I was reading a report by MIT [1] the other day and noticed an interesting remark made by one of their team of researchers: 95% of organizations adopting Generative AI (GenAI) see no measurable financial return, and only 5% of them achieve meaningful value/outcome. [2]

Then I looked around, and I saw a bunch of clients, a lot of AI developers, and product managers falling for the AI hype “for the sake of the hype”.

It might look attractive and make your product appear to be powerful in the eyes of the public. Still, if you’re not working on the right things using the right resources, you’ll literally fall behind — AI is not gonna make you win against your competitors.

Adding AI to your product is an obvious move

A lot of companies add AI for the sake of sounding like outcompeting others; they’d hire an AI Engineer to add a simple “Ask AI” button to their app that doesn’t (always) add value to the core product, and sometimes even slows the business down [3].

When the technology is publicly and easily accessible, with the amount of hype that comes around, everyone wants to jump on board, every business wants to build on top of it, which makes the competition even harder and unavoidable if having “AI” in your product is your end goal.

That being said, no matter what a company’s end goal might be, it is worth noting that:

• Some founders decide to add AI out of a mere FOMO (fear of missing out). They don’t wanna lose, even when they don’t clearly see what the future holds (it happens a lot more than we want to admit when building something). They see the hype and try to invest in something that can bring value in the future without a clear plan.

• Others decide to add AI because they have a predefined set of ideas in mind about how their product will get powerful “once they have AI in it”. So, they write them down, add them to the product roadmap, make an early announcement to get their customer base excited, and try it.

Both groups can win if they go beyond the adoption level, where all that matters is the label, that special “Ask AI” button, so that the goal will not be to say “our product is AI-powered” but “our customers now save 50% of their time using our solution”.

AI adoption can waste your time, money, and energy

If you’re not clear about what you actually mean by “AI” beyond the means of what’s being provided in the media through the tech hype, you’re wasting your time thinking about integrating AI into your users' daily workflow.

AI will always be there in the market, but your customers won’t. Because they might find other and better alternatives to your solution, they might not be satisfied with your product anymore, so they will unsubscribe, delete their account, and move on.

That, to help us ask the right questions (and find answers), such as:

1. What do customers really need for their lives to be better? Do they want AI for the sake of AI?

2. What are their primary recurring pain points? This is even the main reason for you building a product.

3. If we integrate AI into their workflow, will that change anything? Will that help them work faster and save time? Will that bring a better and more efficient solution to their problem?

4. How much should we invest in adopting AI and for what ROI (return on investment)?

Going from Adoption to Transformation

As a software engineer, I have noticed and I’ve now come to believe that a great number of those 95% of companies did not either ask or entertain the idea of implementing the ideas coming out of those questions asked above. Because, AI adoption is one thing, and AI transformation is another. (In my opinion)

Almost everyone can adopt AI, but not everyone can bring a transformation through it, given the fact that transformation comes with:

• Providing a shortcut to their workflow. If there are a couple of steps or tasks that are repetitive, boring, and sometimes annoying to users, working on implementing AI, if possible, is the right move and factor that brings value and profit.

• Reducing the cost of running the business through AI integration in the company’s internal workflow.

• Making use of the user’s workflow metadata in order t find new pain points and provide pragmatic solutions. If AI will help you understand your customers better, with their pain points encountered in the workflow, go for it.

• Practically making sure that AI will never slow the customers down as they are seeking the solution to their business or personal problems.

If it’s not going to transform anything in the workflow, why would you want to add AI just for the fun of it?

Notes

[1] MIT’s State of AI in Business: State of AI in Business

[2] Some interesting points: demandlab’s article

[3] "“Ask AI” is not a feature, it’s a UX pattern", from AI, by Kenn Kibadi