A different view by Jaime Porta

For the past few years, the AI conversation has been dominated by capability. Which model is more powerful? Which platform has the best features? Which company has access to the latest technology?

Those questions still matter. But they are becoming less decisive. Most companies can now access similar models, copilots and automation platforms. A competitor can subscribe to the same technology within days. The real advantage no longer comes from simply having AI. It comes from being able to turn it into better work.

That makes the next phase of the AI race an execution race.

Access is becoming easier. Transformation is not.

Launching an AI pilot has never been easier. A small team can connect a model to internal documents, create a sales assistant or automate part of a reporting process in a matter of weeks.

The first demonstration often looks impressive.

Then the difficult questions begin:

• Is the underlying data reliable?
• Can the solution access the right systems—and nothing else?
• Who owns the process after the pilot?
• How does it fit into the way employees already work?
• What happens when the output is wrong?
• Can it operate across countries, business units and regulatory environments?
• Which business KPI should improve if the solution succeeds?

These are not primarily model questions. They are operating-model questions.

This is where many AI initiatives slow down. The technology works, but the organization around it is not ready. Data is fragmented. Processes vary by market. Responsibilities are unclear. The AI tool sits outside the core workflow, creating another screen for employees to open rather than removing work from their day.

The pilot proves that something is technically possible.

It does not yet prove that the company can execute it at scale.

In sales, value lives inside the workflow

Sales offers some of the clearest AI opportunities: account preparation, lead prioritization, proposal generation, next-best-action recommendations, forecasting, product selection and coaching.

But even a strong model cannot compensate for a disconnected commercial landscape.

Imagine an AI assistant asked to recommend the best product configuration for a customer. To provide a useful answer, it may need customer history from the CRM, product rules from a configurator, pricing and availability from the ERP, product content from the PIM and contractual information from another source.

If those systems use conflicting data, if the commercial rules are undocumented or if ownership is unclear, the assistant will not solve the problem.

It will expose it.

This is why AI adoption cannot be separated from platform integration and process design. The most valuable sales use cases are rarely standalone features. They connect several parts of the commercial system and support a decision within an existing workflow. The objective is not to give salespeople more information. It is to help them make a better decision, prepare faster or remove an administrative step altogether.

That distinction matters.

A chatbot that answers questions may be useful. An integrated assistant that understands the customer, applies product and pricing rules, prepares the next action and records the outcome in the CRM changes how work gets done.

AI scales at the speed of organizational alignment

International transformation adds another layer of complexity.

A process that works in one market may depend on different roles, systems, data definitions or commercial practices in another. Local teams often have legitimate reasons for doing things differently. At the same time, allowing every market to design its own AI approach creates fragmentation, duplicated cost and inconsistent governance.

The answer is not to force absolute standardization. It is to be precise about what must be common and what can remain local. Across international transformation programs, I have repeatedly seen that results depend less on the initial tool decision than on the ability to align processes, data, ownership and adoption across teams.

In one omnichannel transformation spanning 17 countries, measurable progress did not come from launching another isolated channel. It came from connecting the customer journey, clarifying how leads moved through the organization and establishing a more consistent commercial process. Lead conversion increased from 20% to 50%, while digital channels grew from a marginal contribution to a meaningful part of sales.

AI follows the same logic.

The model may accelerate individual activities, but enterprise value appears only when the surrounding process also changes.

Leaders need to move beyond the pilot portfolio

Many leadership teams currently track AI activity: the number of ideas collected, pilots launched, licenses assigned or employees trained. These indicators show movement, but not necessarily value.

An organization can be very busy with AI and still change almost nothing. The leadership conversation should therefore move from “How many AI initiatives do we have?” to “Which business capabilities are we improving, and what must change to make that improvement operational?”

That requires a different set of decisions:

1. Start with a business constraint. Identify a decision, delay, cost or customer problem worth solving. Do not begin with a tool looking for somewhere to be used.
2. Define the measurable outcome. Faster quotation, higher conversion, better forecast accuracy, shorter onboarding or lower service cost gives the initiative a reason to exist.
3. Map the full workflow. Understand where data originates, which systems are involved, who makes the decision today and where human review remains necessary.
4. Assign business ownership. AI cannot remain an experiment owned only by innovation or IT. Someone must own the process, the KPI and the operational result.
5. Design for adoption from the beginning. If employees must leave their main workflow, duplicate information or distrust the output, adoption will stall regardless of technical quality.
6. Build for scale, not only for demonstration. Security, access rights, architecture, governance, monitoring and local-market variation should be considered before the pilot becomes politically difficult to stop.

This does not mean every use case needs a large transformation program. It means leaders should understand the path from prototype to operations before celebrating the prototype.

The strongest AI strategy may look like transformation management

As AI becomes embedded in CRM, ERP, eCommerce, product-information platforms and digital sales tools, the boundary between AI strategy and business transformation will continue to disappear.

Companies will still need technical specialists. But they will also need leaders who can connect technology with commercial priorities, redesign processes, align stakeholders, make architecture choices and create the governance required to move across functions and countries.

That combination is harder to copy than access to a model.

The winners of the next phase will not necessarily be the companies that ran the first pilot or bought the most licenses. They will be the ones that can repeatedly turn promising use cases into reliable capabilities—and those capabilities into measurable business results.

The AI race is no longer only about what the technology can do.

It is about what the organization can execute.