Beyond the Platform: Salesforce meets Claude

Several months ago, Salesforce announced a major product strategy change. The entire Salesforce platform was going headless. A new generation of agents could manipulate Salesforce data and business workflows without logging into the platform. Headless 360 got the attention of developers, who have long been frustrated by the esoteric languages and custom tooling needed to develop within the Salesforce platform. Unfortunately for Salesforce’s investors, much of the functionality required to deliver the headless vision is yet to reach general availability.

Claudeforce

The announcement of Headless 360 also caught the attention of business users. Yet, many dismissed it as a mere back end update. They failed to grasp the possibilities that Headless 360 would open up for their day-to-day contact and opportunity updates. In that respect, the announcement of Claudeforce last week was a turning point. It was an integration that got a lot of attention, far more than most new Claude connectors. Yet, in actual product terms, it's not really a big deal. As such, it was announced during an earnings call, rather than at Dreamforce in a few weeks.

Claudeforce has an impressive scope. It promises access to 37 new AI skills, which cover many day-to-day sales activities. Yet, with the right prompting, Claude has been able to do meeting prep and pipeline reviews for some time. Sure, there wasn't an official one click connector for it. However, the same capabilities were available through the Salesforce MCP server released last year. A native connection makes security reviews easier, which is an especially relevant concern for a core business platform such as Salesforce. Furthermore, a native skill saves time by cutting down development time for AI power users and operations teams.

AI Adoption

The entire episode illustrates the rather basic current state of AI adoption in the typical enterprise. For all the talk of agentic automation, AI is not trusted with repeatable tasks. Business users still don't have confidence in AI output, and for good reason. Consequently, the technology is rarely used for business processes that run without human oversight. Where available, rule based automation approaches are the preferred approach for automating manual grunt work. Sure, those traditional workflows may contain an AI component to them, but the actual trigger and execution is deterministic.

Instead, generative AI is typically used for one-off routine tasks. The classic example is drafting an email. Templated activity alerts are generated using the same CRM automation tools as always. One-off customer communications are often co-written with AI. It depends on the rep, their confidence in the technology, and their writing ability. This has resulted in a general improvement in the clarity and quality of the typical business email. People are dismissive of AI generated content, but it does have its place.

Assisting the Routine

The real benefit of AI tools such as Claudeforce is in tasks that occur regularly, but are never exactly the same. Meeting prep is a good example. Plenty of AI tools already offer skills to support this. Microsoft Teams will surface relevant documents when viewing a meeting invite in Teams. AI notetakers will send a recap of the previous call. The trouble is that none of these features provide the right level of detail. Every call is different, and requires different types of briefing material. A sales call requires different preparation from a 1:1 with your manager. Salesforce can help with the former, but it's not much use for the latter.

However, Claude (or competing AI platforms) can help with preparing for any meeting type, as long as the AI understands the right places to pull the most relevant information for each call. That requires business-specific context only available to a chatbot with access to the entire enterprise knowledge graph. Supplying that context is still very much a human task, requiring a prompt or AI skill developed by the user. The advantage of Claudeforce is that it short circuits the AI customisation process, but only if you're a Salesforce shop.

Fortunately, custom AI skills are relatively easy to develop. As such, they're proliferating across the enterprise, even if they're not always shared when they should be. For operations teams, developing and maintaining custom AI skills and custom AI chatbots is fast becoming an essential skillset. It ensures the entire business gets the full benefit of the AI tools being pushed by management. Otherwise, you're relying on vendor trained AI such as Claudeforce, which may not deliver the right answer for every organisation.

The Quest for ROI

Generic AI features do still have their place. They just limit the potential of the technology. AI chatbots have been a game changer in enterprise. Not because of the generative capabilities of LLMs, although that has been useful. The real ROI has been realised through vast knowledge banks. Claude and ChatGPT are now the central tools for accessing information across the business. In part because they are one of the few platforms that have direct access to all corporate data silos.

Investor pressure and AI adoption mandates have made such widespread access possible, delivering the promised efficiency gains that business leaders have been looking for. Trouble is, those same executives aren't just interested in quicker and more accurate information retrieval. Justifying the cost of AI investments necessitates a direct impact on the bottom line. That requires more sophisticated usage of the technology, which is still a long way from production in most organisations.

Integrate Acquires Calibermind

The combination is interesting. One is an enterprise lead intake platform specialising in managing content syndication leads; the other is a multi-touch attribution platform that reports on revenue by channel. The ambition is to break down the silos that block end-to-end media ROI reporting. Good luck to them, as they're not the first companies to take on this challenge.

Operationalising ABM in the Real World: Your Step-by-Step Guide

Finding it difficult to navigate the complexities of ABM. A new eGuide from CRMT Digital cuts through ABM hype, explaining how to implement and run an ABM program in practical terms. Boost ROI with practical approaches to account-based targeting, personalisation and engagement.

Marketing Operations Roadmap Matrix

A Marketing Operations Roadmap is an essential strategic tool that aligns marketing priorities with your overall business goals. Any roadmap begins with understanding where your marketing organisation is right now, and then defining where you want to get to. This free Marketing Operations Roadmap Matrix from CRMT Digital allows you to do exactly that.