Remember when ROI was like the ultimate business measure? You put numbers in, check the results, multiply by 100, and boom – you have your numbers ready to go. This process worked back then, even when computers first came around and we just wanted to be more efficient. But now with AI, especially generative AI, that way of looking at things isn’t enough. AI doesn’t just cut costs. It changes how companies think and act. If you’re trying to measure big changes with a tool that’s only for cutting costs, you’ll always miss something.
Why the old ROI doesn’t work now
A study by IBM’s Institute for Business Value shows most broad AI projects deliver a paltry 5.9% ROI. It’s not that AI is bad, but people are measuring it the wrong way. It’s like judging a cricket player only by how many runs he scores. Sure, the numbers matter, but you miss how he makes the team better and how they play under pressure.
The traditional ROI model misses the good things AI brings: new ideas, faster decisions, stronger supply chains, and happier customers who stay longer because they feel valued.
The real good stuff is complicated
A PwC survey from 2024 points out that 49% of companies face hurdles in unlocking AI’s value due to problems like excessive data and tricky integrations. Another study we were a part of, on the Future of Enterprise Intelligence, showed that most data leaders struggle to turn data into decisions. The problem isn’t the tech — it’s knowing what to measure and why.
Here’s the truth: if you think AI is only about saving money right away, you’ll probably be let down. The real value is in what happens later: more loyal customers, happier teams, and a company that changes faster than others.
Real things you can measure now
Of course, money still matters — but we need to look at it differently.
Efficiency and savings: AI can get rid of repetitive or boring work to save cost and time while increasing productivity. For instance, in areas like supply chains, AI tools have helped cut inventory costs by almost 15-30%.
Higher revenue: Measure increased revenue and new revenue streams driven by AI initiatives such as tailored marketing. For example, a global conglomerate that shares marketing content across regions and languages can use AI to lower ad costs, shorten turnaround times, and attract new customers.
The other important stuff
The biggest changes are harder to see on paper. These things don’t show up in a quick ROI calculation. But if you don’t pay attention to them, you’ll miss the whole point of AI:
Closer to customers: AI has made customers happier by offering quicker assistance. You can also drive personalised customer service through AI chatbots that can handle more service requests.
Happier teams: After AI takes over the boring tasks, attrition rates can be reduced because employees will be allowed to do better work in tandem with AI tools.
Newer ideas: AI is now helping come up with many new product ideas at some companies and helping them make decisions faster with better data analytics capabilities.
Think about cricket. A bowler isn’t just measured by how many wickets he takes. You also see how much pressure he puts on the opposition, how many balls he bowls without runs, and how many chances he makes for others.
AI’s ROI is like that. Spreadsheets might show savings, but the real value is how they help the whole company. Leaders who only look at the wickets miss the winning plays.
What leaders should do now
So, how should companies rethink ROI with AI? Here are some ideas:
Connect what you measure to your goals, not just the tools. Start with the business problem — faster sales, strong supply chains, loyal customers — and then decide what to measure. Don’t just get excited about dashboards.
Use both numbers and stories. Track savings, but also look at things like employee happiness, customer satisfaction, and how fast you make decisions. They can show you what’s coming financially.
Start small, grow smart. The best companies start with small tests, then grow based on what works. It’s like trying out a young player before putting him in a big game.
Think about problems. AI can bring problems like ethics and bias. Constantly monitoring and adjusting your AI usage policies is important to make customers and regulators trust you.
What’s next?
Measuring AI’s ROI isn’t about getting rid of the old ways. It’s about adding to them. The companies that succeed will measure not just how efficient they are now, but how strong they’ll be later.
In cricket, we don’t just judge a captain by his primary skill (batting or bowling) , it’s about how he understands the game, gets the team going, and handles the crunch situations when it matters. AI’s ROI is about leadership: how you use tech to not just save money but to build a company that does well even when things are uncertain.
The future is for those who stop counting returns in small amounts and start making them in ways that others can’t.