AI is everywhere, and businesses don't know where to start. Here's what consultants are telling clients.

• Companies are turning to consulting firms to help work out how to use AI effectively.
• Their clients are asking about how machine learning and generative AI can transform businesses.
• Consultants told BI that questions also include how to make AI worth the investment.

It's possible to pose almost any question to artificial intelligence.

But when it comes to how to use the technology, many companies are directing their inquiries to consulting firms instead.

Doling out advice on AI is making up a growing share of many firms' work. Some 900 of PwC's top 1,000 consulting clients are now working with the firm on incorporating AI into their businesses, a spokesperson told Business Insider.

In 2023, McKinsey & Company brought in a record $16 billion in revenue, partly due to the generative AI boom. Almost 40% of the company's work now relates to AI. And much of that is now moving to GenAI, Ben Ellencweig, a senior partner who leads alliances, acquisitions, and partnerships globally for McKinsey's AI arm, QuantumBlack, told BI.

Boston Consulting Group, for its part, now generates a fifth of its revenue from AI, and much of that work involves advising clients on GenAI, a spokesperson told BI.

"18 months ago, the conversation was all about 'what is GenAI,'" Allison Bailey, the head of people and organization practice at BCG, told BI.  "Today it is, 'How do I actually drive value with AI and drive meaningful change in how we work?'"

Even as some companies focus on how AI might rewrite corporate playbooks, some businesses are asking consultants how to get started. The question could be as simple as where it's wisest to invest resources and training in AI.

Bailey said the "people topics" are critical to the equation. Businesses want to know how to mobilize their workers to embrace the technology.

Greg Sward, head of strategy for technology, media, and telecommunications at KPMG US, said some corporate tech leaders wonder whether they have the knowledge to make smart decisions.

"Many CIOs are afraid that they don't have the right skills," he told BI. They're also worried about how to keep a handle on the technology and what the regulatory environment might look like.

BI asked several consultancies to share the most common questions they're getting about AI and their best advice. Here are some of the themes they identified.

Where to begin

Many companies are still determining how they might use AI and GenAI, according to several consultants.

Jim Rowan, AI market activation leader and principal of Deloitte Consulting, told BI that companies that are new to AI should start by asking some basic questions:

• What are we trying to achieve by adopting AI?
• Do we have the talent, investment, and systems to support deployment?
• Have we addressed data governance, privacy questions, potential biases, and other concerns?

If a company can answer those questions, it should outline areas where the technology will be helpful. Then, "closely measure and monitor its performance to make sure you're actually meeting your goals," Rowan added.

According to Vlad Lukic, global leader for BCG's tech & digital advantage practice, companies should also have a good handle on their data. That's the crucial base for training GenAI. "The first step is making sure you have your house in order from a data perspective. This enables greater seamlessness down the line, and that is where the magic lies," he told BI.

Roy Singh, global head of the advanced analytics practice at Bain & Company, told BI that those just getting started with AI should focus on short-term goals like understanding the technology or driving productivity gains.

Building workers' skills

Companies interested in AI know technology is often only part of the equation. Another component is ensuring that workers are up to speed.

Joe Atkinson, the chief products and technology officer at PwC, told BI that it's essential to help employees learn how to use generative AI rather than relying solely on recruiting AI specialists, who are in high demand.

Getting workers to use the tools might not be as hard as it sounds. Atkinson said the technology is so accessible that almost anyone in a company could use it to create products or services or to be more efficient.

Leading by example is important. Senior managers should use AI to demonstrate how it can be useful, while employees should be given time to explore the tech themselves, according to Deloitte's Rowan.

"It's important to build excitement for AI adoption and communicate openly," he said. So, companies should host events like hackathons, encourage experimentation, and educate workers on how AI could make them more efficient in their jobs.

Reinventing your business

The advances in AI will affect some companies more than others.

And in industries where AI is already forcing a rewrite of business models, companies need to think comprehensively about how they can overhaul operations with generative AI.

Bain's Singh pointed to the grocery business. Companies like Instacart and Carrefour are looking to "invent the next-generation customer experience" using AI, he said, adding that they're not necessarily just looking for efficiency gains.

Most companies, regardless of how much they might have dabbled in AI, should be flexible enough to accommodate new developments, according to PwC's Atkinson.

He suggests that companies should build AI systems with an open architecture approach, which is a way of designing software that makes adding, upgrading, or swapping out elements easy.

A return on investment

Bain's Singh said companies often wonder what sort of productivity gains and other financial benefits they might expect from using AI.

Yet he's said companies are starting to see measurable gains — sometimes even huge improvements — from AI. This might be in areas like software engineering, finance, or human resources.

Singh said many companies — especially those loaded with knowledge workers doing desk jobs — can expect to notch productivity improvements of 15% to 20%. Sometimes, it's far higher. In businesses where AI can take over repetitive tasks, the boost to productivity can be upward of 50%, he said.

That doesn't mean each worker necessarily gets back half of their time. But if the time it takes to complete some work is cut in half, then someone working in marketing, financial services compliance, or a life sciences regulatory role might eventually get back the equivalent of a full day's work.

Singh said it might take a company two to three years to achieve major efficiency gains but that the opportunity is immense.

In most cases, he said, it's important to set expectations about how new AI is for many businesses.

"We should all have the humility to admit we're very early in the adoption cycle," Singh said.