Lori Schafer in Forbes Tech Council: Beyond Sales And Marketing: The Real ROI In AI

Read the full article in Forbes Technology Council here.

Lori Schafer is CEO of Digital Wave Technology, an AI/GAI-native platform for master data with RAD for unified enterprise solutions.

Years into testing and adopting AI for business, companies have settled into a rhythm with the technology. Marketing teams save time by having AI create ad copy. IT developers build advanced, natural language chatbots for the website. Sales representatives eliminate repetitive tasks like data entry. Companies are putting money into generative AI and now agentic AI, and the common assumption is that AI’s greatest impact lies in front-office initiatives like advertising or customer engagement.

AI ROI In The Operational Core Of A Business

For a retailer, a chatbot’s ability to recommend a swimsuit for a consumer is a great innovation, but the ability to have AI forecast the total demand of that swimsuit, manage the purchase order (PO) process, automate allocation to stores across the country and recommend optimized pricing and promotions is necessary business. Consider healthcare. GenAI can help a pharmaceutical company write a new product label, whereas an AI-powered, centralized data management system can support regulatory compliance efforts and help optimize complex drug pricing to increase profits.

These are two examples of many across industries, but they point out where most enterprise companies will unlock ROI from AI. In fact, a new report out of MIT found that of the 300 GenAI use cases they studied, 95% are delivering zero return. The report said 5% of integrated AI pilots extracted millions in value, while the rest demonstrated no measurable P&L impact.

To get the best ROI out of their AI initiatives, enterprise companies should prioritize operational use cases. In the meantime, as AI continues to innovate, such as ChatGPT launching a buy button, companies should continue to experiment with customer-facing implementations that best fit their business goals.

Four Steps Toward Extracting AI Value

The backbone of a commerce engine is having accurate, unified and trusted master data flowing throughout the organization. Without it, AI applied to pricing, forecasting or replenishment will produce fragmented or unreliable outcomes. However, along with managing data, enterprise businesses must take early steps in preparing their company to get the most from AI. Important tasks to accomplish even before implementing AI include:

1. Identifying Proper Key Performance Indicators (KPIs)

Before a company brings an AI solution into its systems, the organization needs to ensure the technology delivers against highly specific business goals. This means implementing AI with process-specific customizations that enable companies to evaluate the AI based on how it meets targeted and personalized business outcomes rather than how it meets more generic software-related benchmarks.

2. Reducing Costs Where It Matters

ROI and cost savings don’t come from a company merely reducing headcount and replacing people with AI. Companies need to reduce business process outsourcing (BPO) spending and specific needs for agency partners, particularly in back-office operations, where value gets optimized even further once AI is implemented.

3. Finding Partners Who Meet Your Business Needs

Again, AI needs to live and breathe a company’s business goals. Very few companies are moving beyond experimentation, in part because of the complexity of building a custom AI solution that integrates smoothly into workflows. The MIT study found internal AI builds fail twice as often, so companies can benefit from finding AI partners that live within their market.

4. Becoming Flexible And Extensible

At the same time, when seeking technical partners or AI solutions, enterprises should look for AI systems that are both interoperable and extensible. Interoperability offers seamless integration between existing technologies. Extensibility provides scale and the ability to adapt to new technologies as new business needs emerge.

Of course, as companies venture into AI and GenAI, there are also common pitfalls to avoid, such as failing to enact a solid change management plan, not getting executive buy-in and unwillingness to test and learn from AI. With proper steps in place and an AI-first mindset, companies can be on the right side of AI ROI.

Agentic AI Narrowing The AI Divide

No doubt, AI is hard. Chatbots are popular but deliver limited enterprise value compared to deeper operational AI use cases. Companies are also quick to add GenAI into support, content creation and analytics use cases; however, few industries are demonstrating deeper structural shifts to drive AI toward measurable change, according to the MIT report.

Agentic AI, a framework that widens GenAI and develops a system of AI agents handling autonomous decision making, could move the needle among enterprise organizations exploring the operational benefits of AI. The technology enables AI to perceive business goals, take action on those goals, collaborate with other agents smoothly, learn from the AI interactions and autonomously orchestrate complex workflows.

As companies struggle to find ROI from AI, agentic AI embedded into operational business goals could help narrow the divide between those that fail and those that succeed. First, though, companies need to ensure basic steps are in place to support clean, governed data and AI. Otherwise, AI results can be unreliable, inaccurate and most of all, costly.