Salesforce’s announcement of its revamped AI pricing model has sparked considerable buzz throughout the ecosystem. With the introduction of Flex Credits, action-based billing, internal-use licenses, and the innovative Flex Agreement, Salesforce is clearly signaling its intent to align more closely with customers’ evolving needs and budgets.
Early reactions from Salesforce professionals are seemingly positive, praising the new model’s transparency, flexibility, and accessibility. Yet amid the optimism, some voices have highlighted issues with complexity, fairness, and administrative concerns. In this article, we’ll delve into these contrasting reactions, unpacking the positives and analyzing the valid critiques.
The Positives of Agentforce’s Pricing Model
While no pricing model is without its critics, the initial reception of Salesforce’s revamped approach has been notably positive. Experts from across the ecosystem are highlighting increased flexibility, clearer budgeting, and strategic alignment as key advantages.
“The Bridge Too Far Just Got a Little Shorter”
Anna Shaffer, Chief Product Officer at b atomic!, described these changes as hugely beneficial, especially to those who are looking to get started with Agentforce.
“Honestly, it’s a huge improvement. The old model always felt like a black box – hard to predict, hard to explain to stakeholders,” She explained. “Flex Credits make it easier, especially for ISVs and midsize teams who want to pilot AI without committing to a massive license shift. It lowers the barrier to entry.”
Echoing this enthusiasm, Sam Adiv, Founder and CEO of OpenTent, said: “Flex Credits” are a “pricing innovation” to emphasize action/outcomes rather than conversation/transaction. And they are way cheaper to get started with ($500 for 100K actions vs. the previous $2/conversation). I love to see it – this will help more orgs jump into playing around with these tools. I bet it will get even cheaper next year.”
Reducing these barriers has resonated with many – it feels as though Salesforce have really listened to the ecosystem and acted accordingly to address this issue.
In April this year, we covered why Agentforce adoption had been slow up to that point, and spoke to Rob Thomason and Daniel Rudman of Gruve, among other industry experts, to delve deeper. At the time, there was still much confusion around the pricing model, with Rob describing the $2 per conversation model as “messy”. Fast forward to now, and this messy model has now been given some more clarity.
Following up with Daniel, he said: “The changes around Flex licenses you can transfer between seat and consumption, as well as licenses for unlimited internal usage, were a big surprise. This means Salesforce is hearing, as we all have, that cost is one of the major impediments to the wider adoption of Agentforce and Data Cloud. This should help accelerate adoption.”
Rob furthered this sentiment, and appreciates that Salesforce are showing new market behaviors to ensure Agentforce is cheaper and more accessible.
“It is encouraging to see Salesforce adjust their pricing strategy to what is happening in the marketplace. It is certainly not the norm for Salesforce to discount offerings. They just don’t do it ever. The ‘bridge too far’ just got a little shorter for some folks.” Rob Thomason, Senior Salesforce Solution Architect, Gruve
“It’s Far Simpler to Directly Correlate Cost and ROI”
Many respondents have also praised Salesforce for delivering greater clarity around costs and usage. A common frustration with previous AI pricing was its opacity, which made it difficult for businesses to clearly forecast expenses and calculate returns. Under the new model, some Salesforce professionals are optimistic about improved predictability and budgeting transparency.
Joseph Monroe, Senior Salesforce Consultant at Blue Gator, said: “It makes it far simpler to directly correlate cost and ROI. There’s no more intricate math around credits, conversations, Einstein requests, etc. No matter how long the conversation, cost is always associated with actions taken. For simple use cases, this opens the door to lots of opportunity for customers that may have been leery to try it out, or use it with simple use cases that might have been cost-prohibitive [beforehand].”
Similarly, Anna pointed out that clearer tier structures and the transparency provided by Flex Credits empower teams to plan proactively. “This shift toward clearer tiers and Flex Credit transparency is a step in the right direction,” she explained. “For teams with well-mapped processes, it’s empowering.”
Keir Bowden, Former CTO of Salesforce UK at Credera and Salesforce MVP, expanded on this point, appreciating how the revised pricing model not only improves predictability but also directly connects costs to the value delivered by Agentforce.
“The new charging model looks to be more predictable than the previous conversation-based charging, and is tied to adding value,” he said. “You get charged if Agentforce executes an action, which means it’s helping the customer even if it can’t solve the problem entirely on its own.”
“I really like the per-user per-month pricing option for unlimited internal use with the Agentforce licensing. CIOs don’t want their staff worrying about how much it will cost to make use of their AI assistant, but they also don’t want surprises at the end of the month. I think this change on its own will kick-start some projects.”
“Salesforce Is Anticipating Workforces Looking Very Different, Very Quickly”
Beyond practical improvements, some industry experts are recognizing something deeper at play with Salesforce’s pricing shift: a clearer alignment with the company’s broader strategic vision.
Salesforce professionals are identifying this new pricing model – and the Flex Agreement especially – as more of a deliberate move towards a future in which digital and human labor begin to blend, responding proactively to the changing dynamics found in workplaces.
Sam highlighted the significance of Salesforce’s proactive stance and how exciting it is for the future of Salesforce.
“Salesforce is anticipating workforces to start to look very different, very quickly, pointing to a ‘blended’ future with both human and digital labor working together, and resources shifting between the two as leaders figure out their processes and approach. This has been Salesforce’s stated goal for a year now, and the pricing reflects that vision. I personally think that vision is powerful and exciting.” Sam Adiv, Founder and CEO, OpenTent
Building on this idea, Daniel emphasizes the practical reality underpinning Salesforce’s strategy, pointing out that the flexible allocation of resources mirrors the growing role of digital labor. “Digital labor is real, and the Flex Agreement provides the capacity for workers and flexibility to shift things around,” he explained.
Finally, Keir frames this strategic shift in terms of reducing risk. The flexibility to switch investments between digital and traditional roles allows businesses to confidently embrace AI-driven transformation.
“The Flex Agreement, which allows organizations to switch investment between per-seat licenses and Flex credits, also feels like a good move to me,” he said. “This not only removes some of the risk around betting big on AI agents, as if it was a little premature, you can reallocate investment back to the traditional roles, but also allows Agentforce to be rolled out using existing agreed spend rather than having to ask for more money.”
Taken together, these viewpoints reflect a clear recognition that Salesforce’s updated pricing not only addresses practical concerns but actively supports its long-term vision of seamlessly integrating AI into the evolving workforce.
The Concerns Around Agentforce’s Pricing Model
Enthusiasm is widespread, but there are still many existing concerns that have risen from this new pricing model. Questions still remain around the complexity of budgeting for consumption-based pricing, the fairness of billing around actions, and the potential administrative overhead it could create.
“It Is Still Consumption-Based, So It’s Difficult to Forecast”
One of the most prominent reservations voiced by industry experts centers around the practical implications of the new pricing model. Whilst it may be clear in theory, there’s concern that tracking numerous micro-actions could create administrative headaches and budget uncertainties.
When we first discussed why Agentforce adoption was slow, Founder and CEO of Elements.Cloud, Ian Gotts expressed strong concerns around the $2 per conversation pricing model, saying: “They don’t know how to price them yet, and right now, nobody wants to hand out a blank cheque – they want caps and predictable ROI.”
And despite the changes made to the pricing model by Salesforce, Ian identified several flaws that could continue to make it more difficult for businesses to recognize their ROI.
“It is still consumption-based, but with a different metric – so it still is difficult to forecast. The cost is based on how many agents are triggered, provided the action is less than 10K tokens. So every action over 10K tokens is considered an additional action.”
“It puts the emphasis on designing efficient agents. A poorly designed agent could call lots of actions to get to a result.”
Steve Snell, Chief Executive Officer at Attentis Consulting, used an interesting analogy to another technological pricing scheme we saw in the 1990s. “This pricing schema reminds me of my 1991 cell phone data plan,” he explained, “where every month’s bill was an adventure into the great unknown.”
When cell phone data plans were first introduced in the early 1990s, they proved notoriously opaque and complex – people didn’t understand how their usage translated into their monthly bills, and each month would bring surprising, unpredictable charges because pricing wasn’t clearly linked to understandable measures.
By comparing Salesforce’s new AI pricing model to this cell phone billing experience, Steve suggests we could be seeing more of the same confusion and uncertainty going forward.
If you’re unsure of a real-world application, Colm Barry, Co-Founder of capeMBX, believes companies that deal with large numbers of digital interactions or transactions every day may struggle.
“As agents become more prevalent, tracking millions of micro-actions could be a headache for IT and finance, especially in industries with high volumes like financial services or travel.”
Summing up this issue, Rob pointed out that while the new system might simplify budgeting for some, it remains intentionally ambiguous – perhaps reflecting Salesforce’s ongoing struggle to balance flexibility with clarity.
“It’s easier for some people, but still confusing to others,” Rob detailed. “This has been the barrier all along: that the pricing model and what you need to buy is blurry. It may be by design.”
“It’s a Bureaucratic Lock-In”
Moving beyond logistics, another central question raised by respondents revolves around fairness. Is charging based on individual actions truly aligned with delivering measurable business value, or could this method risk penalizing users for necessary but minor tasks?
Daniel questions whether the action-based model genuinely captures business value, saying: “I actually don’t think it is the most accurate measure of value. Although I can understand why they went in this direction, other measures might not be currently practical. This might be the least bad option.”
“Taking an action on the platform – workflows, updating a record, etc. – is not necessarily a value to the business, and not all actions provide equal value. For example, updating a record to escalate it to a human or sending an email, which could be automated for free.”
“The true value in my mind is doing something a human would have taken time to do instead. Measuring this, though, is very tricky. Measuring the impact of an action is likewise really hard.”
Rob echoed this skepticism, saying: “I don’t necessarily think it’s fair, but I can see why they went in this direction.”
Offering a sharper critique, Ian Smith, Design Thinker at Being Minded, described the new pricing model as “nonsense on legs”, and highlighted other concerns around whether Salesforce’s approach stacks up against alternative agentic AI offerings.
“ROI does not compute against alternative overlay AI options for Salesforce. It’s a bureaucratic lock-in. AI is so much more innovative than what you get from Salesforce!”
So while Salesforce’s action-based pricing aims for clarity, there’s still underlying tension around aligning charges with genuine, tangible value.
Remaining Questions and Key Considerations
Reactions to Salesforce’s new pricing model have broadly settled into cautious optimism, but several key questions remain unanswered.
Respondents also highlighted specific areas where further clarity from Salesforce would significantly influence adoption decisions, budgeting strategies, and long-term confidence in Salesforce’s flagship product.
Tooling and Spend Management
A key area prompting further questions from Salesforce professionals is the capability of Salesforce’s tools for tracking and managing AI-related spending. For many, the effectiveness of these tools could be critical in determining whether the new pricing actually simplifies cost management or adds unexpected complexities.
Daniel shared some concerns surrounding Salesforce’s Digital Wallet. “How powerful will the spend management tools in Digital Wallet be?” Daniel queried. “This can influence how closely orgs can track and predict their spend, which adds to the impact of a changed license.”
While Salesforce has taken meaningful steps towards cost transparency, the strength and reliability of its spend-management tools remain an important unanswered question.
Sandbox and Internal Seat Pricing
Salesforce professionals are also awaiting clarification around pricing of sandboxes and internal-use licenses. Given that these costs will directly impact organizations’ ability to pilot and scale AI solutions, unresolved pricing concerns could impact the attractiveness of Salesforce’s offering.
Daniel mentioned that Sandbox pricing “still seems high” for users, and that pricing for the internal seat is “unclear”. Rob expanded on this by looking closely at Salesforce’s internal incentives and the strategic impact of Sandbox costs.
“Sandboxes still seem expensive,” Rob said. “I understand that sandboxes and Dreamforce have traditionally been Salesforce’s biggest profit generators, but they might need to reconsider this pricing strategy to better align with current market conditions. Right now, businesses are cautious with spending, putting many projects on hold – and Agentforce is just one example.”
Sam also added to this sense of uncertainty, highlighting the absence of detailed internal-use pricing, despite Agentforce’s upsides.
“Despite no pricing details yet… I think this will likely be a worthwhile add-on for many of our clients.”
It’s clear that many feel clearer pricing structures around sandboxes and internal licenses could influence more uptake among potential users going forward.
Long-Term Pricing Evolution
Looking further ahead, Salesforce professionals are also curious about what Salesforce’s new Agentforce pricing model might signal for its broader long-term pricing strategy. Could this be the start of a larger shift toward consumption-based or pay-as-you-go pricing across other products in the Salesforce portfolio?
Mike Hanson, Chief Executive Officer of Zephryus, shared a prediction, saying: “[It’s] perhaps a move toward a larger company-wide pricing model change? I’d like to see more pay-to-play pricing from Salesforce so that clients can more directly see ‘my money turns into X’.”
Keir similarly emphasizes the evolutionary nature of Salesforce’s current approach, cautioning that further adjustments to consumption-based pricing are likely.
“Consumption-based charging is fairly new to Salesforce as well as customers, so I doubt this is the last we’ve heard on pricing!” Keir Bowden, Salesforce MVP
Final Thoughts
The new pricing model for Agentforce is an exciting next step for Salesforce’s innovative AI tool, and it truly feels as though Salesforce has taken constructive feedback on board to help users adapt to it as comfortably as possible.
There are still some outstanding flaws and questions that need answering, but I think Keir hit the nail on the head.
“The new approach isn’t perfect, but I’m not sure perfect exists in this space,” he said. “Salesforce has such a range of customers that finding one size to fit all is quite a big ask. I do think the C-Suite can be confident that the investment they authorize will more closely match the costs they incur, so this is likely to help the uptake of Agentforce.”
What are your thoughts on the new Agentforce pricing model? Be sure to leave your thoughts in the comments, and keep the conversation going over on LinkedIn!