Dreamforce 2026 Recap
Andrew Campbell:
Each year, Dreamforce gives me a fresh lens on the work we do in Salesforce. This year, the theme that followed me from session to session wasn't a product (yes, AI was in everything) but a simple question: Always ask “why.”
Before you build a feature, flip on an AI tool, or roll out a new process, what outcome are you actually trying to reach?
In my favorite session of the week, ‘Architect for Outcomes, Not Requirements’ a comparison was made between a project and a cathedral. There are parts of this cathedral that are ornate and attractive; The spire is the compelling business story. Your elaborate configuration and automation make up the beautifully architected spires and vaulted ceilings. But, at the very base of this cathedral is the foundation. That foundation represents the fundamental business outcome you are trying to solve for, on which the entire structure is made. This is your “why.”
That same question kept showing up. In an ‘AI-For-Fundraising’ session, the advice was clear: if you can't explain why you want an AI tool, pause and ask the question. The Atlanta Hawks reminded us that if Users don't trust the CRM, they'll find a way to work without it. Their member-style Account view got me thinking about how our ticketing clients could use Salesforce to elevate the in-venue experience for season ticket holders, whether that's a seat upgrade or a surprise piece of swag. AC Milan showed how six different channels feed a single fan view, and I left wondering where we can help you fill in the gaps.
This year, my goal is to lead with outcomes before tools. Whether it's an agent, a new integration, or a simple flow, every build should start with a clear why and a way to measure it. The payoff is fewer projects that just look good on launch day, and more that actually move the needle for your patrons and donors.
Chris Carney:
Dreamforce 2026 was my 8th time attending Salesforce’s flagship event, and it did not disappoint. If you’ve been paying attention to any Salesforce announcements recently, it is no surprise that AI was the heavy focus of the conference. In a big shift from last year, Salesforce has opened their platforms to all of the various AI tools that you may be using today. While Agentforce (Salesforce’s proprietary AI tool) still has its place, it is now easier to directly access your Salesforce data from any AI tool through standard (or custom) MCP servers. Claudeforce was a major announcement, with the CEO of Anthropic joining Marc Benioff for the keynote session.
What made all the AI content different this year is it was less theoretical and much more applicable. The things I saw people present on with regards to AI are things I am either doing today or will be doing over the next year. Two years ago, AI was something that was looming on the horizon. Last year, AI was here but still had a pretty high barrier to entry for the everyday user. Today, AI is becoming a core tool in the Salesforce admin’s tool belt. Learning how to utilize that tool in a secure, responsible way and deliver accurate, consistent, and predictable results was one of my biggest takeaways.
From a career growth perspective, AI can be a scary concept. There are absolutely admin functions that I perform today that can be fully offloaded to AI in the imminent future. However, the human input will remain a critical component to any successful AI implementation. I can see a future where the Pac Admin team becomes more of a Pac Architect team, spending more time on consulting and strategy to drive success and less time in the weeds adjusting layouts and building fields. It is an exciting time to be involved with Salesforce, and I am looking forward to implementing some of my takeaways with the Pac Community over the next year!
Eugene Edwards:
Each year, Dreamforce gives you plenty of new technology to think about. AI was
everywhere this year, but the idea that stuck with me most wasn't really about technology. It was about how we approach problems before we ever start building.
One of my biggest takeaways was simple: the best Salesforce solution isn't necessarily the one with the most automation. It's the one that helps someone make the right decision and take the right action with the least unnecessary friction.
That was less of a revelation for me and more of a reinforcement — and maybe a challenge. Like a lot of admins, I can get focused on delivering. Someone needs a Flow, dashboard, field, or new process, and my instinct is to figure out how to make it work. But several sessions reinforced the importance of first understanding the outcome behind the request. What decision are we trying to improve? What behavior needs to change? What does success actually look like?
One UX session made a comparison that immediately clicked for me as a former football player: players who have to think too much don't play fast. The same thing happens to our users.
If a sales rep has to navigate several screens, interpret too many fields, remember multiple process steps, and piece together information before deciding what to do next, we've created cognitive load. The system may technically work, but we've made it harder for that person to act. Good UX isn't just about making Salesforce look better. It's about making the next decision clearer.
Another session unexpectedly connected this to my military experience. The discussion borrowed concepts similar to military planning: understand the mission, identify constraints, evaluate different courses of action and their tradeoffs, then make a decision. Sometimes our greatest value isn't immediately returning with a solution. It's helping stakeholders slow the decision down just enough to make the right one — so the organization can move faster afterward.
That thinking also applies to scale. Solving a problem once is useful. Recognizing that the same business problem exists across multiple organizations and designing something reusable is much more powerful.
And yes, AI was everywhere. But one of my bigger takeaways wasn't that every problem needs an agent. In some ways, it was the opposite. AI makes the fundamentals even more important. If our data isn't trusted, permissions aren't intentional, processes aren't understood, or we haven't clearly defined what we're trying to accomplish, adding AI doesn't eliminate those problems. It can amplify them.
So I left Dreamforce excited about AI and where the Salesforce platform is going — but even more interested in decision-making, UX, data, architecture, and outcomes.
Don't start with What can we build?
Start with What are we trying to accomplish? What decision needs to become easier? What is getting in the way today?
Then use Salesforce, automation, analytics, AI — or sometimes something much simpler — to remove that friction.
Because ultimately, the technology isn't the outcome. What people are able to do because of it is.
Harrison Gore: At Dreamforce this year, one of the biggest recurring themes was the shift from simply collecting customer data to creating the context needed to act on it. Sessions from organizations like Tottenham Hotspur showed how unified data can evolve into a true “Fan 360,” connecting ticketing, marketing, service, engagement, and other interactions to create a more complete understanding of the fan. Salesforce’s broader Data 360 vision builds on that idea by making data available wherever work actually happens rather than requiring users to constantly move between systems. That same philosophy carried into Slack, where new concepts like Surfaces and deeper Salesforce integration point toward a future where users can access records, take action, collaborate, and work with agents directly within their existing workflows. My favorite session of the conference, McDonald’s World Cup CRM session, brought this concept to life particularly well. Instead of relying solely on a traditional marketing calendar, McDonald’s demonstrated how using live event data and real-world moments can provide the context that drives the next customer interaction. For sports organizations especially, that creates an exciting opportunity to think beyond scheduled campaigns and build CRM strategies that can react to what is actually happening on the field and using game changing plays to capitalize on fan excitement.
Another major takeaway was how quickly the roles of Salesforce admins, developers, and architects are beginning to overlap as AI lowers the technical barrier to building solutions. Several sessions reinforced that AI can help generate Flows, Apex, interfaces, and agents, but being able to build something faster does not eliminate the need to understand whether it should be built that way in the first place. Salesforce increasingly appears to be moving toward reusable business capabilities that can be surfaced through Flow, Slack, agents, portals, MCP, or other applications, making architecture, governance, security, and system design increasingly important skills. Tools like Flow Optimizer reinforce that focus by helping teams identify technical debt and automation issues at an organizational level, while new Flow capabilities continue turning declarative tools into increasingly powerful application-building platforms. One session described the evolution from the traditional “accidental admin” to the “accidental architect,” which captured the conference particularly well for me. AI is making execution easier, but that makes human judgment, technical understanding, and the ability to connect business problems to scalable architecture even more valuable. My biggest takeaway from Dreamforce this year is that the future of Salesforce is not simply about adding more AI. It is about combining unified data, meaningful context, reusable capabilities, deterministic automation, AI reasoning, and human judgment to create systems that understand not only what happened, but what should happen next.
