Forget the Hype: Harness AI to Generate Real Value for Your Business
The possibilities of AI and machine learning appear to be endless. It’s easy to feel overwhelmed and have a hard time knowing where to start. So, how can your company use AI technology to realize real-time business benefits – without setting the wrong priorities, overshooting your target, or wasting precious time?
To answer these questions, we turned to Errol Koolmeister – an AI and technology expert with a proven track record who has helped companies like Nordea, Vodafone, and H&M add value with AI.
What’s your take on the AI evolution, which has evolved in parallel with your career?
“I’ve been fortunate to be able to experiment quite freely with data models, AI, and machine learning for various companies during my career, thanks in large part to the massive hype in this field. It was incredibly fun and educational, but we seldom wound up with any concrete business benefits. It’s only in recent years that we’ve seen major gains with AI, as companies have started to put the models into production and act on their results. I like to say that while no model is perfect, some of them are helpful. You never know which ones will end up paying off before you put them to the test in real life.”
What are the most common pitfalls when a business incorporates AI into their activities?
“A frequent mistake is prioritizing innovation over operations. By definition, innovation is a loss-making endeavor – at least until the day you finally get things right. You have to start somewhere, dare to try things out, and keep optimizing along the way. You also need to keep in mind that the most important issue isn’t AI technology in itself. It’s how a company manages change. At the end of the day, it all comes down to your corporate culture. That’s why you have to focus on your people, rather than the technology.”
How do you determine which AI projects have the potential to pay off?
“First and foremost, I’d like to point out that there’s no such thing as ‘AI use cases,’ only ‘business use cases.’ Start with what you want to accomplish. That can be anything from shortening customer service response times to coming up with new business models to increase your revenues. Prioritize your projects based on three key assessment variables: value, feasibility, and reusability. That will best equip you to sift through your options and find projects that can garner the support of internal stakeholders with the resources you need. It also increases the chances of a quick ROI and that you’ll learn lessons that can be applied to other parts of the company.
Errol Koolmeister’s 3 Keys to AI Success
1. Improve your operational processes
“You’ll see much quicker returns on your AI investment if you start with what you’ve got. By simplifying and streamlining your existing operational processes, for example, in customer service, you can benefit from the low-hanging fruit in projects where AI delivers real value in just a few months. It might sound a bit boring, but when you do that, you can bank those earnings and reinvest the funds in an ongoing learning process.”
2. Stockpile skills for the future
“When you work with your operational processes, you successively stockpile acquired knowledge and skills for the future. You can channel those resources into a central department that creates routines and processes that you subsequently scale up. Or you can initially get an external partner to help you with the innovation phase. It’s important not to focus too much on pure tech. Tons of companies throw away money on huge platform projects, only to scrap them a few years later. Assume that platforms will be switched out every five years and aim to build up your skillset through iterative processes. That way you’ll deliver value throughout the lifetime of your project, not just at the end.”
3. Find ways to cut costs
“Just like my first tip, this involves finding ways to improve existing processes. When you scale them up, even seemingly minor optimizations can make a big difference for a company’s overall profitability.
Hallon is among the companies that have used machine learning to optimize their processes. Read more about how they learned to predict which actions would be right for which late-paying customer. For example, they discovered that 8 out of 10 debt collection cases are unnecessary and risk increasing churn.