Getting Credit For Invisible Work - MAC155

It's four-forty on a Thursday, and a message lands that starts with two words. Quick question. It is not a quick question. It never is. Forty minutes later you've untangled why two teams were working from different numbers, gotten the right person on a call, and the thing that was going to blow up on Monday is now just a thing that got handled on Thursday. Nobody will ever know it was going to blow up. That is the entire problem.

This piece is about glue work — the mentoring, the unblocking, the documentation, the quiet cross-team diplomacy that keeps an organization from coming apart at the seams. Specifically, it's about why that work almost never shows up in the record of your career, why that gap gets dangerous the moment budgets tighten, and what to actually do about it. There's a filter for deciding which of it is even worth your time, a reframe for why it matters more than it looks like it does, and a documentation habit built specifically for work that leaves no evidence behind.

The Work That Has No Artifact

The term itself is borrowed from the engineering world, where work that falls outside anyone's job description but still has to happen got labeled "glue." The thing itself is not a technology problem, though. It's the ops lead who notices a broken handoff between two departments and just fixes it. It's the analyst who rewrites the recurring report so the next person doesn't have to guess. It's whoever onboards the new hire, mediates the standoff between finance and marketing, remembers why the process is the way it is, and answers the question that would otherwise have cost somebody a day.

Researchers have a colder name for it. Economists at Carnegie Mellon spent years documenting what they call non-promotable tasks — work that clearly benefits the organization but does nothing measurable for the person who does it, and it doesn't get distributed randomly. Women volunteer for it more than men do. They get asked to take it on more frequently. And when they're asked, they're more likely to say yes. Broader workplace research extends that same pattern to people of color, to LGBTQ+ employees, and — this is the part that surprises people — to high performers generally, where discretionary effort quietly becomes the expectation.

So this isn't a story about a bad manager taking advantage of you. It's a story about a measurement gap, and measurement gaps are structural. Every performance system your company has ever run measures artifacts — the campaign that launched, the close that finished on time, the report that shipped, the number that moved. Those things leave evidence behind: a date, a deliverable, a line in a system somewhere. Glue work leaves nothing. The meeting that didn't need to happen leaves no calendar invite. The escalation that didn't escalate generates no email chain. The person who didn't quit doesn't file paperwork explaining that they stayed because you talked them through a bad quarter. Your wins announce themselves. Your saves don't.

Career Quicksand vs. Promotable Glue

Before you go document all of this, you need a filter, because the honest answer is that not all of this work is worth doing — and doing more of it isn't automatically the move. I've watched people spend three years being enormously helpful and end up exactly where they started, not because helping was wrong, but because they never distinguished between two very different kinds of helping.

Run this two-question test on your own work: does this create leverage, and can I turn it into evidence? Work that fails both questions has a name worth remembering — career quicksand. Career quicksand is the recurring manual fix — solving the same setup problem for the eleventh new hire, rebuilding the same broken spreadsheet every month because it breaks every month. It genuinely helps, which is what makes it quicksand rather than just waste. Somebody's day got better, and next month it's your problem again — a permanent tax paid for a temporary rescue.

Promotable glue is the same instinct pointed one level up. Instead of answering the eleventh new hire's question, you write the thing that means the twelfth one never asks it. Instead of unbreaking the report, you fix why it breaks. This is the same distinction drawn back in Delegation and Leverage (MAC-066), applied to the work nobody assigned you in the first place — leverage isn't only about who does the task, it's about whether the task keeps existing. Quicksand feels like helping. Leverage means the helping stopped being necessary.

The Immune System

Think about an organization the way you'd think about an immune system. It's running constantly, doing an enormous amount of work every hour of every day, and the measure of how well it's working is that you feel completely normal. You don't experience its successes — you experience the absence of failure, which registers as nothing at all. The only time an immune system ever gets talked about is during an outbreak, which is to say the only time it becomes visible is the moment it has already failed.

That's the glue layer of your company. It runs constantly, and when it's running well, the organization experiences it as nothing happening. Quarter closed fine. Handoff went fine. New person ramped fine. Here's where the timing gets genuinely dangerous: when budgets tighten and headcount gets scrutinized, an organization in crunch mode depends on that layer more, not less. Fewer people, more handoffs, more ambiguity, more things falling between roles — exactly the environment where the glue is load-bearing, and exactly the environment where the reporting gets narrower and the review conversation shrinks down to what did you deliver. The demand for the work goes up at the same moment the visibility of the work goes down. Nobody budgets for the outbreak that didn't happen.

Your manager isn't sitting on your contributions here. In most cases they genuinely value them and have nowhere to put them — the form has fields for outcomes, and you produced non-events. The calibration-room mechanic from Your Manager Is Not Your Career Sponsor (MAC-139) applies directly: your manager walks into that room without you and has to make your case out of whatever material they're holding. "Everybody likes working with them" is not a case. That's a character reference.

There's a harder edge worth putting on this. The Indispensability Ceiling (MAC-145) described the pattern of being irreplaceable and stuck. Glue work is usually how people get there — you become the human API between two departments, the only one who knows why the process exists, and every one of those is a genuine contribution that, stacked together, builds a role that's extremely expensive to move and impossible to describe. Indispensable and unpromotable sit closer together than they sound.

So the reframe worth carrying out of this section: glue work isn't workplace citizenship, it's organizational risk management. Turnover has a price — burned-out employees are roughly three times more likely to be planning their exit, at an estimated cost of several thousand to twenty-odd thousand dollars per person per year. Ramp time has a price. Rework has a price. Escalations have a price. Every one of those numbers is a number your leadership already tracks and already fears, and the work you're doing lands directly on them. Nobody has ever connected the two out loud. That's the job now — not doing more of the work, connecting it.

The Prevention Ledger

This is where the callback needs to be handled carefully. We spent a whole episode on The Brag Document (MAC-141) — the running private inventory of your wins, kept all year, so you're never rebuilding twelve months from memory the week before your review. If you don't have one, that's the episode to start with; this piece isn't re-teaching it.

A brag document works beautifully on wins, because wins have edges — the project shipped on a date, the number moved by a percentage, something you can reconstruct months later from your own calendar. Glue work has none of that. There's no date, no artifact, and no memory. The Thursday afternoon where you caught the mismatched numbers is completely gone by the following Tuesday. It didn't feel like an accomplishment while it was happening. It felt like a Thursday.

So alongside the wins, keep what's worth calling a prevention ledger. Same habit, different input. Five to ten minutes at the end of the week — the same cadence any good documentation practice runs on — but instead of asking what did I finish, the question is what was about to go wrong that didn't. Four fields, ninety seconds each: what was heading for trouble (the situation, not your action); what you did about it, in one sentence; what it would have cost — a rebuilt deck, a delayed review, a credibility hit, estimated honestly rather than left blank; and who else saw it, by name, because your own testimony is the weakest form of proof available to you.

Write that last part while the smoke is still in the room, not at review time — you won't remember it then, and worse, you'll have already decided it wasn't a big deal. That's a real effect, not modesty. Which Career Body Are You? (MAC-152) covered research showing most people place their own contribution meaningfully below where outside evidence puts it, and glue workers are the most exposed group there is to that error, because their entire output is made of things that didn't happen. A brag document records what you built. A prevention ledger records what never broke.

The Action Plan

Everything above is a shape, not an answer — what counts as leverage in your organization depends on what your organization is actually afraid of this year, which is specific enough to your team and your politics that it's worth working through with someone directly if you're at that point. Four steps, and each one has a ready-to-paste AI prompt attached below, because an assistant will do pieces of this faster than you will — and on the first step, more honestly than you will, since you're going to undercount yourself and it has no reason to.

One. Run a two-week look-back tonight. Open your calendar and sent messages for the last two weeks and list every instance where you were the reason something didn't go sideways. Don't judge them yet. Most people find between four and nine, and are genuinely surprised — it's the first time the work has ever been in one place.

Two. Sort that list with the two-question filter — leverage and evidence. Everything failing both is quicksand. For each quicksand item, the move isn't to do it faster, it's to fix the thing that regenerates it, or hand it back. Pick one to fix this month.

Three. Start the prevention ledger this Friday. Ten minutes, recurring, on the calendar so it survives a busy week. If you already keep a brag document, this is a second heading inside it, not a second file.

Four. Translate three entries into your manager's language before your next one-on-one — turnover, ramp time, rework, or escalation, instead of "I helped Priya get up to speed." Try "I cut the new analyst's ramp from six weeks to three, which is about half a headcount of recovered capacity." Same event. One version is a personality trait. The other is a business result.

Do the first one this week, not next quarter — it takes twenty minutes and it's the step everything else depends on.

The Close

The work you're doing is real. In a tight market it's more load-bearing than it's ever been, and the organization is quietly relying on it while having no mechanism whatsoever to see it. That gap isn't a reflection of your value. It's a reporting failure, and reporting failures get fixed by better reporting — which is a thing you can start doing this Friday afternoon.

So the next time it's four-forty on a Thursday and a quick question lands in your inbox, go handle it. That's who you are and it's worth being. Then take ninety seconds and write down the fire that didn't happen. Because nobody else in that building is going to.

Get the Episode + Prompts

The full episode and show notes are at managingacareer.com/155. Below are all four prompts from the action plan, copied out in full so you can paste them straight into whatever AI assistant you use — change the details to fit how you actually work, but keep the guardrails in each one intact. They're doing real work.

Prompt — the look-back.

Search my sent email and calendar for the last two weeks. I'm looking for moments where I prevented a problem rather than delivered something: places where I reconciled conflicting information, unblocked someone who was stuck, flagged a risk before it landed, mediated a disagreement, supplied context nobody else had, or answered a question that would otherwise have cost someone most of a day. For each one, give me the date, who was involved, what was about to go wrong, and what I did. Exclude routine status updates and anything that was formally assigned to me — I only want the work that wasn't anyone's job. List them individually; don't summarize.

Prompt — the sort.

Here is a list of problems I prevented at work. Sort it into two lists using two tests. Test one, leverage: did this reduce the chance the same problem happens again, or did it only fix this one instance? Test two, evidence: is there an outcome someone other than me could confirm? Anything failing both goes in a list called Career Quicksand. Anything passing both goes in Promotable Glue. For each quicksand item, add one sentence naming the durable fix — the change that would stop it regenerating. Be skeptical rather than generous: if something only resembles leverage, put it in quicksand and tell me why.

Prompt — the weekly entry.

I keep a weekly log of problems I prevented at work. I'll paste a rough note. Turn it into a four-field entry. One: what was heading for trouble — describe the situation, not my action. Two: what I did about it, in one sentence. Three: what it would have cost if nobody had caught it — give a concrete estimate in hours, money, rework, or delay, and state the assumption you used. Four: who else witnessed it, by name. If my note doesn't contain enough to fill a field, ask me a question instead of inventing the detail. Here's the note:

Prompt — the translation.

Here are three entries from my log of prevented problems. Rewrite each as one sentence a manager could repeat in a calibration or headcount meeting. Anchor every sentence to one of four measures: turnover cost, ramp time, rework, or escalation. Keep estimates conservative and state the assumption behind each one so I can defend it if I'm challenged. Do not use the words "helped," "supported," or "assisted" — every sentence has to name a business outcome. Give me two versions of each: one for a written self-assessment, and one I can say out loud in a one-on-one without sounding rehearsed.

Links & References

• Research & sources — Why Women Volunteer for Tasks That Don't Lead to Promotions
• Research & sources — Who's Doing the Glue Work at Your Company?
• Research & sources — How to turn invisible impact into career growth (the Glue Work Matrix)
• Research & sources — What is a Brag Document? Complete Guide for 2026
• Research & sources — Invisible labor at work: what it is and how to address it
• Related Managing A Career episodes — Delegation and Leverage (MAC-066)
• Related Managing A Career episodes — Your Manager Is Not Your Career Sponsor (MAC-139)
• Related Managing A Career episodes — The Brag Document (MAC-141)
• Related Managing A Career episodes — The Indispensability Ceiling (MAC-145)
• Related Managing A Career episodes — Which Career Body Are You? (MAC-152)

TAKE THE SURVEY!

Are you looking for a career coach? If you reach out to me via the contact form, I will arrange an introductory session where we can talk about your career goals and how I can help. If we're a good fit, we can schedule regular coaching sessions.