I watched a capture manager lose a conversation because his intelligence was at his desk and he was not. He was at an industry day, shaking hands with a program office contact who mentioned a contract coming up for recompete. He nodded, said he would look into it, and tapped a note on his phone. By the time he got back to his laptop that evening, the moment was gone and so was the specificity. He could not recall the contract number, could not pull the incumbent on the spot, could not say anything that made him sound like he already knew the landscape.
That is the gap I built the mobile surface of my GovCon intelligence engine to close. Not because mobile is a checkbox every product needs. Because federal business development physically happens in rooms — at expos, in hallways, over coffee at a conference center — and the intelligence has to be in the room too.
Where the work actually happens
GovCon BD is a relationship business that runs on in-person events. Industry days, association chapters, pitch sessions, prime-sub matchmaking, the endless circuit of conferences. The valuable conversations are unscheduled and the useful answer has a half-life of about thirty seconds. Someone mentions an agency, a program, a contract, and either you can engage with substance right then or you cannot.
A desktop tool is useless in that moment. The signal I built into the engine — the recompete windows, the incumbent strength, the eligibility read — is exactly what you want mid-handshake, and exactly what you cannot reach when it lives on a dashboard back at the office.
The most valuable thirty seconds in federal BD is the moment a buyer mentions a program by name. If your intelligence is at your desk, you spend that moment nodding instead of knowing.
Mobile is not the desktop, shrunk
The wrong way to do this is to responsive-resize the dashboard and call it mobile. The dashboard answers “what is my whole pipeline.” The field answers a completely different question: “this specific thing just came up in conversation, tell me what I need to sound informed in the next ten seconds.”
Those are different products that happen to share a backend. So I designed the mobile surface around field interactions, not around feature parity.
| At the desk | In the field |
|---|---|
| Browse the full pipeline | Look up one name fast |
| Deep solicitation analysis | The three-line summary |
| Configure, filter, export | Capture a lead before you forget it |
| Plan the week | Win the next thirty seconds |
The field version does fewer things and does them in one tap. Search a company, get its federal footprint and whether it is a vulnerable incumbent. Hear an agency named, pull its buying rhythm. Meet someone worth following up with, capture the context immediately so it does not evaporate by evening.
The constraints the field imposes
Building for a conference center, not an office, changes the engineering in ways that are easy to underestimate.
Connectivity is hostile. Convention center wifi is a punchline for a reason, and cell coverage in a basement expo hall is worse. The field app cannot assume a connection. So the data that matters for in-person work — the recent qualified opportunities and their key signals — is cached on-device after the morning sync. The summary you need mid-handshake is local. Anything requiring a fresh server round-trip is the thing that fails you exactly when you need it.
Latency is unforgiving in a way the desktop never is. At a desk, a two-second load is fine. In a live conversation, two seconds is the difference between looking informed and looking like you are stalling. The field surface targets sub-second response on cached data, because the human in front of you does not wait.
Capture has to be frictionless or it does not happen. The single highest-value field action is recording a lead and its context before memory degrades. If that takes more than a few seconds of tapping, it will not get done in the moment, and a lead reconstructed tonight from memory is a worse lead than one captured in the hallway with the details intact. So capture is one screen, a few fields, voice-to-text on the notes, and it syncs back to the main pipeline when connectivity returns.
What I deliberately left out
The discipline was in what the field app does not do. No deep solicitation analysis on the phone; that multi-hop retrieval work belongs at a desk where you can actually read the output. No configuration, no bulk operations, no exports. Every feature I added that was not a field interaction made the thing slower to use in the one context it exists for.
This is the same principle that governs everything I ship: build the simplest thing that works for the actual job. The job here is winning a thirty-second exchange and capturing what came out of it. Anything beyond that is desk work pretending to be field work.
The payoff is presence
The capture manager who can pull up an incumbent’s record while still shaking the hand is a different kind of presence in the room. He is not someone who will look into it. He is someone who already knows, who can ask the sharp follow-up question, who walks away with the lead and the context already recorded.
That presence does not come from a better dashboard. It comes from putting the intelligence where the conversation is. Federal BD is fieldwork, and intelligence that cannot leave the office is intelligence the field never sees. Built for missions means the signal travels to where the mission actually happens, which is almost never at a desk.