95% of AI pilots are failing and companies are burning $85K monthly. Here's the reality check every product leader needs before annual planning season destroys their budget and credibility.
Absolutely! Ugly prototypes are often the fastest path to clarity. For 2026, I’m focused on building AI systems that amplify human decision-making and streamline workflows, prioritizing real-world impact over flashy demos.
Starting with solutions first is a big no-no. In my own AI implementation project at the moment, the solutions were what was thrown at us first. While they offer value to know the art of the possible, we first focused on the problems out there instead.
I'd love to hear more about how you handled that situation, because it seems like executives don't listen. Very inspiring you are following the right track!
I love how you make it clear that a roadmap is a series of decisions, not a wishlist. This was a great read. Thanks for resharing this, it’s so relevant now!🩷🦩
Interesting article, thanks! One thing that confuses me slightly is that you seem to mix internal AI inititives (companies building internal AI tools to optimise productivity) and external AI initiatives (companies building external AI capabilities for the users of their product. Your article overall seems to focus on Product but the research you refer to (e.g. 95% of AI pilots fail to deliver business value) is about internal AI initiatives. This makes me curious: how many external AI initiatives in Product fail to deliver user value and in return drive business value through the product? Did you find any research on that specifically? Or do you feel the 95% can be extrapolated to external/product AI initiatives?
Brutal but necessary reminder! Successful AI roadmaps start with real problems, clean data, and smart prototyping, not flashy features.
Yes! And the hardest part is convincing stakeholders that "boring" data work is more valuable than exciting AI demos.
The teams I've seen succeed all do the same thing: they prototype first, even if it's ugly. Saves months of building the wrong thing.
Are you planning any AI initiatives for 2026? Would love to hear what you're prioritizing.
Absolutely! Ugly prototypes are often the fastest path to clarity. For 2026, I’m focused on building AI systems that amplify human decision-making and streamline workflows, prioritizing real-world impact over flashy demos.
That sounds great, Suhrab! You're on the right track, and that's what we should all follow. 👏
The demos will be always demos. 😂
Haha always! 😂
Starting with solutions first is a big no-no. In my own AI implementation project at the moment, the solutions were what was thrown at us first. While they offer value to know the art of the possible, we first focused on the problems out there instead.
Yes to this! 💯
I'd love to hear more about how you handled that situation, because it seems like executives don't listen. Very inspiring you are following the right track!
True. I assure them we will get there to start brainstorming solutions, but need to cover Discovery first.
I love how you make it clear that a roadmap is a series of decisions, not a wishlist. This was a great read. Thanks for resharing this, it’s so relevant now!🩷🦩
Thank you, Pinkie! Many people forget that roadmaps need to adapt. There are many misconceptions in the product management world.
Thank you, Pinkie! Many people forget that roadmaps need to adapt. There are many misconceptions in the product management world.
I love you for being a PM. I'd love to pick your brain🩷🦩
Good article that shakes the false hope that AI agents are bringing on the table.
"everyone wants an AI" but don't understand their systemic risks that are amounting to data breaches. There, I said it. 😁
Exactly! 💯
Just right in time for roadmapping next year.
Specially if your roadmap includes AI precisely! Thanks for supporting Antonio :)
Interesting article, thanks! One thing that confuses me slightly is that you seem to mix internal AI inititives (companies building internal AI tools to optimise productivity) and external AI initiatives (companies building external AI capabilities for the users of their product. Your article overall seems to focus on Product but the research you refer to (e.g. 95% of AI pilots fail to deliver business value) is about internal AI initiatives. This makes me curious: how many external AI initiatives in Product fail to deliver user value and in return drive business value through the product? Did you find any research on that specifically? Or do you feel the 95% can be extrapolated to external/product AI initiatives?