Data and Generative AI: A Window into Your Organisation’s Soul?

1 month ago 10
News Banner

Looking for an Interim or Fractional CTO to support your business?

Read more

For the third consecutive year, HBR Analytics Services has researched contemporary data use, focusing on its applicability to generative AI. The report—“Scaling Generative AI for Value: Data Leader Agenda for 2025”—spans multiple industries and companies globally. It’s time as conversations have noticeably pivoted from abstract use cases and generative AI pilots to a desire to know realise value at scale from investments.

The prior report emphasised the chief data officer’s (CDO) role, but this year’s report includes other influential roles involved in deciding generative AI use.1 It gives a more comprehensive view of how organisations approach AI adoption; the 646 organisations represented use or plan to use generative AI in some capacity.

This study captures insights from multiple business functions at organisations of varying sizes. 39% of respondents work at companies with over 10,000 employees, and 26% work at companies with fewer than 1,000. In line with previous predictions, the most common use cases involve software development, productivity, customer services, and sales and marketing.2

Respondents can be divided into three groups:

  • Leaders (34%) have one or more generative AI use cases established.
  • Followers (31%) are in the early stages of one or more use cases.
  • Laggards (35%) are still considering if or how to move forward.

Rather than summarise the report, I want to pick out key insights and their implications that you can consider as you scale your own initiatives.

The Role of Leadership

Where are we really with generative AI initiatives?

47% of respondents agree or strongly agree that their organisation’s initiatives are going well, while 22% percent are at the other end of the spectrum. What was striking was over 20% of participants involved in generative AI initiatives do not know how they are progressing. Segmenting these responses by adoption rate is fascinating: Nearly 75% of leaders are satisfied with progress, compared to a reasonable 50% of followers but barely 20% of laggards. Reading between the lines, there is a correlation between leaders’ engagement in initiatives, satisfaction levels, and knowledge of how much progress was actually being made. As we have highlighted in numerous situations involving data, the active and visible engagement of a leader is a critical success factor of data-driven initiatives.

Aside from shepherding initiatives, leaders play a second role of bringing clarity of the role generative AI can play in organisations, how it can support business strategy and, critically, the ownership for initiatives. We constantly over-estimate our effectiveness here with 50% of respondents cited this as a challenge.

I am interested in how the CDO role is evolving. In over 33% of organisations, the CDO is the de facto leader of generative AI—which I consider a positive. Most organisations don’t need another band-aid silo created by hiring a CXO role specifically for AI (which is happening in just 10% of organisations surveyed).

CDOs increasingly engage in business strategy, share data ownership with the broader business, and take an active role in leading generative AI initiatives, a heartening positive trend.

Data: The Electricity of the Generative AI World

Nearly all organisations use publicly available large language models. Three competitors can ask a model the same question and get pretty much the same answer, so is differentiation dead?

Not at all.

Your organisation can use data to tailor prompts and outputs, yet over 50% of respondents believe their data foundation is not ready.

Seth Earley, CEO of Earley Information Science Inc., summarises it well: “If organizations aren’t getting good results with AI and generative AI, it’s because they don’t understand the fundamentals. Going from hype to value means focusing on use cases and specific outcomes and getting your data right … And the fundamental issue to get right is data quality. Clean data is the price of admission.”

Major challenges include data cleanliness, integration, and a lack of clear roadmaps for data and generative AI. These challenges cannot just be dumped on IT.

As Tom Davenport, distinguished professor of information technology and management at Babson College, points out: “Data is a broad organizational asset, and it’s hard to put one person in charge of it. What’s needed, and often difficult to achieve, is a common definition of key data elements around the organization and some clarity about who owns data. Further, nobody has been responsible for unstructured content, which now needs to be addressed.”

While most respondents feel overwhelmed by these issues, 87% are investing in solutions, with nearly half focusing on data quality improvements.

Scale Breaks Everything

Time to market remains crucial for differentiation. How quickly can you get an innovation or improvement into customers’ hands? Differentiation in the world of generative AI can be fleeting, so speed matters.3 Respondents cite multiple concerns that restrain scaling initiatives to small customer tests, including liability, privacy, ROI measurement, and unclear roadmaps.

45% of respondents note challenges in getting “business” and “IT” teams to collaborate. This persistent issue spans beyond generative AI and data. It’s time to say enough is enough and recognise that work gets done despite functional labels, not because of them. With data and generative AI cutting across organisations, perhaps we can finely retire that worn, conflict-creating phrase: “the Business and IT.”

Most successful use cases focus on individual functions and lines of business. They are business-led – not dedicated to a specific function like IT – which allows for easier resource allocation and focus.

Don’t Forget Your People

The best data foundation and generative AI strategies mean little if your employees aren’t on board or skilled for transformation. Most organisations have the right people but they may need skill development. I am encouraged to see nearly two-thirds of respondents invest in upskilling or reskilling their existing employees, supplemented by external help. (I’m a fan of the recently released AWS certification on AI, an education I believe most nontechnologists would also benefit from.4) Your people understand your business and can apply technology solutions to real business problems and opportunities.

Generative AI can help reduce undifferentiated work and unlock talent, so you should create space for people to explore generative AI and experiment with new ideas. If you fail to do this, the very people you need may leave your organisation for one that supports them and gets out of their way.5 In the HBR Analytics Services report, UK Government CDO Craig Suckling suggests driving efficiencies to free up time, people, and money for value delivery.

We are in the nascent stages of generative AI adoption, with a whole budnle of exciting technologies on the way. Success requires solid data foundations, leadership, reskilling, and willingness to experiment. This isn’t about a single technology – it is a window into your organisational soul. Get the basics right to break free from productivity constraints and foster innovation.

—Phil

Read more on the research report here.

[1] What’s top of mind for Chief Data Officers going into 2024? Phil Le-Brun, AWS

[2] The economic potential of generative AI: The next productivity frontier, McKinsey

[3] AI Won’t Give You a New Sustainable Advantage, Harvard Business Review

[4] AWS Certified AI Practitioner, AWS

[5] The human side of generative AI: Creating a path to productivity, McKinsey

Read Entire Article