GenAI success depends on talent, data, cash, and culture

Written by HFS Research

Want to win with generative AI (GenAI)? Here’s a leg up from those who have learned the lessons for you-a cohort of enterprise leaders already experienced in using GenAI in their businesses.

We asked GenAI early adopters and enterprise AI experts what goes into creating the optimal conditions for the effective deployment of GenAI; they ranked the options in Exhibit 1.

It all starts with talent, data, cash-and a culture that embraces innovation.

GenAI runs on chips, but its value is powered by (human) talent

It may be a great irony, but it’s true: As technology becomes more essential to generating value, people become even more critical.

The need for sufficient talent (people) skilled in creating and using GenAI is ranked #1 or #2 by 83% of our experienced cohort.

The talent requirement is not overlooked by the massive global service providers, many of which have already announced huge investments in acquiring and training talent. Accenture, among others, is set on doubling its AI-related talent from 40,000 to 80,000.

“Our $3 billion investment is an important part of how we’re going to help create value for our clients. So, it includes what we’re doing with talent. We’re training thousands of people to be able to be relevant to GenAI. – Julie Sweet, Chair & CEO Accenture

Fausto Artico, Global Head of Data Science (models and tech stack) at GlaxoSmithKline (GSK), believes having the right talent and high-quality data are crucial. He said, “If you have the right talent, in the long run, the models you use will make less difference than the data you can access and that talent can make use of.”

High-quality data was ranked the most important success factor

High-quality data is also among our top four critical success factors; it’s the success factor ranked number one by more respondents than any other. GenAIin the real world is voracious for data-the bigger and cleaner, the better. Our early adopters recognize this, and our findings reveal a concerning truth. Without high-quality, accessible, secure, and accurate data, enterprise leaders will have a hard time creating value with GenAI. At the end of the day, GenAI looks for patterns in data and only generates its responses based on the data made available to it. Technologists have been saying “GIGO”-garbage in, garbage out-since the 1960s, but with large language models (LLMs) and conversational interfaces, this issue is magnified beyond human comprehension.

Sheri Sullivan, Global Payroll Operate Leader at EY, fed payroll regulatory updates for countries, anatomy of payslips, along with individual client company policies into an LLM to deliver improved employee experiences in a multi-lingual and multi-national deployment. In initial testing she found HR leaders were thrilled with the results but in a few of the chatbot responses there were inaccurate answers.

When Sheri’s team asked those HR leaders to check the data -specifically, the policies they had provided and were using -they found it was the policies that needed to be corrected. We are in a world of garbage in-garbage out at massive scale with profound implications for how our organizations operate.

“Gen AI can reveal upstream issues enterprises didn’t know they had. We had a couple of cases in which an HR agent told us the bot was wrong, but they found it was their policy that was wrong. -Sheri Sullivan, EY Global Payroll Operate Leader”

The need for investment the survey identified is fairly obvious, particularly in these early days of exploration and innovation and seeking a march on the competition. But our GenAI superstars told us that another powerful success factor is a culture of innovation, and it’s trickier to put your finger on.

Sheri called out the necessity for culture to embrace rapid decision making and reaction. “We learned a lot from our client interacting with the tools.” This rapid, iterative, test-and-learn mindset is core to working within a culture of innovation. Just as “design thinking” suggests, get prototypes in the hands of your intended users as quickly and as cheaply as you are able.

The Bottom Line: Fix your data first, but don’t overlook talent, cash, and culture.

Leaders should recognize that our Big Four-talent, data, cash, and culture-are interdependent and essential. Still, there is no getting away from the importance of data to succeed with GenAI. Without the right data, all the smart people, capital, and healthy context for exploration won’t amount to real business impact.

Our survey data shows a clear-but tough-road ahead for companies that haven’t yet cracked the code on their own data. Many organizations will finally have to bite the bullet and improve the quality of the data they generate, store, and flow throughout the enterprise. No amount of skilled people, bags of cash, or willingness to innovate will make up for data failings. Start there, and don’t wait.

About this research

Your Generative Enterprise™ playbook for the future is a HFS Research and Ascendion research program based on more than 20 in-depth interviews and a survey of more than 100 C-suite leaders and practitioners with first-hand experience implementing GenAI in organizations.

Watch out for more, and join us on the journey at Ascendion and HFS Research to access all our research findings.