Mention the word “AI,” and most leaders immediately think about how they can make their companies run better. And for good reason: AI agents can work 24/7, automate time-consuming tasks, and improve workflows.
Want proof? Look at Salesforce. During a recent week, the artificial intelligence (AI) agent on Salesforce Help handled more than 69,000 conversations and resolved 55% of them, freeing human agents to tackle more complex issues. It’s hard to argue with numbers like those.
But as important as productivity and efficiency are for companies, leaders can — and should — also dream bigger with AI. “CEOs have to decide if they’re chasing small wins or if they’re going to reimagine how the business transforms for the 100x payoff,” said Jessica Apotheker, chief marketing officer, managing director, and senior partner at Boston Consulting Group (BCG).
The most successful companies are already doing this. A 2025 McKinsey survey found that while 80% of businesses considered efficiency a top reason for using AI, those that saw the most return also used it for innovation and growth. With AI’s help, they created new products, challenged their own thinking, and innovated in ways they hadn’t thought possible before. You can do the same.
The 3 ways companies can use AI
In a report published last year, BCG outlined the three stages of AI use: Deploy, reshape, and invent.
Deploy is the stage most companies are in now, where they use AI to automate tasks and speed up work. Reshape is the stage in which businesses transform workflows to operate in new ways. And invent is where companies use AI to think about new value they could bring to market.
“Many organisations aim too low,” the report said. “They focus on incremental automation, which is useful, but leaves so much potential untapped. Often, leaders don’t even ask what new approaches are possible, in part because many lack a strong grasp of what AI can actually do.”
Use AI to envision new opportunities
Dave Birss, the London-based cofounder of the Gen AI Academy, has worked with global brands, growing companies, and startups, teaching them how to reach their potential through creativity and technology. Here are a few ways he helps companies dream bigger with AI.
Challenge your thinking
One of the best use cases for AI, Birss said, is to shatter assumptions. He asks leaders and employees to write down all the assumptions they have about a task, industry, or particular product. If, for example, people in your company assume that Project X won’t work due to pricing, governance, or processes, ask AI to argue against these assumptions. In other words, use AI to challenge everything you believe to be true. “This is an exercise that always brings out absolute gold,” Birss said.
Imagine the future
AI can help companies map out scenarios for the future, Birss said. How? Prompt generative AI to describe what, for instance, the solar panel or mobile phone industry will look like in five years. Then, ask it to suggest new products or services that will be needed. You might also want to explore a trend Apotheker said is on the horizon: “How do you start allowing customers to make transactions directly in LLMs, and what impact will that have on the retail landscape?” Once the AI generates ideas, drill down, ask questions, and refine the ideas.
Look for missed opportunities
Birss suggests leaders use AI to create a quadrant matrix that identifies untapped market needs. Prompt your AI to research a certain market — let’s say, socks — and create a short report. Next, instruct it to create a matrix that compares price to market volume. In other words, what volume of socks are being sold at what price point? Is there an opportunity you’ve overlooked? Could you corner the market on high-end cashmere socks?
Follow the example of these companies
Want concrete examples of how this works? Check out how two companies are innovating with AI.
Tommy Hilfiger, the American clothing and lifestyle brand, uses AI tools to analyse fashion trends, historical sales data, and social media trends to create designs that better align with what customers want. In the past, the company also experimented with AI in the metaverse, inviting consumers to use AI-driven tools to create clothing designs — a move that deepened customer engagement.
Waitrose, a British grocery chain, had a similar goal: It needed to stay ahead of food trends while developing in-house products on tight deadlines. The company used AI to analyse social media buzz, trending ingredients, and home cooking patterns. The result? A creamy, orange-zest-infused Basque cheesecake, which quickly became a bestseller.
These are just a few ways AI can help companies find untapped revenue sources in the market.
Create new products fueled by AI
After you’ve used AI to brainstorm ideas, you can also create products fueled by AI. That’s one of the fun things Charlie Isaacs, chief technology officer for customer connection at Salesforce, gets to do. “I help organisations think bigger by applying AI to different types of situations and different workflows than what they’re accustomed to,” he said.
Some of Isaac’s early rapid prototyping projects resulted in the four robots that roam the San Francisco Salesforce Tower, helping with security. They were created using Agentforce, Salesforce’s platform for building and deploying agents, and technology from Cobalt and NVIDIA.
Whenever a door at the Tower is left open or gets stuck, an alarm goes off — to the tune of about 200 times a day. In the past, a human had to check the door and close it. But now the robots, which are essentially AI-powered roving cameras, are dispatched to the site. They take a picture of the problem and can even push the automatic door opener to close the door.
There’s always a human in the loop, in case further intervention is needed. But the resolution time for door security issues has dropped from 3 minutes, 38 seconds to 1 minute, 32 seconds, and the robots have saved the Tower’s security team 40 hours a week of work. “That is how we’re taking AI to the next level,” said Isaacs. “And when you have a robust transformation platform to build on, wrappered by AI and a trust layer, it makes it easy to innovate rapidly.”
See how robots lend a hand in the Salesforce Tower. (video: 3:13 min.)
Salesforce’s customers are also creating new products and features fueled by AI. Williams Sonoma, for example, wanted to completely reimagine how customers interact with its brands digitally. So, it used Agentforce to create a public-facing sous chef agent named Olive, who helps customers plan menus, discover products, and use recipes.
If a customer needs help planning a holiday dinner, Olive can suggest a detailed menu with recipes and a shopping list. The agent can also recommend the right kitchen tools and table decor, including those that a customer purchased from Williams Sonoma in the past.
Shift mindsets with AI’s help
You may feel bold and ready to innovate with AI. But in the current economy, your board and shareholders want to see ROI, and they want it now. As a World Economic Forum article noted, most executives adopt AI because labor costs are high, efficiency targets are relentless, and shareholders are impatient.
The question for leaders is whether they want to deploy AI for short-term savings or long-term gain. In the best of all worlds, it’s both.
If you need to get people on board for longer-term goals, Isaacs suggests using AI to help shift mindsets within your company and customers. “You could ask AI to create talking points that can help you reason with a customer and say, ‘Look, if you wait just a little bit longer, this could be your outcome,’ and then you can have AI show you what that outcome will look like in the future,” he said.
Apotheker said when you’re innovating with AI, the ROI may be less immediate. “You’re entering new territories, and this is something you do because there’s a real competitive advantage, but it can also be defensive,” she said. “That is, ‘If I don’t do it, somebody else will, and then where will my business be in a couple of years?’”
Ready to innovate with AI?
Our free playbook is your guide to becoming an agentic enterprise.
Get comfortable with (a little) risk
If there’s one constant about innovating with AI, it’s that it involves risk. And many companies are risk-averse right now. “But when you eliminate risk, you cannot innovate because innovation isn’t possible without risk,” said Birss.
A simple way to manage your risk is to build an AI portfolio, which BCG recommends. Apotheker suggests you include a mix of projects from each of the three stages of AI use, and expect some to fail along the way. “You want to prep your organisation for the fact that this is a very iterative process and the first versions will feel scrappy and imperfect. Will it be the futuristic AI you’re dreaming of and will it look absolutely perfect from day one?” she said. “No, but you still need to have that managerial conviction and North Star that iteration is part of the journey.”
Rely on your most valuable tool: Human judgment
As you use AI to innovate — and to generate more ideas in five minutes than you could in a year — remember your most important tool isn’t the tech. It’s your ability to distinguish a good idea from a bad one. Dave Birss calls human judgment the “most valuable of career skills” in the age of AI.
Good judgment, he said, leads to better decisions, better outcomes, and better opportunities. When a team exercises good judgment, it can shape the fortunes of a company.
“Use AI to challenge your thinking rather than replace it,” Birss said. “That is going to be one of the things that creates an ‘us and them’ divide within the workplace and society. The people who use AI to challenge themselves and push themselves into areas they wouldn’t normally consider are going to grow at an exponential rate.”
Why innovating with AI is a smart next move
Efficiency and productivity have become table stakes with AI. Those are big payoffs. But to get the most from the tech, you need to also think bigger. And who knows? With the help of AI, your company might just create the world’s Next Big Thing.
