IBM Consulting makes a big bet on AI-led agentic services

IBM provided a two-day session for analysts to engage with the company’s top global leaders to hear and understand how its way of working is undergoing fundamental changes to deliver faster, provide higher quality, and increase margins. HFS identified a few key takeaways from the sessions and our meetings with IBM executives:

1. The services industry is changing, as we predicted in our Services Tech Vision 2030 (see Exhibit 1). While many providers struggle to adapt, IBM is among the most forward and visible in embracing agentic AI and GenAI applications.
2. A clear operating model for AI-led consulting (what IBM calls asset-based consulting) has not yet been fully established. IBM will need to continue to act quickly before it loses the early mover advantage.
3. IBM is asset-rich as the only services company housed within a global technology company. However, it repeatedly displays a lack of confidence and tends to muddy its market messaging propositions by coming across as too techie. The company needs to drive a powerful business transformation narrative to go with its already established IT narrative. IBM’s BPO services should be an integral part of this transformation.

Exhibit 1: IBM Consulting’s vision is aligned with HFS Services Tech Vision 2030

Source: HFS Research, 2024

IBM Consulting Advantage (ICA) is a solid platform within IBM

Central to the company’s transition to service delivery founded on best-in-class assets is the IBM Consulting Advantage platform. This AI-powered delivery platform is more than just a chatbot or data miner given an AI label; it is designed to enhance workforce expertise by utilizing advanced software assets and methods. The standard interface of a chat prompt is present, but rather than the prompt connecting to one large language model (LLM), the end user can add components that are internally open-sourced and vetted by any IBM employee to create a set of personalized agents and application modules that will provide customized options for the queries and resultant output.

Beyond this, IBM has taken the strengths of watsonx AI models and its recently upgraded release of Granite 3.0 LLM to underpin the platform architecture. A surprise was the announcement of work on small language models (SLM) to focus the LM on what primarily matters to that particular business situation, all running on IBM Cloud. This is years ahead of the expected usage and architecture of business AI forays with significant implications: reduced compute needs and energy use due to its SLM focus increases sustainability and reduces costs; a bespoke platform more than a decade in the making brings the value with more robust architecture; and an interface that enables users to choose from a selection of AI components or create their own. In the past, IBM heavily pushed machine learning (ML) solutions, but it was too early to market for businesses to understand and invest in the technologies. Now, the years of knowledge and work are hitting at the right time.

Betting that ‘invisible AI’ will be the future we need

If this seems exciting for deriving business value from AI, it is. However, this platform is an IBM-only toolset for the foreseeable future. It is not licensed out nor available for general use by other entities, but clients can access the platform within IBM-led projects. This is a strategic move driven by fundamental principles: first, it is a powerful piece of intellectual property expected to aid IBM in establishing margins for services that top consultancies command; second, it isn’t yet fully transparent to platform users as training and technical capabilities are required. In the future, IBM believes AI will be everywhere and non-intrusive. Minimal specialized training will be needed to engage with GenAI and AI agents to complete tasks. In the roadmap of AI development from ‘aid to assistant to team member,’ IBM is deep into the assistant phase and betting heavily on AI team members within the next three years. We expect a first-to-market approach will serve IBM well in making their platform a de facto standard.

IBM recognizes it can’t do it all; partners are needed

By 2026, the company expects AI agents to fill at least 50% of specific roles, indicating a substantial shift toward AI-driven services. This shift is anticipated to unlock value faster, supercharge expertise, and enable further innovation through an ecosystem of partners. Collaboration with an ecosystem of strategic partners such as Adobe, Salesforce, Meta, AWS, Microsoft, SAP, Oracle, and Palo Alto Networks to enhance capabilities and deliver comprehensive solutions is a critical pillar of the solution to bring combined value. In fact, four of the current partnerships have each crossed the $1 billion mark for revenue, with others increasing rapidly. These collaborations leverage each partner’s strengths to foster open innovation and integration of best-in-class solutions, enabling IBM to take advantage of a wide array of technologies and expertise. The partnerships have already contributed to significant productivity gains, with 80,000 active users of the platform reporting 30-50% improvements. By leveraging partnerships, IBM can offer an agnostic approach, recommending the best solutions for clients without being tied only to IBM software.

Improvements in efficiency and client outcomes have been confirmed

IBM’s focus on AI-driven consulting is expected to generate substantial benefits. watsonx is projected to deliver $1.6 billion in AI-related benefits, growing to $3 billion through 2025 as the company strategy achieves several specific improvements in efficiency and client outcomes. The IBM Consulting Advantage platform framework is extensible to many business areas. It has enabled a 50% reduction in effort for tasks such as ideation to sprint creation in product management and a 60% reduction in invoice entry cycle time for a global retail client. In mobile app development, AI integration led to a 45% reduction in cost and a 70% reduction in defects. Additionally, AI governance initiatives have resulted in a 40% increase in productivity, with faster knowledge base productivity and reduced testing efforts. These improvements highlight IBM’s focus on leveraging AI to streamline operations and deliver superior client outcomes.

IBM is all in on developing assets to drive consulting but has areas for improvement

The biggest disappointment of the showcase was that business and strategy consulting, as commonly understood, was not evident. The demos provided were heavily slanted toward application development acceleration. Fundamentally, the discussions on consulting felt more like assistive design thinking, with captured information as the seed of rapid initial code development. This has immense value, yet it is not consulting in the traditional strategic sense. IBM provided verbal feedback on how the platform could support strategic consulting, with Arvind stating that traditional consulting still has about 10 good years left.

Despite strong capabilities, IBM’s BPO services and re-inventing the legacy BPO model were also not in the limelight. HFS Research indicates that the business now controls nearly 45% of the tech spend across the G2000 enterprises. IBM must move rapidly, or this could be a missed opportunity.

A lack of deep industry focus was also evident. The industry defines business problems, and IBM’s new science of consulting must understand and demonstrate industry nuance.

The Bottom Line: IBM is poised to reinvent the future as the world has finally caught up with its vision.

Looking ahead, IBM envisions a future where AI agents play a significant role in its consulting services as it moves toward AI-led consulting services. This transition not only enhances consulting efficiency and innovation now but also positions IBM as a leader in the AI-driven services industry of tomorrow. With unique and deep technology resources, global capabilities, and brand permission to operate in every sector, the only one applying brakes to IBM is IBM.