No ROI in AI yet? Try these six proven tactics for creating real business value

Follow ZDNET: Add us as a preferred source<!–> on Google. 

ZDNET’s key takeaways

• There’s growing concern that AI is a bubble about to burst.
• Smart digital leaders take a tactical approach to use cases.
• They get executive buy-in by focusing on strategic priorities.

With MIT recently reporting that just 5% of AI projects deliver value, there’s a growing fear that the bubble surrounding generative and agentic AI technologies is about to burst.

However, Fausto Fleites, vice president of data intelligence at gardening specialist Scotts Miracle-Gro, is one business leader aiming to assuage such concerns.

Also: A minority of businesses have won big with AI. What are they doing right?

Fleites has built a successful career using machine learning and AI to turn information into insight – and now those experiences are powering innovative deployments in a long-established business.

Formerly in senior digital leadership positions with Sears and Accenture, Fleites started working with Scotts in February 2023. He saw the opportunity to apply his expertise and help the 150-year-old company embrace emerging technology.

He began by building the IT foundations, with technology from AWS and Google, to apply deep-learning models to enterprise data, creating decision-making insights for the executive team.

Also: I unleashed Copilot on my Microsoft and Google accounts – here’s what happened

Over the past year, Fleites has been developing use cases for the exploitation of generative and agentic AI. He told ZDNET how his team’s explorations have engendered interest in AI across the business. “The good thing is that we’ve been very tactical on the use cases and wins, and this success has helped us in the conversations with the other groups that we are targeting.”

Six ways to create value from AI

Fleites’ AI strategy at Scotts is divided into two areas: consumer-facing efforts such as improving search and chatbots, and back-office elements, which center on reimagining internal processes, including rewriting customer service emails. We’ll discuss these in detail later.

–>

From his development of AI-powered solutions to business challenges, Fleites offers six best-practice lessons for other digital leaders:

1. Iterate fast and learn from your mistakes. “One example is that we did our prototype for product recommendations extremely quickly using ChatGPT or Gemini, and we learned the limitations for our business. We pivoted extremely quickly. That culture of rapid iteration, learning from mistakes, and failing fast is key to success.”
2. Go small and listen to the experts. “I would suggest avoiding big bang projects that are multi-year and don’t have any clear business value, because AI is a rapidly evolving technology. Anyone who tells you, ‘I have 20 years of experience doing gen AI,’ they’re lying to you. You must adapt to how the technology is evolving.”
3. Create a strategy and focus on KPIs. “So many companies are just implementing AI because of what they read online. They’re just deploying AI for the sake of having AI, and they’re failing at it. Start by aligning your AI use cases to a business strategy with clear and measurable KPIs.”
4. Work across the business and let people join in. “We’ve formed cross-functional teams that act like startups. The idea is that you can tackle short-term use cases. A cultural shift needs to happen to make your organization more agile, collaborative, and data-driven.”
5. Hone your platforms and perfect your approach. “Invest in foundational technologies to make your data more accurate and accessible. For AI chat, we use the product knowledge that we have as a 150-year company to power our conversations. But we’ve had to format that knowledge in different ways to be more efficient with AI.”
6. Focus on change management and prove the benefits. “We started these discussions and began developing use cases almost a year and a half ago. We still go to the company with all the potential use cases. But early wins in consumer services have allowed us to replicate that success story in other areas.”

Boosting customer services

Fleites’ AI-enabled consumer-facing efforts have produced significant results in two areas: search results and chatbot services. “These two key features are core to our strategy,” he said.

AI search runs via a RAG application in Google Vertex AI, where customers can use natural language to search the company’s catalog to receive answers to their questions.

“Before we introduced this technology, customers had to use the exact term, such as ‘fertilizer’, to get results,” he said. “Now, consumers can express any questions in their own language, and they get results.”

Also: AI use is up, but organizations still aren’t seeing gains, Atlassian study finds

The company is also using AI to improve the quality of customer conversations in its web-based chat agent.

<!–>

–>