Skip to content

How companies truly succeed with AI: from vision to value

Every­one wants to ​“do some­thing with AI” today.
Yet in prac­tice, many ini­tia­tives get stuck in exper­i­ments, stand­alone tools, or proof-of-con­cepts that nev­er actu­al­ly land with­in the organization.

Why do some com­pa­nies man­age to cre­ate last­ing val­ue with AI?
And what keeps oth­er projects stuck at the stage of ideas and demos?

Dur­ing an in-depth con­ver­sa­tion, record­ed as a pod­cast, between Koen, Daan, and Tjel from Hype­r­i­on, one con­stant kept com­ing back: suc­cess with AI has less to do with tech­nol­o­gy, and every­thing to do with strat­e­gy, own­er­ship, and real­ism.

05 November 2025

Start at the top: AI is a strategic story

The biggest mis­con­cep­tion about AI? That it’s a tech­nol­o­gy project. In real­i­ty, every suc­cess­ful ini­tia­tive does­n’t start with data or code, it starts with direc­tion. Before you decide what you want to do with AI, you first need to know why you want to do it.

“We always start top-down,” Daan explains. ​“First we deter­mine the impact of AI on the busi­ness mod­el, the strat­e­gy, and the process­es. Only after that do we look at which domains or use cas­es are relevant.”

From vision to priority

That means an orga­ni­za­tion first needs to consider:

  • How AI affects the cur­rent busi­ness mod­el.
  • Where AI can improve the cus­tomer expe­ri­ence.
  • How tech­nol­o­gy can con­tribute to the strate­gic objec­tives for the years ahead.

Once that impact is clear, focus domains fol­low, Koen likes to call them ​“top top­ics.” Think mar­ket­ing automa­tion, cus­tomer expe­ri­ence, oper­a­tions, or prod­uct devel­op­ment.
With­in those domains, con­crete use cas­es are devel­oped and assessed on:

  • Val­ue: what does this deliv­er for the cus­tomer or the organization?
  • Fea­si­bil­i­ty: how com­plex is it to build?
  • Risks: what are the legal or eth­i­cal boundaries?

A struc­tured approach turns AI from a goal in itself into a means to accel­er­ate strate­gic objec­tives.
Man­age­ment knows what gets pri­or­i­ty, teams know what they’re work­ing toward, and every­one under­stands why it matters.

When AI is looked at on that lev­el, the hype dis­ap­pears and clar­i­ty emerges.

“You first determine the impact of AI on your business model, your strategy, and your processes. Only then do you translate that into concrete domains and use cases.”

Daan Decat AI Adviseur - Cronos group

Why so many AI projects fail (and how to avoid it)

Accord­ing to Daan, bare­ly one in three AI projects suc­ceeds today. And even that fig­ure, he says, is opti­mistic. ​“If you look at real­i­ty, the fail­ure rate is clos­er to 80 or 90 per­cent. Not because the tech­nol­o­gy fails, but because there’s no busi­ness accountability.”

The pitfall: AI as an IT story

Too often, com­pa­nies still push the AI ini­tia­tive over to the CIO or the IT depart­ment. ​“Come up with an AI strat­e­gy,” they say.
But that’s exact­ly where it goes wrong.

AI is a busi­ness project.
It’s up to the busi­ness to determine:

  • What impact AI has on their domain or cus­tomer experience.
  • Where they want to apply AI.
  • What their pri­or­i­ties are.

Once that respon­si­bil­i­ty is tak­en on, val­ue can start to emerge. The role of IT and data remains cru­cial, but more as an enabler than as the driver.

From experiment to collaboration

The com­pa­nies that do suc­ceed are the ones treat­ing AI as a joint effort across three domains:

  1. Busi­ness deter­mines the direc­tion and the value.
  2. IT pro­vides the infra­struc­ture and execution.
  3. Data experts safe­guard qual­i­ty, gov­er­nance, and interpretation.

That col­lab­o­ra­tion keeps AI from becom­ing a tri­al bal­loon. It turns AI into an instru­ment that’s strate­gi­cal­ly anchored in the organization.

For SMEs: start small, think big

It’s not just large enter­pris­es that can ben­e­fit from AI. SMEs stand to gain just as much, pro­vid­ed they approach it real­is­ti­cal­ly and in phas­es. Accord­ing to Koen and Tjel, that process starts the same way as for large orga­ni­za­tions: at the top.

Strategy as the starting point, not technology

“With SMEs, we start with the same ques­tion,” Koen says. ​“Where do you want to take the com­pa­ny? What do you want to achieve with­in three or five years?” That exer­cise auto­mat­i­cal­ly sur­faces grow­ing pains.

  • A com­pa­ny look­ing to expand inter­na­tion­al­ly runs into ques­tions around lan­guage, cus­tomer ser­vice, or regulation.
  • A man­u­fac­tur­ing com­pa­ny look­ing to scale with a lim­it­ed team strug­gles with effi­cien­cy and capacity.
  • A ser­vice provider want­i­ng to respond faster to cus­tomers looks for smarter infor­ma­tion flows.

By first under­stand­ing that con­text, you nat­u­ral­ly dis­cov­er where tech­nol­o­gy makes the dif­fer­ence.

From insight to action

Next comes an inspi­ra­tion phase. We show SMEs what’s pos­si­ble with AI today, let that sink in for a while, and then come back to turn the ideas into some­thing concrete.

That iter­a­tive approach works:

  • It stays acces­si­ble.
  • It starts from strate­gic objec­tives.
  • And it deliv­ers con­crete improve­ments with­out major investment.

AI does­n’t have to be com­plex to have impact. An SME that under­stands where its bot­tle­necks lie can use tar­get­ed AI appli­ca­tions to save time, avoid errors, and grow faster.

AI only works when it’s embraced, well thought out, and human-centered

The com­mon thread run­ning through every sto­ry is clear. Suc­cess with AI does­n’t start with tools or tech­nol­o­gy, it starts with strat­e­gy and own­er­ship.

Whether you’re an enter­prise with mul­ti­ple busi­ness units or an SME tak­ing its first steps, the key lies in three principles:

  1. Start with vision. First under­stand where you want to take your organization.
  2. Ensure account­abil­i­ty. Let the busi­ness decide what mat­ters, with IT and data as partners.
  3. Make it con­crete. Trans­late ambi­tions into real­is­tic, mea­sur­able projects.

AI isn’t a mir­a­cle cure, but a pow­er­ful lever for those will­ing to steer it deliberately.

🎥 Watch the three short videos above, or check out the full 70-minute pod­cast episode with Koen, Daan, and Tjel below:

Let’s sync up!

Send us a message