From Vision to Strategy: Crafting Your AI Business Case for Success

What makes AI different than other business cases

In a typical business case, the focus is on traditional technologies or solutions based on straightforward metrics. A business case for AI goes beyond traditional considerations, encompassing the unique aspects of AI. This includes:

• Data dependencies
• Learning capabilities
• Ethical considerations
• Potential for transformative impact on processes

Through this process, you’ll discover many AI point solutions that offer quick wins for increased efficiency. These are meant for a narrowly defined task, and they tend to be simpler to define and implement. But there’s much bigger value to be had with AI.

The power of AI is how it enables genuine transformation of workflows, processes and organizational structure. These can fuel many long-term benefits, such as improved employee satisfaction and retention, skills acquisition, brand enhancement and a higher valuation of the company.

But transformation isn’t automatic. It requires a more strategic view of what’s coming next for AI and for your business. This is where you start laying the groundwork and it’s in your business case.

Four steps to construct a business case for AI

1. Strategy: Build consensus and alignment among stakeholders.
2. Prioritization: Determine which use cases make sense now and which to save for later.
3. Impact: Use metrics to show how changes will impact people, processes and workflows.
4. Value: Quantify the business value and tell a compelling story.

Push back on unreasonable expectations

Most of your colleagues will have an opinion about AI and what to do, often around generative AI and large language models. These technologies get a lot of attention, but they often set expectations for more value with less effort. This lack of understanding about what’s required to achieve value when implementing AI gets in the way of making meaningful progress.

For example, a financial services business wanted to use AI to make use of all the knowledge contained in their procedures, even though much of it was outdated. They figured AI could quickly search through procedures to surface relevant information for agent interactions with customers. While AI can process a lot of information, it doesn’t know what to activate or what’s “not right.” The bank first needed to select an appropriate AI model, conduct training and perform data cleanup to make it worthwhile.

It’s important to understand what tasks AI can and can’t do, and that understanding needs to be communicated in your business case. AI won’t handle all tasks, but it enables you to keep humans in the loop for tasks that are better performed by people. It also gives you the flexibility to adapt the roles of AI and humans, as needed.

1. Strategy begins with consensus

• Know your primary stakeholders’ needs.
• Build links across organizational silos.
• Anticipate objections.

Find your starting point

Knowing where to begin requires strategic alignment among stakeholders who likely have different priorities, perspectives and levels of technical expertise.

Consider the priorities of your primary stakeholders:

• BusinessThis group wants benefits and value across customer experience, employee experience and operational efficiency, while minimizing disruption to the teams. So, they want technology that’s easy to adopt and brings immediate value.
• TechnicalThey’re concerned with your existing assets, the total cost of ownership (TCO) for AI, minimizing the cost of any change, vendor relationships, best-of-breed solutions, reducing operational risk and minimizing disruption to IT teams.
• FinanceThey value operational efficiencies that reduce OPEX and which the business teams are signing off to achieve. Like the technical stakeholders, they look at existing assets, TCO, ROI and commercial relationships with vendors.

Anticipate objections

Consider conducting a “pre-mortem,” especially if you anticipate objections based on previous initiatives. By working backward from potential points of failure, you’ll better understand perceived risks and how to prevent them. Get your team of stakeholders together and present a simple scenario:

“Imagine that it’s one year from now and our AI implementation goes wrong in some way. For example, we don’t get expected efficiency gains, customers are unhappy or employees are confused. What do you think happened?”

You’ll likely hear a range of responses about why it happened: too busy; vendor solutions weren’t as promised; lack of executive focus; inadequate staffing or expertise and so on.

It’s a useful activity because it’s hypothetical and participants feel safe speaking up and surfacing concerns. Together, you can create a plan that addresses concerns proactively rather than reacting later. You’re also creating a feeling within the team that you care about their input.

Perhaps most importantly, you’ll avoid repeating problems that are known issues in your organization — as you actively bridge organizational silos.

2. Prioritize by use case

• Go deep on understanding use cases.
• Build confidence with quick projects.
• Use more data sources and a centralized knowledge base.
• Plan for long-term transformation.

Choose your top use cases based on potential results

Businesses often assume that adopting AI immediately will make operations more efficient. But every conversation about adopting AI needs to start with an understanding of the use case you’re addressing and the potential results — including how it contributes to differentiating your brand. This process is an opportunity to build working relationships with other stakeholders, and to visualize use case application across your organization.

Once you understand objectives of the individual stakeholders and tie them into the broader objectives of the business, you’ll need to decide on the best options for AI. These conversations help everyone embrace the changes coming with AI by asking them questions rather than making too many assumptions. It also feeds the conversations with leadership.

Well-defined use cases are essential for leadership review. You’ll need to:

• Demonstrate successful use cases: These can be from within your organization or from other businesses where similar implementations were successful.
• Show rather than tell: Think of visuals and anecdotes. These will be much more memorable than overly long presentations.
• Reinforce your position with data: Anecdotes and opinions go only so far.

Lay the groundwork for value realization

Once you prioritize use cases and how they connect to your strategies, next steps help you understand how to get where you want to be:

• Assess the availability and capacity of inhouse resources and skill sets.
• Assess the availability, accessibility and quality of your data.
• Evaluate partners and whether to engage with a Professional Services team.
• With the best use cases identified, decide where AI can help now, where it will have the most impact and what steps you need to take now to ensure a successful implementation.

Agree on quick projects first

Organizing your projects so that you start with quick wins builds confidence for those working with the new technology. It also enables you to prove its value without much investment. Quick wins often start with improving productivity. They’re task-specific and they can be integrated into other processes.

There’s another important advantage: Employees find satisfaction in seeing progress, and this approach plays an important role by not delaying gratification for positive change. There will be more complex projects in your longterm plan, but you’ll gain buy-in from everyone — leadership, stakeholders and users — when you can show measurable results that happen quickly.

Centralized knowledge: The foundation of success

Centralized knowledge is key to leveraging the massive power of AI for use cases where AI provides answers directly to customers or assists agents by serving up relevant information from the knowledge base during interactions. It helps foster efficient decision-making and problem solving, ensuring that the correct information is available at the right time for the right people.

For example, agents might have a favorite knowledge base among several they can choose from to answer customer questions. Maybe this favorite seems more up-to-date, or it has some good information on more technical topics from customer queries. But your chatbots use one specially developed for their needs, and its information covers the most common questions only. It’s also regularly updated.

By identifying your knowledge data and accessing it from a centralized database, you can take full advantage of data you already own. From this, you can implement any number of AI applications from multiple points of view — business, technical, finance — and take your company strategy into consideration as knowledge is shared.

Deployed AI apps will all depend on the same knowledge that’s integrated and continuously updated. As your AI maturity progresses and your expertise grows, so will the value of your centralized knowledge to support it.

Without a centralized knowledge base, your customers aren’t assured of consistent and accurate information. And agents waste time searching for it.

Speech and text analytics: From quick wins to true transformation across your organization

Get value from every tier of implementation

AI at its transformational best

AI transformation enables a new way to support the customer and employee experience, streamline journeys and improve business outcomes through deeper insights. These outcomes go beyond the use of a centralized knowledge base alone. Many AI use cases rely on a wide range of data sources. Bringing this data together opens the door to powerful capabilities, including experience orchestration.

Experience orchestration gives you a way to coordinate technology based on the experience you want your customers to have across all channels of engagement. It analyzes large volumes of many types of data — customer, operational and interactions — to identify patterns and intents. This gives you visibility into where your customers are experiencing issues, such as being unable to transfer money, pay a bill or schedule an appointment. You can use this for proactive decision-making, taking action when it will be most impactful and appropriate.

With AI-powered experience orchestration, you can create value propositions that address customer needs in the moment — going beyond traditional personalization methods.

Transformation is about behavioral change

One you’ve decided which use cases make sense, watch out for a common misstep: underestimating or not thinking through the complexities of process and cultural changes for employees. In fact, the importance of the human factor is frequently overlooked.

Stakeholders must get comfortable as owners of the AI solution as it relates to the use cases, and users need to build new skills and expertise as their jobs evolve. Comfort with tools builds trust and confidence in your ability to lead the way. Manage trust like an asset that you can increase.

There’s also an element of empathy to consider for employees. Change isn’t always welcome — no matter how much it will improve business processes or overall success. That’s because humans are motivated by emotion, not rational arguments. Employees like feeling that what they’re working on is progressing in a coherent way. Plan for it in your business case. One way to address it up front is to break down updates into smaller increments and celebrate that progress internally.

3. Impact: How to measure success and what to quantify

• Extend your use of metrics.
• Get comfortable with variables.
• Look at the big picture.
• Use subjective feedback.

Understand metrics from multiple perspectives

Integration

AI needs to be integrated into business processes to produce measurable results. And those processes need to be adaptable to all the new opportunities enabled by AI. The data points required to substantiate the investment to support your adoption of AI are usually dictated by the use cases you choose.

Calculations

You can choose almost any target goals if you understand the value of the metric and how it will be used to continually improve outcomes. Calculating the value of AI requires looking at results from multiple perspectives.

For example, time saved with automated processes is an estimate, and finding the baseline between human and AI performance can be messy in pre-production. That’s because of the difficulty of calculating ROI based on a single point in time.

Automation

There will be cases where the automation is achieved with the use cases that are highly relevant, with a context and scope of impact that are predictable. In other cases, automation is possible but will carry considerable variability, will require fine-tuning, and value will be realized only after a few months of refinement. Since both scenarios are possible, it becomes crucial to understand and properly express what values come about by design and which ones by experience. It will definitely impact the credibility of the business case.

Timelines

Timelines are complex and vary by use case and goals. For a three-year plan, when will a specific AI tool show that it has reduced costs? These are important metrics. Unfortunately, the performance of machine learning-based AI models can change over time due to any number of variables. But there are ways to find your answers.

Validity

When your AI solutions are built on models offering transparency and explainability, you can see why the tool is making certain decisions and impacting expected outcomes. This means you can trace the validity of these decisions, including errors, biases and anomalies. By evaluating the model and the decision, you can identify strengths, weaknesses and opportunities.

This insight enables you to continually measure AI performance and optimize it in effective ways. Just remember to add budget in your business case for maintenance resources to continually manage the long-term benefits of AI.

Look at the big picture and the individual projects

The use of AI will likely cascade into many areas of your business. Those might be secondary benefits or indicate movement toward a bigger transformation. For example, AIpowered predictive engagement improves customer satisfaction by proactively resolving issues, which reduces agent costs. But it can also drive revenue by engaging certain customers and triggering an action that improves sales conversion. Capture all the value you anticipate; it will strengthen your business case.

Not all business KPIs need to be financial. In fact, focusing on financial metrics alone can cause you to miss critical investments in projects that have longer-term and strategic impacts. Think “flexibility.” It’s difficult to quantify, although it includes responsiveness to users, adaptability and building trust — all of which contribute to a seamless user experience.

These benefits may be more difficult to quantify in financial terms over task- or process-specific benefits.

Then there are indirect metrics that influence customer success, agent retention and growth, such as specific outcomes for agent alignment with the company strategy. This could be measured and benchmarked using post-implementation surveys that spot gaps in training or engagement.

Subjective feedback reveals trends

Don’t overlook subjective data as a form of measurement. As you plan for post-implementation feedback, it should be less about numbers and more of an internal discussion of the AI implementation. You might consider a survey with responses ranging on a scale from one to ten. Topics could be about reducing the number of emails or meetings; whether a process has simplified work or improved it; and the perceived value of AI. Keep it simple and easy to ensure you get the feedback you want; don’t try to cover too many changes or attributes.

Start with a default expectation and over time, multiple surveys will reveal trends in behavioral changes as your AI initiative matures.

You can capture a lot of important metrics, but without a compelling narrative, you’ll miss a major opportunity: Tell the bigger story of how AI supports long-term growth. We call that “value.”

4. Value

• Prioritize value over metrics.
• Look for future value potential.
• Tell a compelling story.
• Continually optimize.

Why value matters more than metrics

Companies often have a hard time understanding the quantification of value for AI. It’s difficult to measure, in part because of the many ways different groups use it, the results they’re measured on and the secondary benefits of AI that cross your entire organization. Think of value as focused on the big picture as you develop your AI business case.

Let’s say an IT team is evaluating speech and text analytics tools for a quick win. They find one that claims 98% transcript accuracy. Sounds good! Another one offers 95% accuracy. With a similar up-front cost, the IT decision-makers consider the highest accuracy as the clear winner. However, that extra 3% they’re getting comes with some costly downsides. The 98% option would require an extra step to feed data to an external AI system, while the one with 95% accuracy automatically integrates data. Giving up that extra 3% means less burden on IT, as it won’t need resources to monitor and manage that integration feed.

By looking at all processes and components together, you’re able to capture true cost. This approach also requires precise language about timelines.

Businesses typically have strategic goals around reducing costs. Some even have measurable targets, like “reduce our costs by 25 percent in three years.” But this still misses total value. You might have identified and prioritized certain use cases that will reduce costs. But how much of the total cost will be reduced and when, and by which changes? In addition to understanding all the components of your initiatives, you must include specific timelines in your calculations.

These are components of determining value, so that resources are applied where they’ll have the most impact.

Build a compelling story

The narrative you’re going to share with decision-makers should illustrate how AI will support business outcomes and support stakeholders’ key priorities. It’s more than identifying the key performance indicators and it’s more than direct financial benefits.

Help your audience visualize transformation, how your business will do things differently and how the AI you’re proposing supports long-term growth.

Looking at value in your examination of use cases is key to securing the resources and support needed. It lays the groundwork for the full utilization of your software investment.

Conclusion

Developing a business case for AI is a journey of discovery that requires a mindset of continual optimization. The more you and your teams work with AI technology, the more innovative uses you’ll find for it. And laying the groundwork for long-term success in a business case is the first milestone.

There’s no single roadmap for AI because there are many ways to harness its transformative power. Start with the results you and other stakeholders want to achieve and tell that story. Make sure that you fully understand — and can substantiate — the metrics that support the story of your business case.

As a champion for AI, you’ll be challenged to visualize innovation — and validate results — in a way you might not be prepared for. Genesys Professional Services can help you with practical guidance on your business case for AI so that stakeholders and leadership feel confident of the results you’re committing to.

About Genesys

Genesys empowers more than 7,500 organizations in over 100 countries to improve loyalty and business outcomes by creating the best experiences for customers and employees. Through Genesys Cloud, the #1 AI-powered experience orchestration platform, Genesys delivers the future of CX to organizations of all sizes so they can provide empathetic, personalized experience at scale. As the trusted, all-in-one platform born in the cloud, Genesys Cloud accelerates growth for organizations by enabling them to differentiate with the right customer experience at the right time, while driving stronger workforce engagement, efficiency and operational improvements.

Visit us at genesys.com or call us at +1.888.436.3797.

Genesys and the Genesys logo are registered trademarks of Genesys.
All other company names and logos may be trademarks or registered trademarks of their respective holders.
© 2024 Genesys. All rights reserved.