Scaling Salesforce AI isn’t a technology upgrade. It’s an organisational choice. This newsletter discuss how leaders can assess readiness, prioritise use cases, and turn pilots into real, organisation-wide business impact.
AI at scale: Why good intentions aren’t enough
Most executives say they want AI at scale. What they actually fund are pilots, feature tests, and AI-generated summaries.
The gap between these two statements explains why so many AI initiatives stall.
IMD’s 2026 AI Trends report highlights a simple reality: AI is no longer just a technology topic. It exposes how organisations think, decide, and operate.
Treating AI as something to “try out” inevitably leads to limited adoption, isolated wins, and missed enterprise impact.
Salesforce is the CRM platform of choice for many companies, and it’s the platform I’ve been helping organisations optimise for AI adoption, from Einstein GPT to Agentforce.
Yet most executives remain in short-term mode, treating AI as a set of features rather than a capability that reshapes how work gets done. Until organisations confront this short-term trap, medium- and long-term scale will remain aspirational.
Scaling Salesforce AI is not a technical upgrade. It is an organisational decision.
Why this question matters
Scaling AI is not about selecting the right features. It is about creating an organisation that can actually use AI as part of everyday work.
Thinking big is easy. Doing the actual work is harder. The challenge is that so much work needs to happen across strategy, data, governance, and culture.
- When AI remains optional or disconnected from core CRM processes, it becomes just another underused enhancement.
- When embedded properly, it changes the rhythm of the organisation: Work moves faster, decisions happen earlier, and teams focus more on outcomes than administration.
- When companies celebrate short-term efficiency gains and stop there, mistaking minor improvements for transformation, it’s cosmetic change.
Quick reflection: The gap between ambition and execution isn’t failure. It is fertile ground for value creation.
AI & the data reality
AI exposes data problems fast. Unreliable data, over-customised objects or manual workarounds are amplified, not corrected, by AI.
Key Principles for Scaling Salesforce AI:
- Look at use cases from a data perspective: Which data will power your AI agents
- Audit your data: Is the right data available, at the right time, in the right format? It’s about the critical importance of identify trusted, advisory, or off-limit sources.
- Governance is essential: Security, privacy, and ownership are non-negotiable.
- Leverage Salesforce tools: Data Cloud unifies systems, Trailhead teaches governance, and Einstein Trust Layer ensures reliable, auditable outputs.
This forms the foundation of a data readiness plan. Without clean, governed, contextualised data, even Einstein GPT or Agentforce solutions become noisy and underused.
Assessing AI maturity in a CRM reality
Most organisations do not start their CRM + AI journey from a greenfield. They already have CRM systems in place. Often heavily customised, unevenly adopted, and politically charged. We inherit history, habits, shortcuts, and exceptions accumulated over years.
AI maturity is inseparable from CRM maturity. Fragmented, inconsistent, or mistrusted CRM systems create an environment where AI can inherit and amplify weaknesses rather than deliver value.
The key question:
Is the organisation truly ready to scale AI on top of its current CRM reality?
Understanding this reality sets the stage for assessing whether your organisation is ready to adopt AI at scale.
General AI maturity: Organisational readiness first
An AI maturity (or readiness) assessment evaluates how prepared an organisation is to adopt, scale, and extract value from AI. It is not a technical checklist. It is a strategic diagnostic across capability, culture, data, governance, and execution readiness.
The assessment identifies gaps: where the organisation is strong, where it is weak, and which areas are critical to address.
The goal: Understand whether the organisation can move from experimentation to more ambitious, scalable AI initiatives.
In the next edition, I’ll discuss a framework to assess AI maturity, pinpoint your current state, and prioritise initiatives. This is the first step toward moving beyond short-term wins to true, organisation-wide transformation.
CTA: Where would you place your organisation today? Reply. I’ll use your feedback to share practical next steps.
Salesforce AI maturity: A platform-specific readiness
For organisations using Salesforce, a second layer is required.
The Salesforce AI maturity assessment tailors readiness to the Salesforce ecosystem, focusing on how it can realistically create value within the existing CRM environment.
It evaluates:
- Current Salesforce landscape (clouds, data model, licences)
- Level of customisation and technical debt
- Salesforce data readiness: quality, structure, governance, compliance
- Process and workflow readiness for AI-driven automation
- Integration with surrounding systems
- Adoption and trust in AI insights among users
- Ethical, security, and policy frameworks
- Scalability plans and plans to adopt other Salesforce clouds
Objective: Ensure Salesforce AI delivers business value without disrupting existing CRM operations.
Our team regularly helps organisations assess their AI readiness, identify gaps, and prioritise initiatives so they can move confidently from pilots to enterprise-scale impact.
Prioritising AI use cases: Pick what actually matters
AI is a luxury problem. Organisations rarely fail from lack of ideas. They fail from indigestion. AI opportunities are everywhere, but resources are limited.
We need to focus, scope and answer one single question:
Where to prioritise your AI investment?
- Focus on impact: Identify high-impact areas where AI enhances daily work. Avoid flashy features.
- Solve the right problem: Target root causes, not superficial symptoms.
- Build for scale: Design solutions that can be reused across teams, regions, or workflows. A use case that can’t be applied 50–500 times is just a pilot dressed as transformation.
- Focus on continuous improvements: Some use cases work best as ongoing product enhancements.
In a future edition, I’ll discuss prioritisation in more detail.
Short-term quick wins: The trap most organisations are in
Most organisations succeed at short-term pilots: Sales reps save time, service cases move faster.
Early wins matter:
- They demonstrate AI works in the real world.
- They build credibility with sceptical stakeholders.
- They show leadership commitment to transformation.
Most organisations stop here. Wins remain isolated, adoption limited, dashboards may look good. Impact stays cosmetic.
The problem: Mistaking early wins for scale is the single biggest reason medium-term impact never happens.
Mid-Term initiatives: Scale and orchestrate AI agents
Once foundations are stable, and the organisation ready to move beyond short-term wins, he real work begins: Embedding AI across the business.
At this stage, multiple AI agents work together:
- One agent flags at-risk opportunities
- Another suggests next-best actions
- A third one prepares follow-ups, escalating to humans only when needed
Mid-term success requires:
- AI literacy across teams
- Standardised platforms and reusable models
- Governance that supports scaling, replication, and continuous improvement
Scaling AI is not about adding tools. It’s about connecting, standardising, and enabling repeatable adoption.
Long-term transformation: Agentic AI as an operating model
More mature organisations can launch transformative initiatives that redefine business models and competitive positioning.
AI becomes part of the operating model.
- Systems and AI agents are connected across the organisation.
- AI becomes a co-pilot, amplifying creativity, judgment, and innovation.
- Agents predict outcomes, execute low-complexity tasks, and escalate intelligently.
- Culture, trust, and decision discipline are key, not technology.
- Continuous learning loops allow the organisation to adapt and innovate in real time.
Operational efficiency is no longer the goal. AI drives value creation.
Very few organisations actually reach this stage. Maybe too early, still. For most, it is not that the vision is wrong. It is that organisational prerequisites are not in place.
Practical question: Can you really scale AI during a global CRM rollout?
In a CRM process harmonisation programme, organisations and consultants already do the hard work:
- Documenting As-Is processes
- Identifying operational pain points
- Designing To-Be processes
- Building and deploying the solution
The question is:
Should we take this as an opportunity to add and deploy AI capabilities at the same time across the organisation?
Introducing AI at this stage can actually sharpen the exercise:
- Clarifies process design: If a process cannot be clearly described in the To-Be model, it cannot be automated or augmented by AI.
- Highlights inconsistencies: Reveals processes that are overly local or dependent on tribal knowledge.
- Prevents costly retrofitting: Avoids expensive rework later in the rollout.
The real question isn’t whether this is ideal. It’s whether it is worth doing, and under what conditions.
I’ll explore the conditions, risks, and trade-offs of combining AI with process harmonisation in a separate article.
CTA: What’s your view? Would you introduce AI during a CRM rollout, or wait until processes are fully harmonised? Reply or comment. I’d love to hear your perspective.
What the journey teaches us
Scaling Salesforce AI is about building an organisation that can absorb it at scale. It is a journey, not a project.
- Reality Check: Messy data, inconsistent processes, sceptical users.
- Short-Term Pilots: Visible wins, limited adoption.
- Mid-Term Scaling: Orchestrated agents, standardised processes, formalised governance.
- Long-Term Operating Model: Agentic AI augments decisions end-to-end, reshaping operations.
Few organisations are truly ready today. Most are stuck in Phase 2, thinking they are in Phase 3. Recognising this gap is step one toward real scale.
Success requires vision, data, prioritisation, governance, and culture converging. Salesforce AI succeeds when teams, processes, and leadership align with technology. Not the other way around.
Takeways for leaders:
- Scaling Salesforce AI is about organisation, not technology
- Start with CRM reality, not AI hype
- Prioritise what scales. Focus on reusable, high-impact use cases
- Measure what matters
- Data quality and governance are non-negotiable
- Transformation is a journey: short-term wins → scaling → AI as an operating model
CTA
What is the biggest barrier to scaling AI in your organisation today? Reply or comment. I’ll share practical next steps based on your feedback.