Beyond the AI pilot

For all the excitement surrounding artificial intelligence (AI), getting an AI tool to work is often only the beginning. The greater challenge is turning it into something useful, repeatable and scalable.

For Justine Therese R. Malabanan, successful AI adoption begins not with technology but with identifying the business problem it needs to solve.

With experience spanning brand management at Procter & Gamble, consulting at Boston Consulting Group, AI strategy at a venture-backed startup and strategy and operations at Google, Malabanan has developed a framework for integrating AI into business operations.

“When the data speaks for itself, the decision is undeniable,” Malabanan says.

However, organizations do not always have complete data. In such cases, she recommends using related information to identify useful patterns or generating new data through pilot projects.

While rapid experimentation allows companies to test ideas without waiting for perfection, Malabanan stresses that successful adoption requires three elements: integration, quick wins and training.

Start with the problem

Integration begins with understanding whether an organization has the infrastructure and processes needed to support AI.

“If a team wants to transition to using AI but all of their existing data are in manual handwritten records, then AI won’t be able to support them until they digitize their documents,” she explains.

Beyond preparing data, companies must identify applications that align with their strategic priorities and deliver measurable results.

Malabanan cautions against overly ambitious projects that are difficult to implement, as well as simple applications that offer little business value.

The ideal starting point is a project that is both achievable and meaningful.

Build momentum through quick wins

“AI is a pivotal shift in the ways of working of companies today,” Malabanan says. “It needs momentum to progress.”

Early successes demonstrate the technology’s value, encourage investment and help build internal support.

Employees who experience measurable improvements can also become advocates for wider adoption.

However, even promising applications can fail without adequate training.

“Employees need to know when to use AI and how to work with it as an augmentation instead of a replacement,” she says.

Training should extend beyond writing prompts to interpreting AI-generated findings, checking accuracy and understanding the technology’s limitations.

Malabanan recalls an organization that developed an AI solution tailored to its operations but struggled with adoption because employees were unwilling to invest time in learning the system.

Expecting immediate results, the team eventually lost confidence in the technology.

She stresses that adjusting to new processes requires effort, but the initial investment can produce substantial long-term benefits.

Leadership drives adoption

Scaling AI also requires clear organizational responsibilities.

Malabanan identifies three essential roles: an executive stakeholder who sets the direction, functional leaders who translate that vision into operational changes, and a program manager who coordinates implementation.

Successful adoption requires both leadership commitment and support from employees using the technology daily.

Before approving an AI pilot, she believes executives should ask one critical question:

“What is the timeline for setting it up before we start to see the benefits of the tool?”

The question forces organizations to evaluate feasibility, reliability and expected returns before committing resources.

Ultimately, Malabanan maintains that AI adoption is not simply about purchasing technology. It is about establishing the processes, skills and organizational support needed to create lasting value.

The starting point is not what AI can do, but what the business needs it to accomplish.

The views expressed reflect Malabanan’s independent professional perspective and do not represent those of her current or former employers