Every business function gets a dedicated tool where the core work happens. Designers once shared files by email until Figma gave them a shared canvas. Project managers managed teams in spreadsheets until Asana and Monday offered one place to plan and execute.
But sales never got their version.
Instead, the sales tech stack focused on logging and reporting on the work salespeople do elsewhere: recording emails and calls, tracking pipelines, and reporting outcomes. AI is changing that. As software moves from answering questions to taking action, the enterprise stack is reorganizing around a new layer: the System of Action. Sales is one of the first places where that shift is becoming impossible to ignore.
The Sales Stack Wasn’t Made to Execute the Actual Work
Legacy enterprise software was designed to capture work. In sales, that role fell to the CRM. It logs deals, tracks stages, and records what happened.
A rep enters a call, updates a stage, and adds notes. On top of that layer, tools like Gong and Clari analyze that data, identifying deals at risk and reps who need coaching. Both layers add value, but neither helps you do your actual work.
Look at what actually fills your day. You write follow-up emails, build and revise decks, prepare proposals, and decide who to loop in next. None of that shows up inside your CRM or the analytics tool. You do it in your inbox or maybe in some document, kind of jumping around between systems, all built to observe you while you work.
More than 78% of sellers missed their quotas last year working on stacks designed specifically to support them. The gap isn’t logging or analysis. Those are solved problems. It’s execution, and nothing in your stack was built to solve that.
Sales Exposed the Limits of a Stack Built to Log, Not Execute
Sales became the clearest proof point that the legacy enterprise software stack was built for a simpler era. That stack had two jobs: record what happened and surface insights about it. For decades, that was enough. It isn’t anymore.
The complexity of the actual work has surged across every department, but sales made the breakdown impossible to ignore. Buying committees that once ran through a single decision-maker now route through entire organizations. Forrester puts the average business buying decision at 13 internal stakeholders and nine additional external influencers. Gartner finds that buyers spend only 17% of their purchasing journey in direct contact with suppliers. The rest happens without the seller in the room.
That means the seller’s job is no longer just to pitch and follow up. It’s to coordinate across a sprawling, largely invisible process they don’t control. A system designed to log what happened after a meeting can’t help with that. And when AI shifted from answering questions to taking action, the gap became a chasm.
A stack built to record and analyze work is still valuable. But when the work itself gets dramatically harder, and AI opens up entirely new ways to support execution, we should be shooting for more than “log and learn.”
AI is Reorganizing the Enterprise Software Stack
Most people assume AI just makes existing software smarter. That’s not what’s happening. AI isn’t a feature added on top of your CRM or your analytics tools. It’s a new layer that can analyze information, recommend next steps, generate the content you need, and coordinate the steps of a workflow on its own.
That capability changes where value gets created. For years, owning the record was the advantage. Whoever controlled your customer data controlled the relationship, because everyone else had to go through that system to access it.
Agentic AI has altered that logic. Once software can connect to data wherever it lives and act on it directly, holding the record stops being the differentiator. The advantage moves to owning the surface where the work actually happens, because that surface generates the context that never makes it into any record, and it’s the only place an agent can act, not just analyze.
You already see this kind of thing in how the shift plays out, day to day. Like an agent listening to your call and drafting the follow-up before you even open your inbox. It also notes when a deal goes quiet for longer than it should, and then it suggests what to do next. You take a look at it and tweak it instead of starting from scratch each time.
Your CRM doesn’t disappear in this shift. Its role changes. It becomes infrastructure rather than the center of daily work. The workflows, custom fields, and administrative overhead that once kept data usable become less important when AI can interpret unstructured data directly and act on it. The record still matters, but it’s no longer where the greatest value gets created.
A Day Inside a System of Action
Here’s what this looks like in practice:
You start your day in one workspace. Not your inbox. Overnight, the deal moved without you: your buyer’s CFO went deep on the pricing page for the first time, and the system flagged it as a risk, with a suggested play and a drafted note reframing the business case. A stakeholder you’d never met joined the evaluation; they’re already mapped, and the follow-up queued for your review speaks to what their role cares about. Your champion needed something for an internal review, so the workspace assembled an executive summary from materials already living in the deal, and it is waiting for your review and approval when you wake up.
After your 10 a.m. call, the room updates itself: next steps captured, the mutual action plan adjusted, the recap drafted and waiting for your edits. A deal that’s been quiet for nine days surfaces, not as a dashboard alert, but with a reason and a recommended move you can approve or change. And on the other side of the table, your buyer’s committee gets answers the moment questions come up. At 11 p.m., without waiting on you, because the AI lives where they’re evaluating.
Notice what changed. You didn’t produce the work and then log it somewhere. You reviewed work already in motion, in the one place both sides of the deal actually operate. That’s the difference between software that watches you work and software that works with you.
Where enterprise software goes from here
This reorganization isn’t finished, and it won’t stop at sales. Every function in your business is about to face the same question: does this job need a dedicated system of action, or does general AI infrastructure already cover it?
The answer won’t be the same across your stack. Some work is generic enough that a broad AI layer handles it well. Other work is specific enough, with its own workflows, timing, and context, that it needs a system built for that job alone. Sales falls into the second category, because the work of selling happens in the space between meetings.
If you lead a revenue or technology organization, the task ahead is straightforward. Map the jobs your teams actually do. For each one, decide where a general tool is enough and where the work is specific enough to need its own system of action.
Gal Aga is the co-founder and CEO of Aligned, provider of the AI Deal Workspace where sellers, buyers, and their AI agents run deals together. A veteran B2B SaaS executive, he brings two decades of experience scaling sales organizations from $1M to $100M ARR to the future of collaborative buying.
