How Businesses Can Get Real ROI from AI
“Almost all companies invest in AI, but just 1% believe they are at maturity.” That’s according to a recent report from global research firm McKinsey, and it highlights a widespread challenge. While the AI hype cycle continues and billions of dollars are poured into the field, businesses are increasingly questioning the actual return on investment (ROI) of their AI ventures.
The core issue isn’t necessarily with the AI systems themselves, but rather with how companies approach implementation. Investing in AI without a clearly defined strategy can lead to disappointment. Businesses need to move beyond viewing AI as a magical solution and instead see it as a tool to be strategically integrated.
With the rapid evolution of the AI industry, it can be difficult to keep up. Generative AI was at the forefront not long ago, and now small language models and agentic AI are dominating the conversation. Instead of trying to chase every new trend, how can businesses take control and succeed with AI?
Industry experts offer their insights on how to achieve a positive ROI:
Start With a Clear Purpose
Many businesses make the mistake of implementing AI first and then attempting to shoehorn it into their existing operations. Peter Ferrari, CEO of Jaca, warns against this approach. “Begin with an in-depth analysis of the core needs of the business and its vision, and then determine how AI can be integrated,” he advises. “I’ve observed numerous companies get overly enthusiastic about AI without considering how it benefits their business. That’s how you squander both money and time.”
Jaca, which offers the unique sugar alternative Allulose, uses AI to analyze customer behavior. Their objective is to predict when customers will make their next purchase and identify the most effective messaging. According to Ferrari, this strategy has decreased customer acquisition expenses by 40% and increased customer lifetime value by 25% for Jaca.
Ferrari stresses that this success was driven by a data-focused approach, enabling Jaca to use AI effectively in advertising its product at scale. The company initially analyzed its business data to pinpoint the most effective areas for AI application, and only then did they began experimenting with AI in their ads. “We would produce a few ad variations and then wait for weeks to determine which one performed the best. Currently, AI enables us to operate fifty iterations at once, modifying the text, images and call-to-action in real time using performance metrics.”
Define Your North Star
Businesses often adopt AI simply to keep pace with competitors, without establishing their unique value proposition. This is a mistake, according to Andy Thurai, VP and principal analyst at Constellation Research.
In a previous article co-authored with Joe McKendrick on Forbes, Thurai emphasized that “the secret sauce to AI success is selecting the right business use case — a robust and expansive business use case.” Thurai also suggested in a LinkedIn post that defining your ‘North Star’ involves “finding your unique differentiators and clearly explaining why customers should care, as well as focusing on how you solve a business problem instead of comparing functions or features with competitors.”
Make Your AI Tools Work Together
Successfully integrating AI across various teams and platforms is, in Ferrari’s words, “a completely different ball game.” “AI integration is like managing a music group,” he explains. “You might have four incredible musicians, but if each is playing a different tune, it will sound awful.”
Jaca utilizes a suite of tools, including Claude for content creation, Midjourney for visuals, Meta’s AI for enhancing ads, and Klaviyo for managing emails. Ferrari emphasizes that if these tools aren’t connected, their full potential isn’t being realized. To overcome this, Jaca uses Zapier, an automation platform that allows these AI tools to communicate effectively rather than operate in isolation.
Prioritize Ethics and Regulations
Many experts predict that AI ethics and regulations will become increasingly crucial, considering the potential harm AI systems pose to society.
One such concern is the spread of AI-powered disinformation. For example, a tragic triple murder in Southport, England, was followed by the rapid spread of misinformation, including AI-generated content, falsely claiming the crime was committed by an immigrant. This misinformation sparked anti-immigrant sentiments, leading to violent protests. Troy McGuire, cofounder of Fintech.TV, stated in an interview that when using AI, “it’s important to remember that AI is not infallible and ensure it’s being used responsibly — especially in the media space.”
Even though AI can save time and help with fact-checking, McGuire believes it’s necessary to incorporate human oversight. That’s why McGuire believes that if you’re in the media business “you need to evaluate your content and cross-reference AI-generated content when necessary,” adding that journalists need to get the facts correct and be transparent, even when they use AI.
Ferrari also argues that a privacy-first approach is essential for any business that wants to use AI. According to him, transparency, regular audits, and adherence to regulations such as GDPR and CCPA are key to ensuring the responsible use of customer data. “Our customers need to know how their data is being used, and they need to trust that we’re handling it responsibly. That’s why we disclose every AI-driven interaction.”
Measure ROI
The most effective way to ascertain that you’re receiving value from AI is to measure its ROI. However, many companies lack a standardized approach to measuring AI’s impact.
By defining specific business use cases, aligning AI initiatives with business strategy, and prioritizing AI ethics, businesses can increase the probability of success. As Ferrari points out, “balancing AI tools against ROI is crucial as these expenditures can quickly become overwhelming and suffocate the business — not simply financially, but also functionally. The companies that succeed in AI aren’t merely those that implement it, but those who do so more effectively than others.”
