Who's afraid of AI failures? (Spoiler: those who still sell licenses "by the pound")

Market analysts and researchers are just doing their job: they analyze current conditions and highlight empirical trends.

Over recent years, various reports have highlighted an undeniable reality: a high failure risk in AI pilot projects and a return on investment (ROI) frequently falling short of initial expectations.

Fair enough. Markets mature through trial, error, and course correction.

What continues to deeply astonish me, however, is the attitude of many professionals operating within the IT and Artificial Intelligence sectors.

The perfect alibi: when it's always someone else's fault

What strikes me is seeing industry professionals maximize and emphasize these negative outcomes in their posts and public statements.

The question begs to be asked: should the very people populating this market be the ones fueling the narrative of failure?

When an AI pilot project fails or misses its targets, the fault is rarely just the client's.

Moreover, in our professionals' posts, we regularly read a shared mantra: "The technology isn't to blame" and "You must fix organizational processes before adopting AI."

At that point, I must confess I am left speechless.

A logical and commercial short circuit

If the technology works and the client isn't to blame, whose fault is it?

The truth is that an AI adoption project is not about "selling software"; it is about delivering a tailored solution that starts with rigorous organizational analysis, evaluates operational impacts, redesigns workflows, and introduces technology only downstream of that transformation.

Otherwise, you simply don't understand how complex projects are managed -specifically those involving artificial intelligence.

Let me add a dose of healthy realism: haven't most experts in our market continued to sell AI exactly the way they sold traditional software over the last thirty years—that is, via licenses, or "by the pound"?

The pricing models for LLM usage and the licenses for CRM/ERP AI add-ons flooding today’s market were not rolled out by some abstract entity; they were sold by the very professionals who now pontificate about others' failures without taking the slightest responsibility.

Software vendors vs. independent solutions

Some industry professionalsun thinkingly praise major software vendors (sellers of ERP, CRM, PLM platforms) for guaranteeing superior results compared to those offering customized solutions outside platform ecosystems.

Yet, the most insightful market surveys tell a completely different story: multi-million-dollar economic returns have been achieved by clients who partnered with entities decoupled from license vendors.

Anyone whose business model relies on selling licenses and who is now scrambling to slap AI add-ons on top is simply trying to perpetuate a volume-driven sales process, not genuine technological and business empowerment.

Conclusions

Today more than ever, before embarking on an AI journey, organizations must exercise extreme caution.

Buying technology "by the pound" will never solve structural problems.

True professional excellence requires a collective soul-searching and a clear choice: partnering with serious professionals who do not limit themselves to self-referential posts, but who tie their own economic results to the measurable success of the solution.

I am certain of this, and I am ready to prove it.

If you wish to discuss this further, let's talk.