> ## Content Index > Fetch the complete content index at: https://www.theadamthomas.com/llms.txt > Use this file to discover other available public pages before exploring further. # The Number Nobody Will Say - URL: https://www.theadamthomas.com/the-number-nobody-will-say/ - Published: 2026-10-20T17:00:27.000Z - Updated: 2026-10-20T17:00:26.000Z - Author: Adam Thomas - Tags: Newsletter $942,068.68. I know that number to the cent because somebody made sure that I did, and twenty years later I still can't round it off in my head. I was in my twenties at DTCC, the company that clears the plumbing underneath the American financial system — last time I looked they moved something like $2.8 quadrillion in a year, a figure so large it stops meaning anything at all. I misunderstood what they call an IPL, an initial program load, and I was rushing at the end of the day. It was a Thursday afternoon. That matters only because Thursday afternoon is precisely when you shouldn't touch a mainframe, and I haven't made a change to anything important on one in the two decades since. The fine came back from the SEC at $942,068.68. Yikes. Here's what I expected: a short meeting, a manager reading carefully from a script somebody in legal had written, and my badge quietly failing to work on Monday. Here's what happened instead. My VP, a man named Tony Dolan, sat me down and told me I was a bright kid and that a million-dollar mistake was going to turn out to be a great investment in my time there. I used to think that sentence was about mercy, and I was wrong about that for years. Look at what he actually did in it. He said the number out loud, in full, to the person who'd caused it. No rounding it down to protect me, no inflating it to make a point of me. Then, having established the real size of the thing, he told me what it now meant going forward. That order is the whole job. ## The reassurance that costs more than the truth Your team already knows that something changed. They watched two people clear out their desks, they noticed the req get pulled, and they can see the AI tooling quietly absorbing work that used to have a name and an owner on it. They have also noticed that nobody above you has adjusted a single date. What they don't have is a number. So what do most of us hand them instead? Some version of we're going to be fine. We're leaner now, we'll be more focused, there's genuinely an opportunity buried in this — all of which may be perfectly true, none of which is a number, and a room full of experienced people can tell the difference between those two things instantly. Here's the part that should worry you, though. Vague reassurance doesn't read as calm; it reads as one of exactly two things. Either you don't know what happened, or you know and you won't say it. So which is it going to sound like on Monday? One makes you look absent. The other makes you look like you're managing your team rather than leading them, and a senior engineer who suspects you're withholding will stop telling you what's actually breaking. You have just traded one hard conversation for a slow, invisible loss of honest resistance. A named constraint can be planned against. An unnamed one gets absorbed. Absorption is the real failure mode here, and nobody ever announces it. The team works the extra hours, silently drops the third priority, and stops writing things down. Meanwhile the plan up on the wall carries on describing a company that doesn't exist any more. Because absorption looks exactly like loyalty when you're viewing it from one level up, it tends to get rewarded right until the quarter that it collapses. ## What the new number actually has to contain So say it — but say what, exactly? "We lost two people" isn't a number your team can plan against either, because that's a fact, and facts are not plans. A usable new number has three parts, and it is reliably the third one that people skip. The Delta is what capability actually changed, expressed as a proportion of the thing your team ships rather than as headcount. Losing a data scientist and an engineering manager from a team of nine isn't a 22% cut; it is the removal of the only person who could size a model problem and the only person who could run a release without you in the room. Say what left in terms of what it used to let you do. The Owner is who holds the gap now, named out loud, before anything else gets committed. An unowned gap isn't really a gap at all — it's a surprise with a delay on it. The Drop is what comes off the plan, announced publicly, with a date attached to it. The Drop is where this stops being a communication exercise and starts being leadership, because it is the only part of the three that costs you anything. Anyone can narrate a loss well. It's also the part nobody trains. Leadership programs spend a day on delivering hard news and next to nothing on deciding what comes off the plan. Naming what you'll no longer deliver, in a room that contains the person who wanted it, is what makes the first two credible. And who's going to say it, if you don't? Skip it and your team correctly hears the whole performance as theater — you acknowledged the constraint and then changed precisely nothing, which is just absorption with better slides on it. Notice what none of this touches, though. Not the vision. The destination doesn't move because your capability changed; only the route to it does. Leaders get this backwards constantly. They protect the plan while quietly letting the vision go soft, when the plan is the disposable artifact and the vision is the only reason anyone joined. ## The gap you can measure this week Here's the diagnostic, and honestly it takes about ten minutes to run. Write down the percentage by which your team's delivery capability changed over the last ninety days — estimate it as the share of last quarter's roadmap you could still ship with the people you have now. Then write down the percentage by which your committed roadmap changed across those same ninety days. If the first number is 30% and the second one is zero, you don't have a communication problem on your hands. You have a plan that has quietly become fiction, and what do you imagine every week it stays up on the wall is teaching your team? That the plan is decorative. Ron Kohavi's research puts somewhere between 60 and 90% of product ideas as failing to move the metric they were built for. That's in the good times, with full staffing and nobody's morale in the tank. Running an unchanged roadmap through a diminished team doesn't beat those odds. It only relocates the failure to somewhere you will find it much later. Tony Dolan didn't make my mistake smaller that afternoon. He made it exact, and then he told me what it had bought. That same move is available to you on Monday. Say the delta, name the owner, announce the drop. Then say the vision again, completely unchanged, so the room understands what the re-cut is actually for. Do the ten-minute version first, though. Two percentages, written on paper, before you say a word to anybody. If those two numbers don't match each other, you already know what your next staff meeting has to be about.