Do LLMs Add Value? A CIO's Perspective on the Real Impact of Large Language Models

Richard Keenlyside
May 7
3 min read

TL;DR: Large Language Models (LLMs) like ChatGPT undeniably add measurable value across industries—from streamlining operations and enhancing productivity to driving strategic transformation. But, like all tools, their true worth is determined by how intelligently they are implemented.

Flowchart titled "Maximising ROI from LLMs" with four colored boxes: "Define the Business Problem First," "Integrate with Data Strategy," "Embed within Workflows," and "Invest in Governance and Ethics." — Maximising ROI from LLMs

Do LLMs Add Value? A CIO's Perspective

As someone who has navigated digital transformation programmes from the boardrooms of global enterprises to hands-on advisory roles with innovative startups, the question of whether Large Language Models (LLMs) add value isn't theoretical—it's practical and pressing.

In my roles spanning companies like Sainsbury’s, LoneStar Group, and across private equity-backed ventures, the injection of AI-driven capabilities such as LLMs has transitioned from experimental to essential.

Understanding LLMs Beyond the Hype

Large Language Models, including OpenAI’s GPT and Google’s Gemini, are capable of natural language understanding, generation, summarisation, and even task automation. But the key question remains: Do LLMs move the needle in enterprise value?

From experience, yes—but with caveats.

Where LLMs Are Creating Tangible Value

1. Productivity and Automation LLMs significantly cut down manual tasks. At M.I. Dickson, we deployed an AI-powered chatbot that improved HR and finance operational efficiency by 40%. These aren't abstract figures—they translate into real headcount reallocation and cost reductions.

2. Decision Support and Analytics LLMs integrate seamlessly with BI platforms like Power BI to provide real-time insights. At a recent client in the utility sector, we utilised AI-driven automation to save 75,000 hours annually. It's in these types of augmented decision environments that LLMs prove most potent.

3. Customer and Stakeholder Engagement Natural language interfaces improve user experience and engagement. Whether through AI-driven search, intelligent document generation or customer support, LLMs elevate digital touchpoints—essential in today’s omnichannel ecosystems.

4. Speed to Market and Innovation Cycles By accelerating content creation, software prototyping, and even legal contract drafting, LLMs shorten innovation cycles. At Quollify, an AI-based platform, we harnessed video and LLM matching tech to accelerate MVP development and funding acquisition.

The Misconceptions

Some executives wrongly assume that plugging in an LLM will revolutionise operations overnight. Without a solid strategy, governance, and change management structure, LLMs can become overhyped distractions. Technology alone is not transformation.

Maximising ROI from LLMs

Define the Business Problem First: LLMs should be used to solve specific pain points—not as a solution looking for a problem.
Integrate with Data Strategy: Without clean, structured data, LLMs can't deliver meaningful insights.
Embed within Workflows: Standalone AI won't scale. Embed LLMs into existing ERP, CRM, and collaboration tools.
Invest in Governance and Ethics: Ensure compliance with data privacy laws (e.g., GDPR) and AI ethics.

FAQs

Q: What types of businesses benefit most from LLMs? A: Any organisation with repeatable processes, high volumes of text or customer interaction—such as retail, logistics, financial services, and healthcare.

Q: Are LLMs secure enough for enterprise use? A: When implemented with proper controls, LLMs can be secured through private models, local hosting, and robust data governance.

Q: Can LLMs replace human roles? A: They’re better suited to augment, not replace—think augmentation of skills, not displacement.

Closing Thoughts

LLMs do add value—but only if implemented with a clear purpose and tied to business outcomes. As CIOs and strategic leaders, we must balance innovation with practicality. AI is no longer a buzzword; it's a business enabler.

Richard Keenlyside is the Global CIO for the LoneStar Group and a former IT Director for J Sainsbury’s PLC.

Call me on +44(0) 1642 040 268 or email richard@rjk.info.

Follow me on X https://x.com/cioinpractice & LinkedIn https://www.linkedin.com/in/richardkeenlyside/

Would you like me to generate a complementary social post or image for this article?