The Harsh Reality: AI Investments are nowhere near the CEO Agenda
Written by co-founder Ossi Syd.
I don’t have data on this, but based on empirical experience, I would stake my reputation on the claim that 90–95% of AI investments are nowhere near the required level of systematic approach – or anywhere near the CEO’s and Executive Team’s agenda.
Understanding the context of the claim
In recent months, one of the main themes in the global AI debate has been the ROI of AI initiatives. In other words, the question of whether the enormous sums invested globally in AI have produced real value.
At a macro level, some have drawn analogies to the widespread adoption of computers in the 1980s: back then, ROI only began to materialize with a delay of several years relative to the investments. Perhaps the spread of the internet in the 1990s followed a similar dynamic – learning and searching for direction went on for surprisingly long before the new way of operating started to deliver significant results. On the other hand, despite the massive bubble and its eventual bursting, the majority held a strong belief that the general direction was right and that there was no going back.
AI isn’t one-dimensional
“Bubble!” has been shouted this time as well. The AI hype is thick, and there are more snake-oil salesmen around than ever. Still, the analysis remains largely thin and idealistic; one rarely encounters a deeper scrutiny of the logic behind the alleged bubble. Furthermore, the prevailing attitude toward AI is often one-dimensional, focusing narrowly on large language models and their immediate challenges rather than the broader paradigm shift.
Based on my own experience with AI implementations in companies (S&P 500 included) since the late 2010s, I am not worried about a bubble. Regardless of where exactly we are on the development curve, we are dealing with information technology that is more autonomous, more capable of learning, and more expressive than classical deterministic IT. From this, new kinds of value can be obtained – and in many cases already have been obtained in practice.
“Experimenting with AI or testing something with it isn't an investment in itself.”
Why is AI ROI so difficult to make visible?
If value exists – or at least is about to emerge – why is making it visible so difficult in the case of AI?
Back to basics. We should not allow the abstract and multifaceted nature of AI (or hype) to confuse us. First, let’s focus on the semantics of investment. By definition, an investment is “the allocation of money or resources with the expectation of generating future profit or income.”
Let's be honest: Not everything we throw money at deserves to be called an investment. An investment is systematic activity and has a certain structure. It has objectives derived from strategy, specific resources allocated to it, and it is monitored and managed. It is on the executive team’s agenda, and specific individuals are accountable for it. Specifically, experimenting with or testing something isn't an investment in itself, though it can certainly be a valuable starting point for one or a part of the process.
You probably already know the ropes
For example, if a paper mill company identifies strategic potential it wants to pursue, it invests in capturing that market by, for example, building new capacity – a new mill. Or it invests in improving an existing production plant so that its capacity and/or quality increases.
The investment is planned carefully, and its follow-up is continuously on the top management agenda – perhaps even supported and mandated by the board. The investment is managed systematically and is continuously compared against strategic goals and their metrics.
The investment is also obviously multifaceted: it is not just a new paper machine with better technical properties. New capabilities and capacity may be sold to customers in advance. New personnel are recruited and trained. New management systems are built. IT systems are developed. Processes are renewed to match next-generation technical capabilities. Logistics are optimized.
You likely aren't in the paper industry, and perhaps you aren't planning to invest in a new mill either. Nevertheless, the same fundamentals apply: the investment is constantly on the management's agenda, it has objectives, it is continuously monitored & steered, it has a systematic structure, and it involves many different dimensions, beyond technology.
We decided to do something about it
Sure, it’s easy to claim that it’s simple and we (you) should just get it done. But:
The culture and frameworks for executing investments in traditional contexts haven't appeared out of thin air. They are the result of long-term evolution. If suitable structures and frameworks don't exist, they must be built.
So, let's stop complaining, pick up our shovels, and get to work. Renessai has been collaborating with the top experts from the academic world, and together we have developed a model for our clients to build the necessary framework for measuring AI's ROI.
Expect to hear much more from us on this topic in the near future.
