Read the opening of chapter 1
The opening of chapter 1, exactly as printed.
Chapter 1: The Senior BA's Mandate: Judgment Over Evidence
Eleven weeks before DirectClaims was scheduled to go live, you told the steering committee to kill it. Not delay it — kill it, with the vendor contract still open, two directors whose bonus targets were tied to the launch date, and a build that was, by every visible metric, on schedule. You didn't have a failed test case. You didn't have a blown SLA. You didn't have a single document that said "this will fail."
What you had was four things, none of which meant anything on its own. The vendor's public roadmap had quietly dropped an integration they'd committed to in writing eight months earlier. No announcement, just an absence on the next version of the same PDF. Two VPs, asked the same scoping question three days apart, gave answers that contradicted each other in a way neither seemed to notice. Pilot telemetry showed a decline in adjuster engagement that a plausible alternate story — seasonal claims volume — could explain just as well as your read of it. And three of the sponsoring VP's direct reports had left within six weeks, which happens, and means nothing, until it's the fourth data point in a pattern you're the only person tracking.
Each of those, cross-examined individually, falls apart. A lawyer would take any one of them apart in ten minutes. Together, they told you something specific: the business case underneath the platform had shifted, quietly, out from under a schedule that hadn't shifted at all, and launching the build would commit the company to eighteen months of running a platform whose business case the people who'd have to live with it no longer believed. You said so. The project stopped.
Keep readingShow less
Four months later you're in the same room, with most of the same eleven people, and the company has moved on to a smaller integration project that shipped on time and, as it turns out, is the reason claims processing didn't buckle during a regulatory reporting deadline nobody had fully modeled when DirectClaims was still alive. That's a good outcome. It is not proof. Nothing rules out the version where DirectClaims would have limped to a mediocre finish and everyone would have adjusted. The CIO, going around the table in the retrospective, gets to you and asks you to walk the room through exactly what told you to call it.
You reach for the artifact that makes this easy, and there isn't one. There's no ticket, no requirements doc, no meeting minutes that add up to "kill it" on their own. There's a workshop transcript here, a telemetry export there, a vendor PDF you happened to have diffed against the prior version out of habit. None of it, read cold, gets anyone in the room to your conclusion. What gets them there is you, live, walking the reasoning back out loud, in the order you actually built it, in a room where at least two people would benefit from you being wrong. That reconstruction — not any document you produced along the way — is the thing being graded. It is also, whether or not anyone in the room would phrase it this way, the actual test of whether you're the kind of BA this organization keeps handing its ambiguous, high-stakes calls to.
This is the mandate this book is about. Not "gather requirements better" or "communicate more clearly." You've done both of those competently for years, or you wouldn't be reading a book pitched at this level. The shift that defines senior BA work is that the deliverable stops being an artifact and starts being a defensible synthesis, produced under conditions where the evidence is deliberately incomplete, because if it weren't incomplete the decision wouldn't have needed you. Everything in this chapter, and most of what follows, is about operating well inside that condition rather than pretending you can engineer your way out of it with more documentation.
You know the four core responsibility areas — decode, optimize, document, communicate — as a description of what a business analyst does. At this level, that description is true and almost useless. What matters now is that each of the four is a place where you are either accumulating standing or spending it, whether or not anyone frames it that way out loud.
Decode, taking ambiguous business signal and turning it into a position someone can act on, is where the DirectClaims call lived, and it's the highest-variance of the four precisely because your work there is invisible until it's tested. Nobody audits a synthesis while it's being built. They audit it after the fact, usually at the moment it's most inconvenient for you, which is exactly what happened in that retrospective. Every time you decode ambiguous signal into a recommendation, you're making a forward bet against your own credibility, and the size of that bet scales with how much weight the recommendation carries and how little corroborating evidence exists. The senior-level skill isn't better decoding technique. You already have that. It's knowing, in real time, how much of your standing a given call is worth risking, and being honest with yourself about whether you're making the call because the signal warrants it or because you're bored of ambiguity and want to resolve it.
