Back in June, we openly agreed on a stance that seems obvious but needed saying anyway: not everything needs an AI agent. My colleague Mariel Domingo explained that yes, Salesforce would keep pushing Agentforce use cases and shiny demos, but it would also become clearer what actually required AI automation versus what didn’t. This has certainly come true.
In the wake of this year’s Dreamforce conference, new products, and yet more (questionable) renames, the community appears to largely agree with one particular sentiment. So, Salesforce: do we really need AI for that?
Full Steam Ahead Into AI
With innovation comes a lot of things, and in Salesforce’s case, it particularly comes with expansiveness. Ever since the release of Agentforce in 2024, the company has shifted its messaging, pivoted its product strategy, and stormed full steam ahead into the world of AI.
It’s something that Co-Founder Parker Harris said is necessary to avoid “becoming obsolete”, and something impossible to ignore from Salesforce’s sales pitches, events, and new releases. With every new AI product or tool – or renamed product or tool – the push for adoption grows stronger. Suddenly, AI is the solution not only for the menial tasks it was designed to tackle, but for entire sales, content, and marketing processes.
The importance of discerning when and when not to use AI in everyday workflows is more important than ever. Sure, allocating AI resources over human resources into every task you can think of might look appealing to stakeholders, and it certainly does wonders for Salesforce’s own adoption numbers.
However, as time goes on, it becomes more apparent just how exaggerated some of Salesforce’s AI use cases may be, despite the company’s own guides to help professionals decide when it is best to use it.
‘90% of the Uses I’ve Seen This Week Don’t Need AI’
For many, this year’s Dreamforce conference was another reminder of just how intense Salesforce’s AI push is. Matt Pieper, the Founder of Right Sized Tech, took to LinkedIn amid the conference to share his thoughts on the matter, saying he believed 90% of the use cases he saw that week “don’t need AI.”
“As I’ve talked to people this week, it’s apparent that a lot of AI ‘use cases’ are being made because people just don’t know what’s possible today,” he wrote. “The majority of users of gAI/LLM/Harnesses/etc. are barely scratching the surface and are using an expensive Google.”
Although he shared his appreciation for one of Salesforce’s latest offerings, Claudeforce, he maintained that many of the issues he had personally heard about could be adequately solved with core functionality or without the use of AI.
Many agreed with Matt’s sentiments, with Paul Battisson, the Founder and CEO of Groundwork Apps, responding that many of the “use cases” he kept hearing about could be solved with solutions like reports and dashboards, better configured page layouts, or a configured path with some guidance.
‘We Are Blatantly and Openly Choosing Profits Over People’
Matt Gerry, Salesforce CTA and content creator, also shared his thoughts on Salesforce’s AI push, admitting that Dreamforce had “left him sad.”
“Every single session and keynote I went to left me somewhat disappointed in the direction we’re choosing to focus on and lean into in tech,” he wrote. “We are blatantly and openly choosing profits over people in a way we have yet to experience (despite it being somewhat of the norm these days).
“AI could be used for good things it could, but with rare exception that was not the focus of Dreamforce. Its focus in nearly everything I saw was choosing profits over people, choosing to enrich the already unbelievably wealthy individuals of this world as opposed to giving an opportunity to those less fortunate.”
He emphasized how he felt AI had created a disconnect between supporting communities and creating use cases for mega corporations, when the focus should always be put on the human side of work. This is not the first time Salesforce’s marketing and messaging have been openly criticized, and the company ran into particularly hot water after CEO Marc Benioff made an ICE “joke” at an internal event.
“I was there too and I would agree with your analysis,” Fabrice Talbot, the Founder of AI Practitioner Circle, responded. “However, nothing new there, Salesforce is a show, I respect that, and you should not expect anything else.”
When AI Has A Place, and When It Doesn’t
Understanding when to use AI versus when not to use it is one thing, but understanding what a valid use case is versus just heavy marketing is another.
Mariel says she believes AI certainly does have its place in making messy and unstructured work easier to navigate, but that it isn’t as simple as that.
“I think it’s awkward that I wrote that article back in June when Salesforce released their framework, which basically said that using an agent is the ‘anti-pattern’ for processes you can actually map out,” she said. “This year’s Dreamforce made AI and agents sound like the answer to everything!”
“At the end of the day, it’s as simple as asking yourself why a step needs reasoning in the first place. If the question cannot be answered in one sentence, it probably doesn’t.”
Final Thoughts
It is very likely that even Salesforce understands not every workflow or process or even business problem needs AI. A lot of its marketing is centered around how useful and wonderful AI can be, but it has to. However, that doesn’t mean it isn’t up to the discretion of professionals to decide whether they need to use AI for a certain task or not.
Salesforce will keep pushing you AI tools, products, and more. Explore what you’re interested. Use what you can. Take everything else with a pinch of salt.
