How to successfully integrate AI into your business strategy: Lessons from experience
These days, there hardly seems to be a conference or a business conversation that doesn't involve AI, Generative AI (GenAI) or Large Language Models (LLMs). That’s true in 2024 and doubtless for many years to come. As AI becomes part of the business strategy for more companies, spending on AI is projected to reach nearly $300 billion by 2027. Generative AI will account for 35% of that total — as compared to just 8% in 2023.
What’s the current state of AI strategy?
The field of AI is so broad and fast moving that it can be hard to decide where to begin. Even looking at the range of different AI techniques, the choices can be daunting. For example:
- Machine Learning: A method of training algorithms to learn from and make predictions based on data. Techniques include supervised learning, unsupervised learning and reinforcement learning.
- Deep Learning: A subset of machine learning that uses neural networks with many layers to analyze various factors of data.
- Natural Language Processing: Enables interactions between computers and humans using natural language instead of code. Example use cases include language translation, sentiment analysis and chatbots.
- Computer Vision: A field of AI that enables computers to interpret and make decisions based on visual data.
- Generative AI: Enables the creation of new content, such as text, images, or music, based on the data the AI was trained on.
These categories often overlap, and many AI systems use a combination of these techniques to achieve their goals.
Is it time to make a big move on AI?
As an executive or business leader, there can be a sense of pressure to try out new technology, or even make it part of your technology estate. Investors are increasingly expecting businesses to signal their intent and deliver on AI, with some companies using it as evidence of the seriousness of their plans.
For example, Klarna recently committed to reducing costs by leveraging AI. Microsoft is constantly talking about massive interest in their Copilot offerings, and there have been many proofs of concept (PoCs) in that area.
What could possibly go wrong?
Unfortunately, despite all the hype and sizzle, nearly half of all AI projects don't move beyond the proof-of-concept stage.
You could argue that’s the point of a PoC: To succeed or fail fast. On the other hand, that’s a big investment of money and resources to not yield an outcome, and there’s a risk that expectations get badly out of line with reality.
In the current cycle of AI adoption, too many PoCs are mere explorations — driven by a desire to get involved with a hot new technology without a clear business strategy for AI.
Despite all the hype, nearly half of all AI projects don't move beyond a proof-of-concept. Surprisingly, the best business strategy might be to make AI a bit more boring.
The path to success: Building a business strategy for AI
The technology is very important — but success is also about getting some fundamentals right.
- Understand the business case — or acknowledge that you’re experimenting — but don’t mix the two.
Many PoCs move forward without a clearly understood business case. Whether intentional or not, they end up being almost entirely experimental. This is not inherently a bad thing. It’s OK to explore new tech to build corporate knowledge and experience, but be honest. If that’s what you’re doing, acknowledge it and don’t confuse it with expectations of tangible immediate returns. - Focus on business outcomes.
If you’re not experimenting, be clear about what business outcome you're trying to drive and — like business change and life generally — be very clear about your vision. - Embed AI within the broader company ecosystem.
Think about the interactions and integration with other platforms and business processes needed for production and full implementation. Invest in data quality and AI security (the security of the AI model itself, plus the data used for model training and inference). You must also ensure observability of the AI stack. - Make AI part of your broader business transformation.
AI doesn’t exist in a vacuum — and it’s not the only factor needed to ensure success. You need to clearly define what kind of training is required, and any broader business changes needed to make AI work. - Keep the Board and Executive Committee involved.
Is your AI in line with your ethical standards or business values? How do you keep that on track? Creativity and entrepreneurship are important, but so is good governance. If your business has privacy or other regulatory constraints, consider a Sovereign AI approach. Your Board will thank you.
Putting AI to work for business
It may seem surprising, but the best AI business strategy should include becoming a bit more boring! By that I mean, getting the fundamentals right. In the next few years, businesses will move away from experimental AI PoCs and focus more on practical use cases that deliver tangible value. Customer service will be a primary area. Microsoft Copilot is a great example: helping generate emails and responses that drive an overall improvement in quality and raise average performers up to best-in-class.
At that point, AI will become truly powerful but also be one of many tools available for organizations to drive productivity and transform. If you are looking for more insights into how to leverage AI more effectively in your business strategy, reach out to me, or read more expert opinions on the subject.
Posted on: October 25, 2024
