AI is a new employee. Onboard it like one.

When I talk to leaders about AI right now, a common challenge comes up: how to turn rapidly changing capabilities into meaningful business results. According to our survey of 550 product and technology leaders, 91% of executives say they’re feeling more pressure to prove the ROI on their AI initiatives. That pressure makes it harder to balance exploring new tools with choosing investments that will deliver lasting value.

In those conversations, I keep coming back to the same starting point: what leaders and their teams already know about their businesses, customers, and markets. AI opens up new possibilities. That knowledge helps determine which ones are worth pursuing.

Business expertise gives AI strategy direction

New AI capabilities and competitors’ investments raise practical questions: What’s worth testing? Where should we invest? What can we afford to wait on?

Answering those questions takes both an understanding of the technology and a clear view of the business. Where are customers running into problems? Which constraints are holding teams back? What would a better outcome look like?

I’ve found that the most useful AI conversations start there. AI can help uncover gaps, test assumptions, and process information in new ways. The knowledge teams bring gives that work context—and helps connect an interesting capability to a business result.

Don’t let the tool tell you what your business needs

We recently worked with a company exploring an AI-powered chat experience for its customers. The team had used AI tools to map out the experience and possible customer scenarios. The results looked promising.

Then we started walking through those scenarios using what the company already knew about its customers. The gaps became obvious. The proposed experience didn't reflect the order in which customers actually make purchasing decisions.

AI hadn't been able to uncover something the business was missing. The business knew something the AI was missing.

This problem can show up in plenty of ways.

• An AI-generated product roadmap might prioritize features that sound compelling but don't address the problems customers pay you to solve.
• An AI sales assistant might recommend an action without understanding the signals experienced salespeople use to distinguish a promising opportunity from a dead end.
• An AI workflow might optimize one step of a process while creating additional work somewhere else.

Across these examples, the missing piece is business context.

Your company knows more than what’s stored in its data

Some of your company's most valuable knowledge may never have been formally documented. It lives in the experience of people who can look at a customer, opportunity, or problem and instinctively know what matters.

We often describe that as a “gut feeling,” but there is usually a decision process and decades of experience underlying the feeling. Part of our work at Modus Create is helping organizations uncover that reasoning and turn it into context AI can actually use.

A useful comparison is onboarding a new employee. Access to the tools is only part of the picture; understanding the business takes context. For AI, that means making relevant goals, priorities, processes, and institutional knowledge available—and keeping that context current.

To provide that context, you first need to understand what you're trying to solve. Ask yourself two questions:

1. What's the problem?
2. How do you know it's a problem?

I’ve found that those specific questions often uncover something different from the issue a client initially identified. What looks like a technology problem may actually be a symptom of something happening elsewhere in the business. Asking how you know there’s a problem helps separate the symptom from the underlying cause before you start building a solution.

The first step isn’t choosing an AI tool. It’s identifying a problem worth solving.

Build a strategy that learns

Of course, knowing your business inside and out doesn't mean every assumption you make will be correct. The goal is to create a faster way to find out what works and adjust when it doesn't.

That's changing how we think about product development. Instead of treating the backlog as a list of requirements we assume are right, we can build around what we need to learn. As new tools reduce the cost and effort of development, teams have more opportunities to test ideas, gather feedback, and change direction based on what customers actually do.

Deployment is the beginning of finding out whether something delivers value. Before we build, we define the outcome we expect and how we’ll measure it. Once it’s in production, we look at whether people use it, whether it helps them accomplish what they need to, and whether those results support our assumptions. What we learn informs the next decision.

Our research shows that while 58% of decision-makers consider ROI or revenue impact a top metric for AI-powered products, only 40% say the same about user adoption. If people aren't actually using what you built, the ROI conversation may be premature.

From individual productivity to company productivity

Much of the early value companies have found in AI has been at the individual level: helping someone research, write, analyze, code, or complete another task faster. Those gains matter, but they represent only part of the opportunity.

The bigger opportunity comes when those gains extend beyond individual tasks and begin connecting knowledge and decisions across the organization. Customer feedback can inform product priorities. Institutional knowledge can guide AI-assisted decisions across teams. What the company learns from one initiative can inform where it invests next.

That's the shift from individual productivity to company productivity: using AI to improve how the organization makes decisions, not simply how quickly individuals complete tasks.

Start with what you know

Nobody has AI completely figured out. The technology is changing too quickly to know exactly what the next tool or capability will bring.

The tools will keep changing. A clear understanding of the business gives teams a basis for deciding what to try, what success looks like, and when to change direction.

The opportunity is to bring that knowledge into the work, test it against what customers actually do, and use what we learn to make better decisions. That’s how AI becomes useful beyond any single task or tool.

See what's blocking AI returns and map out a clear plan forward. Learn more.