Published · Reviewed by Horacio Poblete, CEO and founder
Before asking for proposals, write in one sentence what leadership needs to be able to do when the program ends. Something like: "every director prepares the monthly results meeting with AI and reviews the analysis they get from their team with judgment". That sentence becomes the yardstick. A proposal that doesn’t explain how it gets there, using your company’s documents and decisions, falls behind, however good the curriculum looks.
With the sentence in hand, invite a few providers and compare their answers to the same questions. Corporate presentations help little at this point: everyone talks about transformation, and almost no one shows what your board will do differently.
What to ask for in the proposal
Copy these questions into the request for proposal. The answers say more than the provider’s portfolio.
- What material of ours will you work with? Which documents, decisions and meetings do you ask for before starting?
- What changes in the program for the board and for managers?
- What will each participant be able to do on their own at the end, and how will that be checked?
- How does the program handle governance: confidential data, usage rules, human review?
- Which tools do you use, and what happens if the company already has a different one under contract?
- What stays with the company when the program ends: method, criteria, prompt library, agents built?
- How will results be measured, and what do we need to record before starting?
- Who leads the sessions, and what does that person build with AI today?
- Is the program live and designed for us, or an open class with some adjustments?
- What don’t you do?
The last question is the most revealing. A serious provider can say where its work ends. Ask Almai the same questions too. For the seventh, the guide on how to measure the results of AI training shows what to record beforehand and what to look at afterwards.
How to tell whether AI training is practical or only theoretical
A practical program ends each session with something from your work done a different way. The board agenda prepared with research across several sources, a vendor’s proposal analyzed with judgment, the difficult email to the team rewritten and discussed. If by the end of the session no one has produced anything they’ll use that week, it was a class about AI, not with AI.
Signs that it’s practical:
- Asking for material beforehand: the provider wants real reports, presentations and decisions to prepare the sessions.
- Exercises by role: the CFO works on analysis and budget, the sales director on proposals and sales forecasts.
- Hands-on: most of the session is people using AI, not watching someone else use it.
- Critical review: the program teaches you to be skeptical of the result, not just to generate text.
Signs that it’s theoretical: the history of AI, a long glossary, trend slides, other companies’ cases instead of yours and a certificate as the main deliverable.
Business school, consultancy or specialized firm
Each type of provider tends to get different things right and to stumble on others. It’s worth looking at what each one usually does well before looking at the name.
| Type of provider | Usually good at | Where it tends to stumble |
|---|---|---|
| Business schools | Strategic vision, management repertoire and exchange among executives from different companies. | In the open class: the content has to serve many companies at once and rarely touches your budget, your report or your Monday meeting. |
| Consultancies | Diagnosis, process design and implementation projects, often with good industry knowledge. | Training becomes an accessory to the project and the knowledge stays on the consultancy’s side. When the contract ends, the company has the solution, but doesn’t know how to maintain it. |
| Firms specialized in AI | Being up to date, with people who build with AI every day. | Quality varies a lot: some are tool resellers with a course wrapped around them, others speak the language of technology rather than that of the people who decide. Check whether the program starts from the company’s decisions or from the product. |
| Course platforms | Scale and anytime access, which helps level vocabulary across operations. | Recorded content rarely changes an executive’s routine, and it doesn’t separate what each level of leadership needs. |
Between a business school and a firm specialized in AI, the choice depends on the goal. If you want to broaden an executive’s strategic repertoire, the school does that well. If you want the board working with AI and deciding priorities on the company’s own material, look for someone who works with that material and builds alongside you. The question on how to choose AI training sums up these criteria in a few lines.
AI consulting or leadership training
In most cases, leadership training comes first. A board that already uses AI in its own work knows what to ask a consultancy for, evaluates the proposal better and doesn’t outsource the decision about priorities. When there’s an urgent, well-defined project, the two can run together, as long as someone on the board owns the decision. The full answer is in consulting or training: which comes first.
Warning signs in a proposal
- ROI as a percentage: a return promised as a number before the provider knows the company is a well-formatted guess.
- Focus on one tool: the program revolves around a product, and what’s learned expires with the next version.
- The same content for everyone: board, managers and operations in the same room, with the same exercises.
- Demos instead of your material: ready-made examples that always work, because they were chosen to work.
- Vendor dependence: agents or workflows only the provider knows how to maintain, with no method transferred.
- Silence on governance: not a word about confidential data, usage rules or review.
Dependence is where Almai’s view is clearest. We build agents during the programs, together with leadership, as a way to transfer method and technology, never as a separate product. The criterion applies to any provider: in the end, the company needs to know how to maintain, adapt and evolve what was created.
When executive AI training is worth hiring
It’s worth it when the board doesn’t yet use AI in its own work and, even so, already approves AI budgets, tools and projects. It’s also worth it when there are scattered pilots with no owner, when the team already uses AI and no one reviews the quality of what comes in, or when leadership disagrees about what to do first.
It isn’t worth it when the board plans to send representatives instead of taking part, when what’s wanted is to roll out a tool that’s already been chosen, or when no one will own the next steps. In those cases, the money buys a good day of conversation and little change.
If the board doesn’t work with AI, the company is only playing at transformation.
