Salesforce’s Agentforce World Tour 2025 moved across the globe, from London to San Francisco, uniting admins, developers and consultants around the future of service enablement.
Our tech team attended sessions across the 2 days to capture the most important trends and takeaways for anyone working in tech strategy and transformation.
Below is our report on the 2025 tour’s biggest learnings and lessons to bring into your technology transformation.
AI is here to stay and must be personalised
Speakers emphasised that generative AI, predictive AI and service automation are no longer “future bets”, but baseline expectations.
It was repeatedly mentioned that generic AI won’t cut it. The focus in 2025 is:
• Training models on your proprietary service data.
• Surfacing insights that matter for your agents, like next best actions, tone suggestions or policy guidance.
• Ensuring governance and data privacy in regulated industries.
• Training models on your proprietary service data.
• Surfacing insights that matter for your agents, like next best actions, tone suggestions or policy guidance.
• Ensuring governance and data privacy in regulated industries.
Salesforce showcased Einstein Copilot and new customisation layers to make AI assistants genuinely useful for contact centres and field service.
Key insight: AI must be tailored to your workflows. Off-the-shelf tools need to be adapted to deliver real productivity and security.
Agent productivity = business value
Agentforce Tour demos showed how new console experiences, AI summaries, and next-best-action prompts reduce friction for agents. Freeing them to deliver better service with less training overhead.
Salesforce’s own event promo called it: “A new era of productivity, personalisation, and AI.”
The human element still wins
Despite all the talk of AI and automation, nearly every session had the same underlying theme: Your agents are your brand. Enablement isn’t just about speed – it’s about empowering agents to be empathetic, knowledgeable and trusted advisors.
Best practices highlighted included:
• Continuous coaching and microlearning
• Embedded knowledge articles in console
• Easy escalation with full context
• Continuous coaching and microlearning
• Embedded knowledge articles in console
• Easy escalation with full context
Ensure whoever is leading your AI initiatives have been well briefed on this and are training your tools with robust learning scenarios.
Trust and data ethics matter more than ever
As one speaker put it: “If you don’t earn trust with AI, you lose customers.”
Agentforce sessions repeatedly returned to:
• Transparent AI
• Explaining recommendations to agents
• Protecting sensitive data in generative outputs
Agentforce sessions repeatedly returned to:
• Transparent AI
• Explaining recommendations to agents
• Protecting sensitive data in generative outputs
Salesforce positioned its trust layer as a competitive differentiator, which is not yet present in other CRMs. If you’re working with a service provider, be sure to ask how they are protecting your data.
Our final thoughts about the upcoming normal for service businesses:
• AI isn’t optional, but must be tailored
• Data unification is a competitive edge
• Human connection is your differentiator
• Using tech to build trust is key
• AI isn’t optional, but must be tailored
• Data unification is a competitive edge
• Human connection is your differentiator
• Using tech to build trust is key
We have defined and are rolling out our controlled AI use cases to help create efficiencies across the business, which pass on countless benefits for our clients. Change management, as with any tech roll-out, is the big internal challenge to get right. As a tech and data focussed company, we have change engrained in our values and how we approach day-to-day work.
If you’d like to know more about how our setup is benefitting our teams and clients, feel free to reach out.
