing something with AI. Most of them are just experimenting, doing pilots and proofs of concept. Due to this experimentation approach, it’s hard to get a lot of economic value out of AI as these companies struggle to get systems into production deployment.

Most use of AI is tactical in nature rather than strategic. Objectives tend to be oriented towards process and decision-making improvements, which is not that different from what you can do with analytics too. And as it’s proving to be increasingly difficult to find skilled workers an obvious though quite low-level objective is also the automation of jobs. These less ambitious low-hanging fruit projects using AI are overall more successful than moon shots.

In general, there is a slow movement towards the democratisation of data science, AI, analytics and even automation, which Tom feels is going to accelerate with the rise of generative AI.

Most finance professionals have not been early adopters of either analytics or AI. Probably the single most aggressive adopters are in marketing functions, perhaps followed by supply chain and logistics. Even HR is ahead of finance. Typical use cases for AI in the finance department are forecasting, fraud detection (esp. in the financial services sector), optimising collections and internal and external audit (matching invoices to collections, identifying anomalies).

2/ How aggressively are companies embracing AI?

To assess the level of AI adoption across these companies, Tom referred to the 2022 ‘State of AI in the Enterprise’ report by Deloitte. Divided into two dimensions – the amount of fully deployed AI use cases and the amount of value these organisations are getting – companies are more or less equally represented across 4 categories:

• 27% Transformers: high deployment and high achieving; but even the transformers are not nearly as aggressive as they might be
• 24% Pathseekers: these companies have high outcomes despite having fewer deployed used cases
• 22% Underachievers: these companies have many use cases but are not getting much value from them
• 28% Starters: low deployment and low achieving

A more interesting result from this survey is that both categories of high-outcome organisations use AI to do new and different things, whereas low-outcome companies are much less likely to do so:

• 50% of high-outcome organisations use AI to penetrate new markets and segments
• 48% use AI to develop and enhance new products, programmes and services
• 48% use AI to enable new business models

3/ What are the traits of AI-fuelled companies?

Less than 1% of large companies around the world that are doing something with AI are truly AI-fuelled. These organisations generally outperform their peers in growth and profitability, and they have better business models, make better decisions, have better relationships with customers, offer better products and services, and charge more profitable prices. The companies and use cases in the book are not digital natives but legacy companies such as Airbus, Anthem, Ping An and Capital One, which have to deal with all the organisational change issues involved in any form of business transformation and certainly involving AI.

• They use AI across the entire company, having many systems (10+ to 1,000+) in production deployment and using multiple technologies.
• They are not using AI to optimise existing ways of work but to reimagine and improve work processes (e.g. Shell is transforming the way it does maintenance for refineries and pipelines using drones and image recognition with AI systems which dramatically accelerates the time that it takes to inspect an entire refinery).
• They have a large group of employees who are fluent in AI and who know how it can be applied. These really aggressive democratisation efforts both in data science as well as automation make it really easy for non-technical people to work with AI.   
• They devote a lot of attention to voluminous, high quality and unique data to avoid deploying the same kind of applications as everyone else.
• They make long-term commitments and large investments.
• They have a framework for ethical, trustworthy AI in place (e.g. Unilever has an outside auditing firm evaluating every one of their use cases on fairness and transparency).
• They have a well-defined governance structure and substantial talent.

4/ What are these all-in companies trying to achieve?

• Creation of new products, new services, or new business models
• The most common objective is not just to strive for operational improvement but operational transformation.
• The least common objective is to create new (positive) customer behaviour; a good example is insurance companies who nudge their customers towards safer driving behaviour.

About Tom Davenport
Tom Davenport is a Professor of Information Technology and Management at Babson College (US). As a pioneer in process innovation, analytics, big data and AI, he is a world-renowned speaker and thought leader who has written or edited 20 books and over 250 articles in top-tier academic publications. His work is often referred to as a must-read or must-hear and he’s on a mission to provide cutting-edge insights on how companies can use analytics, big data and AI to their advantage.

Ready to learn more about the power of digital transformation? Come and join us

The Centre for Financial Leadership and Digital Transformation conducts action-oriented research involving the finance function of tomorrow and also serves as a knowledge platform. Finance leaders who want to develop a competitive and effective finance department or accounting office, embrace the technology dimension, and stay up-to-date with the most recent technological developments impacting the finance function could benefit from our unique knowledge platform and research.

Find out more