A framework for improving ROI on AI investments

By Matt Sabo

There is a desire for deeper AI integration across all industries. But token-based billing, or consumption-based pricing models, are driving up costs. Making AI investments a bigger financial risk.

Directing AI spending to efforts that will deliver the greatest business value or ROI needs to be a priority. Keep reading this AI ROI guide to learn how to best do that.

What factors determine AI ROI?

Demonstrating AI ROI can be complicated. Your goals may include achieving substantial cost savings in person-hours or realizing new productivity gains by implementing an AI tool. AI can certainly help achieve these, but the actual value of an AI investment is often more nuanced.

Here are some specific areas where an AI investment strategy can create ROI for your business:

• Quality: In areas like coding, AI can help you produce higher-quality work. This is often one of the fastest returns, showing up before a potential uptick in productivity.
• Consistency: AI can also help make your overall outputs more consistent, avoiding dips in quality or pace. 
• Productivity: Rather than looking at AI as a tool to reduce person-hours, think of it as a tool to get more production out of those hours. Be aware that meaningful productivity gains take time to achieve and may vary by role.

Identify specific problems and opportunities

Never start planning your AI investments by thinking about AI. The result will often be aimlessly spending money with little to show for it.

Instead, identify specific problems that need solutions or opportunities to capitalize on — and only then start looking for AI tools. This keeps you focused on concrete business value.

Consider AI costs and secondary benefits

The cost of a particular AI tool should always be a factor. Cost is a growing issue: To help recoup their investment costs, companies like OpenAI, Anthropic and Microsoft are shifting to more expensive billing models to better align revenue and expenses.

Carefully analyze how much it will cost your business to make gains in quality, consistency and/or productivity and if you can achieve meaningful results with less usage.

However, there are additional financial benefits AI can deliver that you may not have considered. Many investors want to know that you’re integrating AI into your core business, so doing so could help attract additional capital or boost the value of your company.

And in certain cases, you can even claim federal R&D tax incentives for AI investments you make or deduct the full cost of AI expenditures upfront.

Why many AI initiatives fall short of ROI expectations

Despite AI’s potential improvements to quality, consistency, productivity and more, many AI business investments fail to deliver measurable ROI. This can be caused by:

• Lack of focus: If your AI investments are not focused on solving specific business problems or developing clear business opportunities, you are likely to waste time, energy and money.
• Ineffective change management: You will get the most out of AI if your staff knows how to use it effectively. Approach AI implementation as a process of change — one that includes mindset shifts and training before you can expect to see results.
• Highly complex and fast-evolving technology: AI is evolving at light speed. Changes are constant. As a result, most organizations simply can’t keep up with the latest developments, let alone figure out how to make good use of them.
• No advisory support: AI is complex. If your organization does not have AI advisors to guide you in implementing AI tools more effectively, you’re at a major competitive disadvantage.
• Weak governance: Without the proper guardrails in place, businesses run into problems like shadow AI, in which team members use unauthorized AI tools with unpredictable consequences.
• Poor data readiness: “Garbage in, garbage out.” The saying is true. If you don’t have clean, organized data to feed into your AI models, you won’t generate useful outputs.

A framework for maximizing AI business value

Here’s a basic framework you can use to turn your AI investments into business value or ROI:

1. Start with policy and governance

Your governance policies set the tone for how you use AI in your business. Effective governance drives results and reduces your risks. This needs to be your start point.

2. Partner with an AI advisor

Staying current with a technology that is evolving so quickly requires a lot of bandwidth. An advisor can help you understand where AI is now, where it’s going and how your business can best use it.

3. Identify specific problems, opportunities, use cases and people

Your AI investment needs to be focused on specific problems, opportunities and use cases. Also important: Identify people within your organization who are excited about AI and can serve as early-adopter champions. Let their enthusiasm spread amongst their peers.

4. Have a plan to monitor AI use and control costs

How will you monitor AI use within your organization and control costs? As more AI providers move to token-based billing, this will become even more essential to keeping costs manageable. You’ll want to ensure you have controls and risk management policies in place to both understand how your organization uses AI and stay on budget.

CEOs — How will you measure AI ROI?

How do you effectively measure your AI success or ROI? The focus is often on usage or licensing rates, but these are largely vanity metrics. Since different roles use AI in different ways, comparing usage rates is an apples-to-oranges exercise.

Instead, focus on metrics that show you whether AI is making a measurable impact on your business results. Are you experiencing changes in productivity, quality, customer satisfaction and other KPIs? Is there a meaningful difference in these areas between employees who use AI and those who don’t?

If your answer to some of these questions is yes, that’s ROI.

Where AI is delivering value today

You can use AI to create value across departments. Key areas to explore include finance, operations, HR, sales and customer experience.

Finance

AI can help your finance team accelerate or fully automate many repetitive accounting and bookkeeping tasks. For example, you can deploy AI to tackle accounts payable tasks like processing invoice images and comparing the data against what’s in your system. AI tools can also assist your team with financial modeling and reporting.

Operations

AI can automate routine operational tasks like putting leads into your CRM, data entry, coordinating meetings or generating meeting summaries and moving information from point A to point B. AI can also help you review regulations, understand if you are in compliance with oversight rules and document compliance efforts.

Human resources

AI tools can assist with the hiring process. For example, you could ask an AI tool to compare a job candidate’s resume with the posted job description and then generate a list of questions to ask during their interview. This scenario is a good example of why you need a strong AI governance policy in place, as you could violate security and privacy regulations here if your AI use doesn’t align with standards.

Sales and customer experience

AI can help sales teams research prospects or potential clients. Customer experience staff are using AI to help keep up with existing clients. For example, you can set an AI agent to search for news about a key client or prospect, or to pull insights from a database like ZoomInfo.

You can also use AI to quickly scan recorded sales or customer calls to create scripts and follow-up questions based on what has proven effective with your specific prospect or customer base.

Transitioning from pilot programs to enterprise-wide adoption

You can launch an AI pilot with relatively little effort. However, notable ROI usually won’t happen until you scale to enterprise-wide AI adoption.

This process demands a deeper level of change. Here is a four-layer framework for scaling your AI use to maturity:

• Layer one: Data. Prioritize building an organized data foundation that your AI tools can draw on.
• Layer two: Execution. Create AI agents that know how to use your data to execute simple tasks.
• Layer three: Intelligence. Create domain-level agents that serve as the equivalent of department heads that can contextualize your business strategies, goals and daily operations to your lower-level AI agents.
• Layer four: Conversation. Layer a communication agent over everything else. It should understand your business and be able to direct other AI agents and provide strategy recommendations to humans.

FAQs about AI ROI

Here are answers to commonly asked questions about AI ROI:

How long does it take to see ROI from AI?

It varies depending on your AI tools. With an AI assistant, for example, you want to see production improvement within 4-6 months. If you don’t, it’s not being implemented correctly, it’s not the right tool, or the wrong people are using it.

Bigger investments may take longer. While brand-new companies are starting with enterprise-level AI in mind, existing companies often need to shift their mindset before seeing AI ROI.

Mindset changes lead to process changes, which in turn drive outcome changes, so focus on mindset before investing heavily in enterprise AI. You won’t pay off enterprise-level investments quickly, but you should start seeing some advancements right away.

What is a good ROI for an AI investment?

Don’t fixate on a number for here. Cutting-edge companies are looking at AI expenses as a percentage of payroll. They are treating the money you’re invested into AI credits as an employee-related expense, like a uniform or a piece of equipment.

The percentage of payroll that AI should make up will vary by industry and by role. Look for benchmarks specific to your industry to get a clearer sense of what a reasonable percentage might look like for your business.

Which AI initiatives typically generate the fastest ROI?

Knowledge agents or AI assistants can deliver fast ROI. These are quick implementations and can help your team by making it easy to access organizational information, get answers to common questions and remove small barriers to productivity.

What are your next steps toward better AI business outcomes?

To improve your ROI on your AI investments, work with an AI consulting firm. A good advisor will know your specific industry, understand the latest developments in AI technology and help guide your business forward with an eye toward how AI continues to evolve.

Wipfli helps businesses use AI to solve organizational problems and drive growth. Learn more about how they can help you develop an effective AI strategy.