With Agentforce, Salesforce brings new opportunities for clients and partners

At its annual Dreamforce event in California, Salesforce officially unveiled Agentforce, an all-in-one agentics and Gen AI solution. This new offering may finally enable Salesforce customers to get the most from their investments in Customer 360 (CRM), Data Cloud, Einstein, Mulesoft, Slack, and even Tableau. For integration partners, Agentforce may make it easier to implement new Salesforce products and boost the value they provide across enterprise technology, business processes, and operational needs.

Salesforce’s vision of how agents can be applied to business is compelling, and the fact that the company chose to showcase this with a retail client isn’t lost on HFS. In recent research, we have characterized retail as being at an inflection point, primarily due to the need to modernize legacy systems, make better use of data, and adopt emerging technologies for competitive advantage in saleable ways.

Agentforce leverages all of Salesforce’s tools fluidly

So, what is Agentforce? Simply put, Agentforce is Salesforce’s productization of its artificial intelligence, large language models (LLMs), and autonomous computer agents to augment employees to facilitate the flow of data across human-in-the-loop tasks in service, sales, marketing, and commerce. By adopting Agentforce, Salesforce is promising to enable companies to scale their workforce, improve the flow of data, and create more contextual interactions. Agentforce leverages all aspects of the Salesforce suite of offerings and is best used by clients who are doing the same.

Unlike many keynotes, Salesforce walked its attending customers, partners, and analysts through a comprehensive view of how Agentforce, alongside its many other products, works inside Saks (a North American retailer). It demonstrated how Saks is going live with its agentic program to boost customer experience, productivity, and profitability. By spotlighting a single client’s holistic needs, Salesforce clearly articulated how GenAI can be scaled across the client’s operations, people, and customer service. The solution leverages the firm’s private data (Data Cloud), desired brand experience (Customer 360), insights (Einstein), and collaboration tools (Slack) to showcase a vision of the possible.

While questions remain and examples are just emerging of what it will take the average Salesforce customer to apply this new solution to their application stack, the potential is there. Agentforce can boost customer experience by pulling critical customer data from across systems and data repositories and using tools from Slack to robot call center agents to respond and action engagements in a contextual, meaningful manner. As agents can extrapolate from all the systems Salesforce touches, they can also prompt the human in the loop to add the final, personal touch to the experience, all without resetting the engagement.

Agents will be vital to scaling GenAI across an enterprise

Agents are expected to have a significant impact on the business. They will impact culture, change management, and how people work. In a recent study examining whether agents would affect jobs, HFS called out agents and their pending impact on systems and people, for better or worse. The jury is still out on the future of the workforce; however, with what we saw from Salesforce, we expect many clients and their advisors to delve into just how these solutions augment, enhance, or replace tasks done traditionally by people.

Nonetheless, the power of agents to leverage data, applications, business processes, and talent to enable collaboration, product development, and customer experience is clearer than ever. Like Salesforce, other software vendors are crafting a lot of innovation in the GenAI area; however, for enterprises to get the most from these efforts, they’ll need to work with their services and advisory partners to implement these solutions at scale. These support firms will be essential in mapping these solutions’ technological and human impact. With 44% of companies, in a recent HFS study, showing genuine concern over the impact of AI on their workforce, firms must work with partners to keep their humans happy.

Agentforce deepens the Salesforce moat, but it doesn’t replace the need for companies to build their agentic solutions

Make no bones about it: Salesforce has deepened its moat with Agentforce. For clients using Salesforce products, the pressure is on to adopt the vendor’s complete suite and Data Cloud to gain all the advantages it showcased. Salesforce’s Agentforce will work well across its own product suite, but it isn’t clear how well it will integrate with other record systems in large firms. While its Data Cloud and Mulesoft (API) can facilitate data management and exchange, it’s unlikely other software vendors will stand idly by as Salesforce corners the client experience market.

Building your agent versus buying is also part of many enterprise budgeting conversations. While Salesforce runs on its own cloud, many other agents run locally or are cloud agnostic, enabling firms to access data and systems where and how they prefer. Coding is still part of Agentforce. While Salesforce showcased how Agentforce can be designed, tweaked, and optimized using its Salesforce Platform (a low-code offering), others have gone straight to a no-code model, making solution development events more straightforward. From a cost perspective, clients with potentially lower development and run-time costs can facilitate the build agents.

Cost questions remain

Throughout the Salesforce keynote, one thing was clearly missing. There was no mention of the licensing models or costs associated with the compute, tokens, or integration of Agentforce across the vendor’s solution or the metered costs related to using Agentforce in production.

To design, deploy, and maintain a complete Salesforce ecosystem of products and have Agentforce running across all of them, usage will grow, and firms must budget accordingly. Salesforce must clarify how it will charge for these solutions. In some cases, its partners and Salesforce may need to evolve the traditional SaaS licensing model from fixed per-seat or usage pricing into more of an outcome-based model. After all, shouldn’t these agents be held accountable for their contribution, just as their human counterparts are?

Note: In a follow-up session, Salesforce shared that its agents may cost $2/session, but they made it clear that’s the list price, and big clients will have customized pricing.

The Bottom Line: Salesforce demonstrated what it takes to scale AI across the enterprise.

By going all in with AI and agents across its products, Salesforce showed how its clients can do the same. The demonstration by HBC-owned Saks showed how Agentforce can span its systems and teams to provide enhanced capabilities across product data, operations, and customer experience. While the costs of running Agentforce haven’t been disclosed, the opportunities for Salesforce’s millions of clients, IT services, and integration partners should drive a lot of interest. Finally, we expect to see significant scaling of GenAI across large enterprises’ IT and business systems.