Overcoming the AI confidence crisis: How IBM and Salesforce empower UK enterprises

I've noticed something striking. While customers are excited about the transformational potential of AI, many are hesitant to move beyond simple pilots. This isn't just my observation—the data backs it up.

The recent IBM Institute for Business Value (IBV) survey of Salesforce users revealed that only 16% feel truly confident in using AI to achieve business results. What's more telling is that 45% are interested but either don't know where to begin or have concerns about the implications.

From my experience working with clients at IBM Consulting®, I've come to believe that this confidence crisis stems from the fact that AI is far more than just another technology upgrade. The capabilities it unlocks require organizations to fundamentally rethink their operations and their business models, from how they collect and integrate data to their business processes and human plus digital workforce.

Take Agentforce, for example. I often show our clients how it enables both assistive agents that augment human workers and fully autonomous agents capable of handling multistep processes. These aren't your typical chatbots; they're intelligent assistants that can work independently or alongside your team, handling routine processes while knowing when to escalate complex cases to human colleagues. This digital labor is about augmenting workforce productivity, not replacing it.

While the technical implementation might seem straightforward, integrating an agent into an established workflow raises numerous questions about the roles, actions and channels you want to establish for the agent—not to mention the guardrails around trust and reliability. And that's before we even tackle the data challenge, which is the most critical factor in an agent's success.

In my client engagements, I've noticed a consistent pattern: moving from pilot to production often becomes a structural issue because their most valuable data typically resides outside Salesforce. Taking full advantage of Agentforce requires organizations to consider end-to-end processes, which means helping ensure tight integration between Salesforce and other enterprise systems holding crucial data.

Our IBV report findings reflect what I see in the field. While 97% of Salesforce users collect diverse data, only 24% use it to transform their customer experiences. In my experience, this gap often exists because organizations struggle to access their own data effectively.

This is where IBM solutions make a real difference. Integrating IBM® watsonx Orchestrate^TM into Salesforce Data Cloud helps UK enterprises unlock access to data from widely used systems such as SAP, Workday and Microsoft® 365. Its thousands of prebuilt connectors and skills enable the intelligent automation of complex cross-application workflows. Connecting front-to-back processes seamlessly can maximize returns on agentic AI investments.

I'm particularly excited about our Salesforce LLM Open Connector, which allows users to create API gateways and proxy servers connecting LLMs such as IBM Granite^® to Einstein Studio. Granite 3.x is our latest AI model, designed specifically with enterprise needs in mind. What makes it special is its ability to work across multiple languages and understand complex business contexts while adhering to IBM's stringent AI Ethics principles.

We've built it for trustworthy enterprise usage, carefully selecting and vetting all training data to help ensure it meets the highest standards for mission-critical applications. It’s openly available for businesses to use and adapt, which shows our commitment to transparent innovation. This has been especially valuable for my public-sector clients and those in heavily regulated industries, who can keep all data and AI activity on trusted models running in IBM's UK data centers.

Another of my favorite points to discuss with clients is IBM's membership in Salesforce's Zero Copy Network. I've seen the frustration of valuable data trapped in legacy systems, and Zero Copy allows teams to use data from mainframes, databases and hybrid cloud sources without moving or duplicating it. This means that both human teams and AI agents can access decades of valuable information directly through Salesforce, removing the pains of straddling multiple systems.

When clients ask about implementation, I'm proud to share how we're using AI to implement AI. Our IBM Consulting Advantage platform has transformed our Salesforce solution delivery. By using our Intelligent Delivery Suite for Salesforce, our teams can complete tasks in minutes that previously took days, resulting in increased and more consistent quality, cost savings and a quicker time to market for our clients.

IBM has decades of experience and proven data strategy and integration tools that complement the Salesforce AI tools. Looking at the bigger picture, we also have crucial experience in delivering end-to-end AI solutions for our own organizations and for clients. That means we can advise you on how to get started and where best to invest.

While AI advances are happening almost daily, success requires a marathon mindset rather than a sprint. I encourage clients to adopt an outcome-based approach: identify where AI can deliver business value, break the problem into smaller chunks that deliver ROI, get it done and then iterate on the solution. I've seen firsthand how these iterative projects can flourish and attract more capital investment, ultimately becoming significant success stories.

What excites me most about our partnership with Salesforce is helping organizations find their unique path to AI adoption. Whether it's improving customer service, streamlining operations or enabling more informed decision-making, we're helping organizations across the UK harness the power of AI in practical, secure and effective ways.

If you're considering how AI can transform your organization's use of Salesforce, email me to learn more.