Large language models, currently being associated with the term AI, are certainly the most hyped technology of the current era. However, there is a question that is repeatedly being asked by many business experts recently. That question is: Will AI ever make big money for the Big-Tech?

There may be some postulations that the investments in these AI Technologies were not made to make money (at least initially) but to enhance current products and services. But the pressure is building to start generating some sort of revenue from AI-enabled products and services to start recouping the cost of research and development. But all such efforts are based on the answer to one question: Will the customer be willing to pay the cost?

The customer will happily pay for anything that they perceive will deliver value. The word “perceive” here is very important because, during a period of extreme hype, the customer may perceive that technology will deliver value, blinded by the light from the shiny thing, but it may end up not delivering the perceived value. This applies to the approach most Big-Tech companies are leveraging to make money from their AI investments. That approach is to “enhance” existing products and services leveraging AI. Will the customer happily keep paying additional money for products and services they have been using for decades if it now has additional “AI-enabled” features?

The gist is that if you can show value to the customer, specifically when it comes to existing products and services, the customer may agree to the markup. But herein lies the challenge. The customer has probably been using your products and services for many years and it has integrated into the day-to-day lives of their employees. Any enhancement through AI, unless transformational, if not innovative, will not come across as enhanced value to those used to the product for decades. Or at least enhanced value at a magnitude that justifies a markup. Many software products have come a long way in last three decades. If you compare a 1990s version with a version today, the difference will be beyond stark. However, through hundreds of upgrades and iterations, we did not realize the journey of progress.

Recouping revenue from AI through existing products and services is more challenging than launching entirely new products and services. However, clever and effective marketing may help build the perception of perceived value, allowing a markup that the customer is willing to pay. However, the marketed features need to justify the enhancements.

If you take Samsung’s example, you will find that S24 series sales has eclipsed S23 sales. One of the factors behind this has been the AI features. From form factor to a majority of other features, the S24 series is almost similar to S23 series. What differentiates these, is the Galaxy AI features. Through marketing, Samsung was able to highlight the features and based on the increased sales volume as compared to last year, we can safely assume that customers so perceived value in those AI features. Now in order to maintain the momentum of growth, those features need to deliver as per expectations.

What happens if the value does not meet expectations? You run into the Humane AI pin situation. In case you are not familiar with this product, I suggest you research it. What happened here is that while the sales upon launch were great, considering the high price point, the marketed features promised value for the price point. However, the company has seen unprecedented returns.

Mobile phones are different from business and enterprise software, and so are the expectations. While our phones are becoming an extension of our day-to-day lives, we still do not see them as computers. Hence, any technology on a phone may come across as a transformative feature compared to a computer-based technology. This translates into the fact that our expectation of enhanced value from enterprise and business software is much higher than that from a smartphone.

This perception makes it challenging for technology service providers when they try to recoup benefits. Another way to make revenue from these AI technologies is to offer them to other companies. There are obviously benefits and risks involved in this approach. The risk category involves your partner learning from the technology and then launching their own feature. This is not difficult in today’s era. In that case, you not only lose a source of revenue, but you also end up creating a new competitor. Examples of this are plenty in the hardware world, but it can actually happen more easily in the software world, no matter how tight the contracts are.

So the only robust, short-term, sustainable, and strategic way, if you already have a very good and established portfolio of software solutions, is actually to build significant enhancement through AI in your products and services and couple them with a marketing strategy that effectively establishes the value of those AI-enabled features and functions. The mid-term goal should be to leverage existing expertise to build products and services that come across as distinct and transformational.