Adopting a level-by-level approach to AI implementation can help increase success – and ultimately ROI. Explore our AI transformation team’s proven, multilevel strategy.
Many organizations are pouring money and time into exploring and developing AI solutions and rightfully wondering if they’ll see a tangible return. Taking a level-by-level approach to AI implementation can help confirm a return on investment (ROI) at each stage, help build confidence, and maintain momentum while avoiding implementation fatigue. Showcasing the ROI of AI implementation at different levels can also help create a foundation of support throughout the organization.
Our AI transformation team developed a multilevel strategy for AI implementation that is tested and proven based on work with many organizations – including our own.
Explore AI transformation services
Get quick wins and long-term, tailored AI solutions from our AI transformation team.
Level 1: Educate internal teams at every level
Successful AI integration requires support at every level of the organization, which means a foundation of education is vital. If people don’t know what they are using or how to use it, adoption will never happen. Tailoring education for different roles, departments, and levels can help bring some of the more abstract concepts of AI to life. Executives might support the AI initiative, but they might not be the ones to come up with specific use cases or implementation ideas. Frontline workers might have great ideas for use cases but need executive-level support to execute these initiatives effectively.
The key to success at this level is internal selling that showcases how AI can make lives better versus theoretical discussions. For example, an executive-level training might focus on how AI implementation can benefit the organization, ideas for implementation across departments, and potential ROI. In contrast, frontline workers need to understand how AI can make their daily lives easier and what aspects of the technology they can implement, which might call for more in-depth, hands-on training. This top-down and bottom-up approach allows for buy-in at all levels, and it offers an opportunity to get the entire organization excited about potential AI implementation projects. When we empower our people with the right tools and education, we see a surge of innovative use cases and AI applications tailored to their everyday workflows. We’ve seen many skeptics convert to AI advocates once they witness the technology’s transformative potential firsthand.
How are we doing this at Crowe?
Internally, we’ve made AI our number one priority at the board and executive level, which helps us drive a variety of AI initiatives and implement multiple AI solutions. We’ve also engaged our internal teams through education sessions, working groups, and guilds and by naming dedicated change enablement professionals whose role is to spread AI adoption and educate our people on its benefits.
Level 2: Start with out-of-the-box tools
With a solid educational foundation in place, the next step in successful AI integration is implementing out-of-the-box AI tools. Organizations likely already have access to many of these tools as part of their current software suites, but they might not have investigated how these tools can deliver immediate productivity gains with less overall risk. Some packages include the ability to create AI-powered meeting notes, transcripts, and email drafts. Implementing these tested and proven tools allows for quick wins, which can be used to build a case for more extensive AI programs and initiatives that are tailored to the organization.
How are we doing this at Crowe?
We have implemented commercially available tools to help with a variety of tasks, from email drafting to document analysis and more. Regardless of industry or role, we’ve found people often struggle with “blank page syndrome” when starting documentation from notes and other source materials. Many out-of-the-box AI tools can condense pages of notes into coherent first drafts, which allows our teams to focus on reviewing, editing, and enhancing that content rather than spending time creating it from scratch. We’ve used the success of these programs as a foundation for more extensive AI implementation programs, including the development of our own proprietary generative AI platform.
Level 3: Automate high-volume, low-analysis processes
With a foundation of education and a frame of out-of-the-box tools, the third level of AI integration involves solving organization-specific issues. AI’s ability to understand and interpret unstructured data opens up new possibilities for streamlining everyday tasks, so organizations should start identifying high-volume, repetitive processes that were previously impractical or impossible to automate using classical programming techniques.
One prime example is processing customer emails and automatically entering them into a work item system or sending back a request for quote. Traditionally, these tasks would have been too complex for rule-based systems, but AI can now handle them with high accuracy and the potential impact is significant. For instance, if a request for quote process that typically takes seven minutes can be cut in half, the resulting enterprise value increase could be quite significant for a large organization.
How are we doing this at Crowe?
Two use cases ripe for AI implementation in tax and accounting are data cleansing and quality control. Tax and accounting firms often receive similar data from clients in thousands of different formats, requiring highly skilled professionals to spend significant time on basic data work. AI can streamline this process by automating the low-value administrative tasks so staff can focus on higher-value planning and analysis.
Another promising use case is applying tax rules through data mapping. In its simplest form, this process involves taking client data, inferring its meaning, and then applying the relevant tax rules. Historically, this process relied on humans manually reading through documentation and matching rules. AI can now – highly accurately – perform this data mapping, followed by professionals who review and validate the output rather than handling the entire process manually.
Finally, AI is proving invaluable for document interrogation, such as scanning long documents like leases, contracts, and reports to elicit key information. Rather than professionals manually reviewing pages of dense material, AI can rapidly parse through and highlight critical clauses, findings, and commonalities. This approach helps derive insights far more efficiently from large document repositories.
Level 4: Reinvent the right jobs using AI
The most transformative level of AI implementation involves reinventing the right jobs in a way that allows AI agents to take over entire administrative functions, which frees up humans to focus on high-value, strategic work that requires uniquely human skills like empathy, creativity, and relationship building. For example, imagine a sales professional who, instead of spending countless hours on administrative tasks, such as entering quotes in the system or looking up inventory across multiple data sources, can now dedicate time to cultivating client relationships and exploring new market opportunities. Or tax professionals who, rather than manually documenting findings, can rely on AI to generate first drafts from meeting notes, allowing them to concentrate on strategic analysis and decision-making. This level is currently a future state for most organizations, as it requires a robust foundation built on the three previous levels to be truly successful.
How are we doing this at Crowe?
At Crowe, we’re developing AI agents capable of handling entire workflows, from reviewing contracts, manipulating data, and reviewing large volumes of information.
With the range of tools and applications continuing to grow, AI implementation is and will continue to be an ongoing process for most organizations. Using this multilevel integration strategy can help organizations implement AI more successfully and serve as a guide as the technology continues to evolve. How do we know? We’ve used it ourselves.
By being at the forefront of AI adoption, Crowe has navigated the challenges and resistance that inevitably arise during implementation. We’ve fought through the growing pains and emerged on the other side with valuable insights. We strive to avoid the costly mistakes that come from taking on too much, too soon without a solid foundation. This hard-won experience enables us to guide clients on their own AI journeys, helping them bypass many of the pitfalls and realize ROI before progressing to more transformative applications.
Our AI transformation team delivers tailored AI solutions that can help companies solve problems, optimize performance, and transform their business.
