Projects are launched that make perfect sense on paper, but we are not sure who will use them or whether there is a sufficient volume of users to make the project profitable.
I see many cases where the system we have set up is so complex that it would take less time to do it without automation. We create them just because the options are available, not usa phone numbers list
because they make much sense. As I was saying to my colleagues at Gen AI Circle (it's a community of experts and professionals who want to be experts in AI), they are “ tool oriented, not problem/need oriented ”.
And on the other hand, there are companies that want to implement these systems more because they say they apply AI or because they are looking for a shortcut that does not exist, than because they are actually going to be used for real.
Can you really imagine a company that sets up and actually uses a 100% AI and automated system to generate content? A blog, a freelancer, etc. yes, but a company?
Assuming that a blog is launched, do you think that content will provide real value to its audience or will that blog/newsletter become something that generates large volumes of content that few or no one reads because it is of low quality? So have we made progress or harmed the brand? What good has the work and investment been done?
I am of the opinion that innovations should be used only in those areas where they really add value, only in those areas that improve the quality of work and make a difference.
In my opinion, we should experiment, because it is the way to learn and develop, but only implement what is really useful. If we generate high expectations, and what I mentioned in the previous paragraph happens, what favor are we doing to the development of AI?
We have moved from the initial excitement of the hype with generative AI to a phase in which the difficulties in integrating and scaling these solutions become evident. In a previous article I gave you the updated data of the Hype Cycle of artificial intelligence published by Gartner in 2024, where we saw that generative AI had already passed the peak of the initial craze.
I am going to use a McKinsey article as a basis to talk about how to move forward in this scenario, wasting as little time as possible and taking advantage of the real value contribution of AI in our companies.
1. AI Projects: Innovate with a Realistic Objective
A common mistake I see in many agencies and SMEs is launching AI projects without a clear purpose. The idea of experimenting with new technologies can be tempting, but it is essential to ask yourself: does this project really bring tangible value to the business? Or are we simply trying to follow a trend or because we are looking for a short-term benefit? (which we rarely achieve in the end).
McKinsey highlights that one of the biggest problems is the proliferation of AI pilots that never scale (according to its study, only 11% of the companies surveyed have implemented a scalable AI-based system ).
Not only does this disperse resources, but it also frustrates attempts to see a meaningful return on investment. Instead of spreading themselves thin across multiple experiments, companies should focus on those projects that can actually have a clear and measurable impact on their business.
In the image below you can see a simple matrix that will help us select the most appropriate projects to avoid these problems (in reality it is a matrix that attracts the effort of a lifetime).