Successfully leveraging AI in business and professional services firms

AI innovations are creating seismic structural changes for many companies

Oct 22, 2023

Key takeaways

AI is growing rapidly, and its adoption is expected to increase moving forward.

These applications can drive several process efficiencies, but associated risks may emerge.

AI will disrupt functions—such as billing models—within professional service firms.

Artificial intelligence Business transformation
Generative AI Business services Management consulting Professional services

Artificial intelligence (AI) is changing how companies in all industries conduct business, creating opportunities for enhanced insight and efficiency and lowering costs for consumers. But what can AI mean for professional services organizations and how can it influence and strengthen day-to-day operations?

A rapidly expanding market

AI has seen explosive growth recently as use of the technology has expanded. For example, according to Future Market Insights, the generative AI market is expected to be worth $10.9 billion in 2023. The market is expected to reach $167.4 billion by 2033, expanding at a 31.3% compound annual growth rate throughout the forecast period.

In addition, IBM’s AI Index tracks the adoption rate of the technology across all industries. They find that 35% of companies in all industries are using AI to date, with another 42% implementing a formalized AI strategy within their business. While these companies may not be using AI yet, they have a plan to use it moving forward.

Ultimately, AI is growing extremely quickly and it’s being rapidly adopted, but many professional services firms may not understand what it requires and how to effectively adopt it within their business. 

The impact on personnel

First and foremost, we see AI as a net job creator for professional services companies and all industries. Companies will need to staff positions such as AI software engineers, data architects and automation administrators—people who will actually design and implement AI solutions and test and train models.

Companies do have options, as they can hire that necessary talent in-house or outsource necessary personnel. But those positions will need to be present and relevant in any company that adopts AI.

With these new staffing demands, firms need to keep the tight labour market in mind. Even with some wavering and more economic stability, unemployment rates are still at record lows. And the hiring environment is even tighter for professional services firms, as companies usually require a higher level of education and specialized degrees. Considering these challenges, firms might be best served to leverage external providers for AI resources and services. 

Benefits and risks

When AI applications are effectively implemented within professional services firms, they drive several process efficiencies that mimic many manual, repetitive tasks that humans currently perform. At this point, firms are having a hard time hiring people to perform those tasks in the first place. So, an immediate replacement of professionals will not occur in those positions. Instead, a firm may go from typically hiring 20 new associates each year to bringing on only five to 10 because AI covers some of the functions at the basic task level.

However, it’s important to note that AI doesn’t possess human-like consciousness, subjective experiences and the ability to make ethical judgements. Many tasks within professional services firms require that perspective, so in many cases, AI can only generate responses by predicting what will come next in a sequence, based on training data.

For example, a law firm needs subjectivity and consciousness to make ethical judgements, and an AI model will have challenges determining what is right and wrong in a specific situation. Therefore, rather than a complete replacement of human tasks, firms are implementing a “humans in the loop” model, with employees working alongside AI technologies to validate and review accuracy of outputs and decisions.

Despite the limitations of AI, many tasks can be accelerated and efficiencies can be quickly realized. For instance, other potential gains in productivity include using AI for drafting agendas, research communications and proposals. But the emphasis should remain on drafting, because anything that comes out of AI systems should be reviewed and validated.

With the level of scrutiny that falls on information that comes from professional services firms, users need to be careful about the information they choose to use. The algorithms that some free AI applications are built on have not been updated for a year or more, meaning that there is no guarantee of accuracy. And some generative AI applications are prone to “hallucinations,” which happen when data is produced based on fabricated or fictitious source information.

“Notably, a number of lawyers used affidavits in court cases citing multiple court cases found in an AI tool—but these cases were not real,” says Sonya King, an RSM director and professional services senior analyst. “This is another example of AI not being able to determine what is and isn’t real, and emphasizes the importance of verification and validation.”

Firms also need to be careful about the information they feed into AI solutions. Any data that is typed into some commonly used AI tools becomes part of that database, and that could trigger privacy issues between the firm and their clients.

A significant change to billing

Another way AI will start disrupting professional services firms will emerge with how value is estimated and billed to a client. Firms commonly charge clients based on billable hours, with whatever time associates or partners put into work directly driving revenue. But now, a task that may have taken humans 10 hours in the past can be accomplished by AI in 10 minutes. That means getting information to clients quicker, but a reduction in billable hours.

This shift will likely result in a change in billing structures, with more value-based billing or fixed-fee engagements to capitalize on the true value that firms are providing to clients. Companies will be making significant investments in AI to drive efficiencies, and to recoup some of that, they should be able to collect a similar amount of revenue.

With that said, we are seeing many very educated clients out there who are pushing professional service firms to use AI so they can take advantage of cost savings. Moving forward, there will certainly be some compromise in a real disruption to billing models. What may have been 100 hours of work in the past could now take 50 hours, with a client possibly billed for 75 hours in a fixed-fee engagement.

“Both the client and the firm win as billing processes change with the evolution of AI, but there will be some give and take to adjust to fixed-fee engagements or outcome-based billing,” says King. “This is a issue that will really shake up the industry.”

Risks of inaction

Regardless of the potential challenges and necessary changes to some existing processes, firms need to evaluate AI applications and where they can enhance operations. Your competitors are certainly determining where AI solutions fit the business and likely implementing plans to decrease fees to fit the emerging fixed-fee model. With AI strategies in place, they will have more time to spend on value-added activities instead of spending time on manual tasks that don’t drive value.    

Getting started

The time to act on AI is now; however, many firms may not know where to start. In most cases, the technology is not actually where to begin, your data is. Again, AI technology is based on a sequence of data and predicting what is coming next. If you do not have good data or an effective data structure, an AI solution will likely not deliver on its transformative potential.

In the professional services industry, a few leaders are driving AI innovation with homegrown solutions or early-stage applications, but many firms are still in the planning phase. Promising platforms are on the horizon, and organizations are putting plans in place. If you are not at that point, the most important step is to make sure your data structure and integrity is ready for emerging AI models.

RSM contributors