The companies leading in AI have already left email behind. They are building the future of work on shared context, not private inboxes, says Slack’s chief marketing officer, Ryan Gavin.
The companies advancing fastest in AI transformation have one thing in common: They realized early that you cannot transform the way employees work on an email-era foundation.
I talk to enterprise leaders every week about what true work transformation demands in this AI-first era. Two clear patterns are emerging across industries, and it is becoming impossible to ignore.
The first pattern comes from companies that still run on email. Whether they intend to or not, the inbox functions as the operating system for employees. Decisions play out in reply-all threads. Urgent requests stack up. Outside of meetings, the inbox becomes the center of gravity for getting work done. When someone leaves, their institutional knowledge often vanishes with them. Tools for chat, project management, and automation are added, yet the gravitational center never shifts. Everything collapses back into email. In these environments, the benefits of AI show up at the margins, such as faster document creation or small automations, but the fundamental work practices remain unchanged.
The second pattern is modeled by companies that are often AI-native: Cursor, OpenAI, Anthropic, Vercel, Box. At these organizations, email barely gets used. The inbox is a lightweight channel for external correspondence, checked only when needed. The real work happens in open, shared environments where conversations, data, systems, AI, and agents operate in one connected flow. Employees do not use AI to accelerate the old way of working— they use it to redesign how work actually happens. Workflows are agent-first. Processes are continuously reimagined. People shift from doing tasks to orchestrating systems that have full context of their roles, their teams, and their priorities. At these companies, one employee can easily operate at the scale of 10 in a traditional enterprise.
“You start with Slack as a tool. Then you realize it’s how you run the company.” —Philip Hess, CEO, reMarkable
The distance between these two operating models is massive and widening every week. For many enterprises, the barrier lies not in the adoption of AI. It’s unlearning 30 years of email-era habits that have quietly, and, increasingly, become a liability. The biggest barrier for unlocking the productivity promises of AI for your people is, in fact, the old way of working.
Email has been “good enough,” but now it’s a liability
A chief information officer (CIO) told me recently that the board had given a clear mandate: Get more return from the company’s single largest expense, its people. For most enterprises, labor and its productivity levels are not just a line item. This correlation is why today’s most successful firms tend to lead in revenue generated per employee. This CIO had funded dozens of AI pilots, many technically successful, yet employee productivity was only marginally better—nowhere near the step-change needed, nowhere close to the expectation that AI could double or triple output with the right foundation. The uncomfortable truth was that the models were fine. The foundation they were sitting on, the foundation for how the company actually worked, was not.
I hear versions of this story every week. Systems of record are organized. Customer relationship management (CRM) systems are connected. Email archives are indexed. But the reality of how work gets done lives in places AI cannot see: private threads, impromptu chat debates, evolving documents, and quick decisions made across work groups and channels. By the time anything reaches a CRM or a recap email, the context that gives it meaning is already gone. The AI is not underperforming. It is operating blind.
With AI, that no longer works.
AI needs complete context: what is happening, why it matters, how it evolved, and what should happen next. It cannot reconstruct intent from siloed inboxes or a patchwork of disconnected chats. It cannot learn from work it cannot observe. When employees rely on email as their operating system, AI gets pushed to the margins, offering small conveniences instead of transformational productivity gains.
To cope with email’s limitations, companies layered more tools on top: chat for speed, video for alignment, docs for collaboration, project trackers for accountability. Each helped in isolation, but together they created a fractured digital workplace. The average worker now loses nearly four hours a week just regaining context after switching between apps. That is the toggle tax. And it only grows as more AI is added to the mix.
So, when organizations attempt to layer AI on top of an email-centric foundation, they get exactly what that structure enables:
• AI that needs everything overexplained
• AI that hallucinates
• AI that dead-ends
• AI that becomes another tool to manage instead of a multiplier for the workforce
Moving from systems of record to systems of action
Customers are often surprised when I tell them that since I joined Slack nearly two years ago, I send fewer than five emails a week. For leaders coming from traditional enterprises, it sounds impossible. But I learned quickly what AI-native companies already understand. Work does not happen in email. It happens in conversations. It happens in shared, open, interoperable environments where people, data, systems, and AI all see the same information at the same time and can act on it. At companies such as OpenAI, Rivian, Anthropic, and Vercel, Slack channels are the connective tissue between research, product, operations, and field teams. It is where they build. Human collaboration and AI agents coexist in the same conversations where decisions are made.
“In this age of AI, I am not simply thinking about innovating. I am trying to reimagine what businesses look like from the ground up!” —Prasad Swaminathan, Group SVP and Global Head of HR, Adecco
These companies are not simply adopting AI. They are shifting from email-driven systems of record to systems of action, where intelligent agents observe, learn, and act alongside teams. In these environments, AI is not bolted onto old workflows. It becomes part of how work is created, validated, and executed. Employees move from being the activity engine to being the orchestration engine.
Seeing customers such as OpenAI, Anthropic, Box, Writer, Vercel, and others build their businesses on Slack as the work operating system has convinced me that unlocking AI productivity for employees is not a technology problem. It is the challenge of overcoming the inertia of the old way of working. The companies seeing real returns from AI are not the ones with the biggest budgets or the most advanced models. They are the ones willing to reimagine how work happens, and that reimagination must start by moving work out of the inbox.
The future of work starts where AI has context
The leaders reimagining how work gets done are pulling away from the rest. They are treating their work platform as infrastructure, just as essential as CRMs or data platforms. It’s the 2% of their budget that maximizes the return on the other 98%. They are asking the harder questions: Where are decisions truly made in our company? Where does work actually happen? And can AI see any of it? The most innovative companies in the world now share the same understanding. AI cannot transform work if it only sees the record of what happened. It unlocks its full power when it can see the work itself.
“Agentforce is the interface that will turbocharge productivity across all our Salesforce interactions, including Slack.” —Aaron Levy, CEO, Box
This is why companies that operate in Slack see a fundamentally different return on their AI investments. Slack gives AI something email never can: complete conversational context. The full history of how decisions were made, the reasoning behind them, and the real-time signals that show what is happening, who is involved, where work is stuck, and what should happen next. AI can only act with confidence when it can see the context that humans have always relied on to understand work.
When AI works inside Slack, agents can recommend actions instead of searching for clues. They can automate workflows instead of waiting for instructions. They can resolve issues before they become problems. They can operate at the speed of the team, with the judgment of someone who has been in every conversation.
The companies leading this shift are already seeing employees operate at two or three times their previous output because they have removed the friction that slowed them down and given AI the visibility it needs to act. They did not achieve this by adding more AI tools. They achieved it by moving work out of the inbox and into an environment designed for intelligence, action, and shared context.
“I finally have a place where I can communicate with my entire organization to share experiences, wins, and learnings. Plus, we’ve seen massive productivity gains. We couldn’t be happier with our choice to implement Slack.” —Varun Krishna, CEO, Rocket Companies
The future of work will not be built in private threads and siloed inboxes. It will be built where people and AI—such as Agentforce—can work together in the open. And that future is already happening in Slack.
To explore how leading organizations are building their foundation for the AI era with Slack, visit Slack.com/customers.
Note: This content was created by Slack.