Enter the Agentforce

September always marks the return of conference season to the marketing world, and not just for event marketers. Both Salesforce and Hubspot hold their main customer conferences in which they present their vision for the next twelve months. That vision has the potential to shape the entire industry, given the importance of their platforms and their long-standing status as CRM thought leaders. This year was no different, especially given the mismatch between Generative AI hype and real-world experiences of the technology.

Agents

Both Salesforce Dreamforce and Hubspot Inbound took place last week and, as predicted, were dominated by Generative AI talk. Both firms announced major new AI platforms for customers, while providing a succession of demos to highlight the incredible potential of the technology to improve the customer experience. However, marketers have seen many promising AI demos in the last two years. The difference is that platform owners are only now introducing the tools for businesses to adapt AI to their workflows.

As is typical, Salesforce made the more ambitious pitch. They announced  Agentforce, a low-code builder for AI agents. Microsoft made a very similar announcement at a special event just a few days prior. The difference is that few companies do low-code development better than Salesforce. Agentforce uses many of the same customisation tools as the rest of the Salesforce platform, which should aid adoption. Admins familiar with Salesforce Flow should feel right at home when building agents and chatbots using Agentforce - as it uses pretty much the same UI. Agents are not just limited to Flow though. There are also APIs and Apex trigger support for coding more complex use cases.

Models

For all the much-vaunted flexibility, Salesforce have announced a series of pre-built agents that cover the standard customer service and marketing use cases that are frequently handled by chatbots today. On top of that, agents can be internally facing too. Sales coaching is another of the pre-built agents that is available for admins to customise. The Agentforce prompt builder then allows those agents to be tweaked with bespoke training and custom guardrails to fit business requirements. The most useful piece is the integration with Salesforce Flow so that agents can trigger assigned flows based on real-language queries from customers or internal users.

It's this integration with business workflows across the Salesforce ecosystem that makes Agentforce so interesting. The same agent could be used in a business process that spans Sales Cloud, Service Cloud and Marketing Cloud. Salesforce sees agents as a way of pushing companies to buy into all of the different product clouds across the platform, rather than just focusing on the core CRM functionality. The vision for Agentforce very much involves embedding service and marketing workflows into agents even for companies that use other platforms for support or email marketing.

Foundations

Salesforce are even bundling the basic features from the Sales, Marketing, Commerce and Service clouds into the enterprise tier of the platform for free, reversing a product split made a decade ago. The new feature bundle is called Salesforce Foundations and was also announced at Dreamforce. Foundations is pitched as a way of leveraging agents for basic customer service or marketing functionality even for businesses that don't subscribe to the relevant Salesforce products. There are heavy usage restrictions on Foundations, which is being pitched as a try-before-you-buy model, so as not to cannibalise existing subscriptions.

It is pricing which will ultimately decide whether Agentforce is a success. Agentforce is heavily integrated into Data Cloud, which uses consumption based pricing. That could limit adoption. Mark Benioff wants to see one billion AI agents created within Salesforce by the end of next year. That's probably too ambitious. However, there are definite use cases for Agentforce. In time, I suspect the internal facing applications of Agentforce will become dominant. It provides an excellent UI for busy sales reps to request quick updates to their accounts and opportunities. In the interim, Agentforce will definitely have Drift and other live chat services concerned.

HubSpot Debuts Breeze

Meanwhile, in Boston, HubSpot announced their new Breeze AI. This encompasses many of Hu Spot's existing AI capabilities as well as some new ones, which naturally include a set of AI agents. As always with HubSpot, there is a strong emphasis on ease of use over flexibility, which means that each Breeze agent is limited to a specific use case. These cover data and reporting as well as content generation use cases.

Microsoft Announces Copilot Agents

Microsoft's main customer conference isn't until November. That didn't stop them from announcing a whole range of new features for Copilot last week, including an enhanced Agent builder for Copilot Studio. This integrates with Teams and Dynamics to support customer experience use cases. Improved customer services and enhanced sales efficiency were among the highlighted benefits of the announcement.

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.