The Software the Executive Never Had

Thirty Minutes Before the Meeting

In 2010, I was running SugarCRM. The company existed to capture customer information. Every email logged, every call noted, every deal stage tracked, every contact record kept current. The system held it all. Before every important customer meeting, our most experienced salespeople still spent thirty minutes piecing together what was actually happening on the account.

They opened five tabs. They re-read the last email thread, scanned the deal history, looked at the latest internal back-and-forth in chat. Someone got the assistant on the phone to ask what had been said on Tuesday's call. They walked into the meeting somewhat oriented, never fully prepared.

The information had been captured. The context had to be assembled. No software did the assembling, so it fell to the salesperson, by hand, every time.

It wasn't for lack of trying. We hired sales engineers. We built workflow automation. We shipped our customers more capability inside the product every quarter. The thirty minutes stayed. The reason the thirty minutes stayed was that none of those investments owned the assembly. They owned more capture, more workflow, more tooling. The assembly was a different kind of problem, and we didn't have a name for it yet.

I watched the same scene play out, in a different shape, when I was running VA Linux a decade earlier. Before a board meeting or a customer call, I'd spend an hour or two reconstructing what I should already have known. The data sat in twelve systems. The understanding sat in none of them.

What I was missing both times was the assembled, current, situation-aware picture that a senior leader needs to walk into a moment ready. The information itself was abundant. The picture was the thing no system owned.

I have watched this same pattern in every company I have operated in or around since. Founders, partners, CEOs, COOs, Chiefs of Staff. The scene plays out at different scales, in different industries, on different stacks. The systems they use capture more activity than any previous generation of leader had access to. The systems do not relieve them of the work of carrying the picture forward.

For thirty years, the productivity stack has grown more powerful at recording what happened. It has not grown a category of software responsible for what happens next.

This essay names that missing category.

A Pattern Across Companies

The thirty-minute scramble is a category-of-work story, and it shows up across industries that look very different from the outside.

I spent time recently with a public-company COO in industrial services. She runs an operating cadence that includes six meetings before noon, each with a different stakeholder group, each requiring different context. Her calendar told her where to be. The seven tools she lives inside told her parts of why. The picture she actually walked in with was something she assembled in the seven-minute gaps between meetings, on her phone, in the elevator.

A growth-stage CEO I worked with last year has a different version of the same problem. His variable is decision density. Twenty-plus consequential decisions per week, each with downstream effects that compound across the quarter. He has a Chief of Staff. The Chief of Staff hits a ceiling, because the work of carrying the executive's context has outgrown what any one person can do, even an excellent one.

A fractional Chief of Staff I know runs three engagements at once. Three principals, three contexts, three sets of relationships and commitments. The story is sharpest in her case. The market need preceded the tools by years, and the tools still haven't caught up.

In each case, the stack was different. The industry was different. The role was different. The pattern was the same. The system that should be carrying the executive's picture across decisions, commitments, and goals does not exist, so the executive carries it, or the Chief of Staff carries it, or it does not get carried.

When a pattern shows up in that many places, across that many tools and that many sectors, it is a category problem.

A Missing Category

For most of those thirty years, the assumption was that the gap was a feature problem inside one of the existing categories. The right answer would be a better calendar, or a better CRM, or a better meeting tool, or a better note-taker. Each generation of tools added more capability inside its category. Each generation left the executive's actual job, deciding under incomplete information, exactly as hard as it had been.

The feature framing missed what was happening one level up. Each tool captures its slice. The picture across the slices, where the executive's work actually lives, falls outside every tool's responsibility.

A category is a different unit of analysis than a feature. A feature improves a tool. A category answers a question about responsibility. The question is whose job it is to keep the picture current across the executive's decisions, commitments, and goals. The answer should leave the leader walking into every moment with the situation already assembled. For thirty years, no category of software answered.

The gap that thirty years of feature work could not close is what Executive Intelligence names.

Information, Memory, Knowledge, Context

Four words get used as if they meant the same thing. They don't, and the distinction is what makes this category visible.

Information is what was said, written, or recorded. The email exists. The call happened. The deal-stage field is filled in. Capturing information is what most enterprise software has done well for a long time.

Memory is the system's ability to find that information again on request. A good database remembers. A good search engine remembers. Memory turns capture into recall.

Knowledge is the synthesis of information into something the system can reason over. A document index has knowledge. A trained model has knowledge. Knowledge moves a system from "I can find this for you" to "I can answer questions about this."

Context is something more specific, and it's what the executive needs. Context is the maintained, current, situation-aware picture of the work. It holds the decisions in flight, the commitments outstanding, and the goals in play. It stays ready for the next moment on the executive's calendar.

Every modern tool delivers some combination of the first three. Almost none of them delivers the fourth. The executive does, by hand, before every meeting, every decision, every conversation. The cost of that manual reassembly is the largest ongoing tax on senior leadership, and almost nobody puts a number on it because it doesn't show up on a line item.

I will come back to that cost in detail in the next essay in this series. The point here is the distinction. Information is captured. Memory is retrieved. Knowledge is synthesized. Context is maintained.

Context is the one a system has to own.

Why the Category Was Missing for Thirty Years

Three structural forces kept this category from existing. Each of them has loosened in the last few years, and the loosening is the reason the category is buildable now.

The first force was point-tool architecture. The productivity stack was assembled one tool at a time, each solving its own slice: CRM for customer information, email for correspondence, the calendar for scheduling, document tools for artifacts, chat for the back-and-forth, and project trackers for delivery. The architecture grew up vendor-by-vendor. Each tool was responsible for its own data. None of them was responsible for the picture across all of them. No vendor had a reason to build that picture. Each one made its money by keeping the executive inside its own slice. Point tools have built a powerful stack. Point tools have never produced an executive's situational picture, because they were never asked to.

The second force was the complexity threshold. For a long time, the executive could hold the picture in their head. The tools were few enough, the team was small enough, the deal count was low enough, the relationships were narrow enough. Memory and discipline scaled with the work. Then the work crossed a threshold. The number of tools, the velocity of decisions, the breadth of relationships, and the count of in-flight commitments grew past the point where any individual could carry the picture. The work began to ask for a system. None existed. The Chief of Staff role grew up in part to absorb that demand on the executive's behalf, which is one of the reasons the role exploded in the last fifteen years. Companies pay a premium for that work now: the average Chief of Staff base salary hit a record $167,954 in 2025, up 8.5% in a single year (Chief of Staff Network 2025 Salary Report). A human Chief of Staff is a partial answer, and the work has continued to grow past what one excellent person can carry.

The third force was the technology gap. The architecture this category needs wasn't something software could build a decade ago. It has to maintain a continuously current picture of an executive's decisions, commitments, and goals. It has to pull from many different source systems. It has to treat people and projects as first-class entities whose state changes over time. The graph databases weren't mature at scale. The language models couldn't reason over the entities with the precision the work requires. The infrastructure to ingest, encrypt, and synthesize at executive scale wasn't commercially available. The category required architectural primitives that had to exist first, and those primitives have only come into reach in the last few years.

The clearest sign the primitives have arrived is that operators have started building this for themselves. I recently spent time with a founder in confidential computing who got tired of waiting and built his own version. He runs his company on it. When he described the core of what he had built, it came out as accounts, people, and projects held as linked objects in a knowledge base, with the system analyzing his own working sessions to get better at helping him. He had arrived, on his own, at the architecture this category requires. The capability is real. What doesn't scale is asking every executive to build and maintain it themselves.

Three forces, none of which has been true for very long, kept the category from forming. Two of them, the complexity threshold and the technology gap, have changed. The third, point-tool architecture, has stopped solving the problem the executive actually has.

Something else changed at the same time, and it is what turned a tolerable gap into an urgent one. AI has moved the executive's work from doing to deciding. Software now handles the routine production that used to fill the day, and what remains is judgment: the decisions, the tradeoffs, the calls only the leader can make. Every other function already has an AI accelerator for its core work. Sales got one, engineering got one, support and marketing got theirs. The pattern shows up in the returns: 89% of executives say AI has sped up their work, and only 6% can point to organization-wide returns from it (Atlassian State of Teams 2026). The speed arrived. The organization-level payoff did not. Leadership is the function still waiting, and its core work, deciding under incomplete information, is exactly what this category was built for.

The category became buildable around the same moment the category became urgent. That's why now.

What the Category Is

Executive Intelligence is a category of software with four defining properties.

It is a layer of software. The category sits above the existing work stack. It reads the executive's existing tools and builds the record from the work itself, so the executive never stops to log a decision, file a commitment, or update a system. It maintains a continuously current picture of their work. It delivers context back into the tools they already use. Email is still email. The calendar is still the calendar. The CRM is still the CRM. The category lives one level up from any of them, where the picture across the stack is assembled. The position is above the stack, by design, because that's the only architectural location from which the picture can be carried.

It is responsible for follow-through. The category's job is to make sure the decisions, commitments, and goals a leader sets in motion carry through, and to keep a current, situation-aware picture of the work while they do. It holds that picture continuously, carries it forward between sessions, and keeps it current as new information arrives, the way a payroll system maintains payroll, the way an accounting system maintains the books. Follow-through becomes the thing the system owns, as a first-class responsibility.

It operates across the executive's decisions, commitments, and goals, across time. These are the three dimensions of the executive's work that have to stay current. Decisions made earlier in the quarter inform decisions made later. Commitments to people are tracked across their lifecycle. Goals set earlier pull the rest into line, so the day's decisions and commitments stay pointed at what the executive is working toward. The category's responsibility is to hold all three in the same model, so they reinforce each other. On its own, the system keeps the work aimed at what the executive set out to accomplish, so the goals themselves carry through. Today commitments sit in task tools, goals sit in planning systems, and most decisions live nowhere at all. This category gives decisions somewhere to live. The system reads the meeting notes, the emails, the conversations, and the documents, and it knows what was decided without anyone stopping to record it. A decision made offline is no different once it surfaces: the system pulls the context around it together, the people, the project, the history, and the earlier calls it builds on. The executive never assembles the record. For the first time, a system of record for decisions exists, built from the work itself.

It operates at the scale no person can maintain. The executive's role generates more decisions, more commitments, and more goals than any single human can carry, even with the help of a very good Chief of Staff. The category exists because the scale of the work has passed the threshold of individual memory. Software has to do the carrying.

Those four properties together name the category. Take any one of them out and what's left is a different kind of system. A layer of software that maintains context only inside one tool is a productivity feature. A system that maintains commitments without decisions is a task manager. A model that holds the picture for one session and then resets is a copilot. A capability that works at the scale of one project but not at the scale of an executive's portfolio is a meeting assistant.

Four properties. One category. The name is Executive Intelligence.

What the Category Does for the Executive

A system that does that work delivers three capabilities the executive has never had from any other category of software.

It maintains preparation. The executive walks into every meeting with the situation already assembled. The decisions on the table, the commitments outstanding, the relationship history, the open threads, the context the moment requires. The preparation arrives ahead of the moment, because the model has been maintained continuously and the system already knows what's on the calendar.

It maintains decision continuity. Decisions made earlier survive the meeting they were made in. The record scales with the stakes. A routine call stays a single line, and a consequential one carries the full reasoning: what was decided, what was ruled out, who disagreed, and the evidence behind it. When the same question comes up again in a different room, the system brings that record back. The executive can build on the previous decision and skip the re-litigation. Decision amnesia is one of the hidden taxes of senior leadership, and a system that maintains decision continuity removes it. The system goes further. When a decision is made, the system predicts how it is likely to turn out, drawing on its maintained picture of the people, the projects, and the history the decision touches. The executive reviews that prediction and adjusts it where their judgment differs. The system logs the expected outcome, then later sets it against what actually happened. The executive judges the call on the reasoning that produced it, separately from whether the outcome was lucky. Over time, the executive can see how well their own judgment holds up.

It maintains context across multi-entity operations. The executive runs a portfolio of projects, a network of relationships, and a set of commitments distributed across multiple teams and outside parties. The system maintains the picture across all of them, simultaneously, and surfaces what the moment requires. The executive's job is to allocate judgment across that portfolio. The category's job is to keep the picture of the portfolio current, so the judgment lands on the right things.

Those three capabilities define what the category does for the executive. They are the operating consequences of having a continuously maintained, evolving picture of the work. They are also what the executive recognizes immediately the first time they have them, because they are the work that has been done by hand, badly, for thirty years.

Six Adjacent Categories

Six categories of software each solve part of the executive's day. None of them does the work Executive Intelligence does, and the distinctions are worth naming, because the executive is otherwise asked to evaluate the category against the wrong yardsticks.

Productivity software increases output. Better calendars, better task managers, better document tools. They help the executive get more done in a given hour. The executive's harder problem is the picture they walk in with, which sits one level up from output.

Dashboards and BI tools come closest to the idea of an assembled picture, and they are the most serious prior attempt to build one. They pull metrics from the source systems and arrange them into a single view: revenue, pipeline, headcount, and the operating numbers that track the state of the business. A dashboard is the right tool when the question is what the numbers say. The executive's harder problem is the picture the numbers can't hold: the decision made last quarter and the reasoning behind it, the commitment a partner is waiting on, the relationship history that shapes how the next meeting goes. Dashboards assemble what can be measured, and the executive's situational picture is mostly the part that can't.

AI copilots answer questions on demand. They wait for the executive to ask, then produce a response. The architecture starts fresh every session, because there's no maintained model behind the chat. A copilot is a useful tool for ad-hoc questions, and it lives in a different category from a system whose job is to be ready before the executive even formulates the question.

Meeting assistants capture what happened inside one meeting. They transcribe, they summarize, they extract action items. They process a single event well. Maintaining the picture across every meeting, every decision, and every relationship the executive carries falls outside what they were built for. They record what was said. They never record what the executive expected to happen, or check it against what did.

Knowledge bases store artifacts. Documents in folders, with search on top. A knowledge base is the right tool when the executive can name the document they need. The executive's harder problem is the one where they can't name what they need, because what they need is an assembled picture of a situation, and no single document holds that picture.

General-purpose assistants serve any task. Breadth is what defines them. Executive Intelligence is built around the opposite design choice: one job, executive readiness, with everything outside that scope refused. The narrower the responsibility, the more reliably the system carries it.

Each of those categories is useful for its own purpose. Each of them is a different category from Executive Intelligence. The executive who tries to assemble Executive Intelligence from those six categories ends up doing the assembly work by hand, which is the work the category exists to do.

What Changes When the Executive Has It

The first sign that an executive has Executive Intelligence is rarely something they can point to inside a tool. It shows up in how the day feels.

The scramble between meetings stops. The executive walks in with the assembled picture already in front of them, and the time they used to spend reconstructing context gets spent on the conversation that is actually happening in the room.

Decisions stop being re-litigated. The reasoning behind every prior decision is carried forward and surfaced when the same question comes up again, so the executive can spend their judgment on the new variable in the room and treat the prior decision as settled.

The Chief of Staff, where one exists, stops being the executive's working memory. The Chief of Staff's job moves up the stack, to the work humans are uniquely good at: judgment on hard cases, relationship work, strategic positioning. The mechanical reassembly that used to consume their week is done by the system.

The team starts running on a shared picture. The CEO, the Chief of Staff, the operating leaders, the executive assistant, the partners around the table all see the same assembled context, kept current in the background. Alignment stops being something the executive has to manufacture in every meeting.

The portfolio stops feeling like a stack of separate fires. The executive sees the full set of decisions, commitments, and goals in one model, with the items that need attention surfaced in priority order. The work of triaging the day moves from the executive's attention to the system's responsibility.

Those are the operating consequences of a system whose job is the executive's situational picture. Each one shows up as a kind of friction that simply stops, which is why the executive feels the change before pointing to its cause.

The Vocabulary This Series Will Use

A few terms will recur across the next six essays in this series. Each carries a precise meaning, and each is load-bearing for the argument.

Executive Intelligence is the name of the category.

The World Model of Work is the technical name for the architecture that makes the category buildable. It is a continuously maintained model of the executive's work. It lives outside the language model. It draws from the executive's source systems and anticipates what the next moment requires.

The Knowledge Graph is the durable substrate at the foundation of the World Model of Work. People, projects, decisions, commitments, and goals as first-class entities, with state that evolves over time.

Follow-through is the responsibility the category owns: whether the decisions, commitments, and goals a leader sets in motion carry through. Context continuity is how it gets there: understanding held continuously, so the executive never has to reconstruct it.

Context blindness is the condition the category exists to end: the structural gap between what the work stack captures and what the executive needs to know. The next essay in this series takes context blindness apart in detail.

Those five terms are the vocabulary the executive, the analyst, and the operator will need to talk about this category as it forms. They will be used consistently across the series.

The Executive Intelligence Papers is a series of seven essays by the founders of Brief on the formation of a new category of software.