Salesforce has moved fast into AI. Einstein 1, Data Cloud, Agentforce, Slack AI, new releases every few months. 

If you are responsible for revenue, operations or technology, it may feel exciting and overwhelming at the same time. Licences are getting more powerful, but your reality may still be dashboards exported to Excel, manual handoffs and AI pilots that never quite go live. 

This article cuts through the noise and focuses on what has really changed in Salesforce, where most organisations get stuck, and how to start unlocking value in a realistic way. 

1. Salesforce is now an AI and data platform, not “just CRM” 

The biggest shift is strategic. Salesforce is repositioning Customer 360 around Einstein 1 and Data Cloud, so your CRM becomes a real time data and AI platform, not only a system of record. 

Data Cloud sits at the centre. It connects data from your Salesforce orgs and from external systems, builds unified customer profiles and makes that context available to AI, analytics and automation. If AI is the brain, Data Cloud is the nervous system. 

On top of that, Salesforce has introduced Agentforce. This is an AI helper and agent embedded inside Sales, Service and other clouds. It can summarise records, suggest next actions, generate content and trigger Flows from natural language, without forcing your teams to learn a new tool. 

The message is clear. Salesforce is no longer adding AI as a side feature. It is rebuilding the platform around AI and data. 

2. Why most organisations are not seeing the value yet 

When we talk to leaders, three pain points appear again and again. 

Fragmented data 
Customer interactions, contracts, product usage and finance data live in different systems, owned by different teams. AI cannot make good decisions if it does not see the full picture or if basic fields are incomplete and inconsistent. 

Complex, fragile processes 
Years of urgent requirements have created layers of flows, rules and workarounds. Nobody feels comfortable touching them. New AI features are dropped on top of this, which makes everything feel more unpredictable instead of more intelligent. 

Lack of AI governance 
Everyone wants automation. Nobody wants a compliance issue or a bad customer experience. If there are no agreed guardrails, legal and security teams slow things down, often for good reasons, and AI projects stay in pilot mode. 

The result is a familiar pattern. The organisation pays for powerful Salesforce capabilities, but day to day, teams still rely on manual work and side spreadsheets. 

3. What to focus on now 

You do not need to implement every new feature to benefit from this new Salesforce wave. You need a sequence. 

Start with the data that matters most 
Pick one or two high-value slices of data, such as customer profiles and recent interactions, and focus on making them clean, connected and available in Data Cloud. The goal is not perfection. The goal is an initial, trusted foundation that AI can use. 

Choose a small number of focused AI use cases 
Instead of “doing AI everywhere”, pick one or two scenarios with clear outcomes. For example, inbound lead qualification, first-line support for a defined set of cases, or an internal knowledge assistant for your sales or service team. Design them so that humans stay in the loop, especially at the start. 

Design governance from day one 
Agree what AI agents can and cannot do. Decide where a human must approve actions, what is logged, and how issues will be reviewed. This makes legal and security part of the solution, not blockers at the end. 

Invest in people, not only in licences 
AI features will keep evolving. Your real advantage is people who know how to work with them. Train admins, architects and business owners on how to design prompts, flows and guardrails. Help users understand when to trust AI, when to question it and how to give feedback. 

4. How Target Everest fits into this picture 

If you have not come across us before, we are a specialist consultancy that works at the intersection of Salesforce architecture, data, AI and strategy. We focus on turning platforms into performance, not just technology into licences. 

In practice, that means we help organisations: 

  • Analyse their current Salesforce landscape and identify where data and processes block AI. 
  • Design an architecture that makes sensible use of Einstein 1, Data Cloud, Agentforce and Slack AI, without over-complicating things. 
  • Prioritise a small number of AI use cases that map directly to revenue, service or efficiency outcomes. 
  • Run safe pilots and then scale what works, with governance and change management built in. 

The outcome is simple. Fewer disconnected pilots. More visible value, faster, from the Salesforce investments you already have and the AI capabilities that are arriving. 

If you recognise these pain points and would like to explore what an AI-ready Salesforce roadmap could look like for your organisation, we are always open to a conversation.