For months, the Agentforce adoption debate has revolved around numbers that are pretty easy to misinterpret. Deal counts can include pilots, for example, and token volume can be skewed by a handful of heavy users. Even Annual Recurring Revenue (ARR), on its own, doesn’t always tell you whether customers are truly scaling or simply testing the waters.
That’s why one Q4 FY26 stat stands out – more than 60% of Agentforce and Data 360 bookings came from existing customer expansion. In other words, most of the quarter’s momentum came from customers who already know Salesforce and decided to spend more.
Expansion is a much stronger signal than first-time deals. Companies can experiment with a pilot, but naturally, they only usually increase spend if a product is delivering measurable value. The question now isn’t just who is buying Agentforce, but who is doubling down, and what that says about real adoption.
Why Expansion Changes the Agentforce Narrative
Over the past year, we at SF Ben have repeatedly asked whether Agentforce adoption was as strong as the headline numbers suggested.
Back in December – when 18,500 deals and 3.2 trillion tokens were announced – we considered this pretty impressive, given the skepticism from a lot of businesses we had seen that year. But we also asked the harder question – did this represent real, genuine adoption in the field, or just better packaging and financial storytelling?
At the time, penetration told a more cautious story. Roughly 12% of Salesforce’s 150,000 customers were on an Agentforce deal, and around 6% were on paid plans. As we noted, even if token counts were high, most Salesforce customers are not actually using Agentforce. The narrative looked like deep embedding among a minority rather than a broad ecosystem default.
That’s why, at the start of 2026, we predicted that Salesforce didn’t need explosive Agentforce growth this year – just something credible.
As Salesforce MVP Andy Utkan told SF Ben: “I would expect even a mediocre adoption of Agentforce and some of the newer technologies to do wonders for the investors.”
This argument outlined that Salesforce didn’t need a ‘hockey-stick’ moment, just steady proof that Agentforce wasn’t stalling. “Steady adoption,” we wrote, “could improve Salesforce’s revenue and investor confidence, which is what the company arguably needs right now.”
But adoption wasn’t just a numbers problem, but also a product experience issue. Early last year, implementation experts described Agentforce as “very buggy”, prone to losing context, fragile across sandboxes, and lacking clear documentation.
Pricing was called “a gray box”, the original $2-per-conversation model was described as “a complete non-starter” for smaller companies, and many buyers wanted to see real agents operating in production before committing.
There was also a persistent myth that you needed the perfect org, with technical debt, messy data, and legacy architecture all blocking the agentic AI push for many. As Senior Salesforce Consultant Joseph Monroe bluntly told SF Ben: “I’ve never seen one have perfect data. I don’t think I will ever see it.”
In other words, Salesforce wasn’t just battling skepticism about AI hype, but also on-the-ground issues like readiness, pricing clarity, and operational maturity.
Against these ongoing issues, the recent Q4 statistic that more than 60% of Agentforce and Dara 360 bookings came from existing customer expansion takes on a different weight, and that many customers are finally moving away from the experimental/curiosity phase. Plenty of these customers would have navigated the roadblocks mentioned – pricing confusion, documentation gaps, etc. – and still decided to increase their commitment.
It doesn’t mean that penetration is suddenly universal, nor does it mean every org is production-ready. But it does suggest that at least some early adopters have moved beyond “proof of concept” into repeatable value. And that is precisely the kind of “mediocre but credible” adoption curve we argued would matter most.
Final Thoughts
Agentforce doesn’t need to be everywhere yet – it just needs to be sticky where it lands. If more than half of bookings are coming from customers doubling down, that hints at something more durable than headline deal growth.
After a year of roadblocks, skepticism, pricing shifts, and product maturation, expansion may be the clearest sign yet that Agentforce is starting to earn its place in the enterprise stack.
