Two skills. Roughly the same effort to learn, roughly the same few weeks of being bad at them in private before they became automatic.
The first one got me name-checked in a leadership meeting I wasn't in. Someone asked in a review whether we could see the numbers split by region, and instead of saying "I'll send that through tomorrow", I built the pivot and dropped a slicer on it while the room watched, and we carried on talking about the actual question. Two days later my manager told me a director had mentioned it.
The second one saved about forty hours a year. I rebuilt the monthly reporting pack as a Power Query refresh, and a job that had eaten two days at the start of every month became a button and a coffee. Nobody has ever mentioned it. Not once, not in a review, not in passing.
The obvious advice here is that I should have told people. I did tell people. It didn't work, and it was never going to, and the reason is structural rather than a failure of self-promotion.
Credit attaches to problems people can still feel
The pivot got noticed because there was a person in the room, at that moment, holding a question they wanted answered. The gap between wanting the answer and having it was about ninety seconds and everybody watched it close. That is a complete story with a beginning and an end, and the audience was already assembled.
The refresh got nothing because it removed a problem that nobody was standing in. The two days of copy-paste weren't causing anyone else pain. They were causing me pain, quietly, at month end, in a way that had been true for so long it had stopped being a fact and started being weather.
And this is the part that makes "just tell them" fail: automation deletes its own evidence. The moment the pack refreshes in ten minutes, the two days stop existing anywhere except in your memory. There is no before state left for anyone to compare against. When you say "this used to take two days", you are asking someone to take your word for a number they never held, about a problem they never noticed, in a conversation you started. That lands as either a complaint or a boast, and often both.
There's a version of this that's easy to be bitter about. I'd rather be accurate: nobody withheld credit from me. There was nothing left to give credit to. The work was real and the evidence evaporated on contact with success.
What actually helps
Install the number before you fix it, not after. This is the whole game and it costs one sentence. Before you touch the manual job, say out loud, to someone who isn't you: "the monthly pack takes me about two days, I'm going to see if I can cut that down." Now the baseline lives in someone else's head, put there when you had nothing to gain by inflating it. Afterwards, "it's about twenty minutes now" completes a story they already started following. Said cold, six months later, the same sentence is just a claim.
Hand over an object, not an outcome. A saving you describe is a number. A thing someone can press is a thing. When the refresh was finished I should have walked it over, put it on their screen, and let them click the button and watch the pack rebuild. Ten seconds of someone else's hands on the work does more than a paragraph in a status update, because it turns your invisible thing into their experience of it.
Don't manufacture the visible moments, but be ready for them. The pivot worked because someone asked. If I'd opened that meeting by announcing I could build a slicer, it would have been a party trick and read as one. The skill that gets noticed is the one you can produce, without ceremony, at the exact moment someone else wants it. That means the preparation is invisible by design and the payoff is entirely a matter of timing. It's worth knowing this so you stop waiting for the invisible work to produce its own moment. It won't.
This week
Find the recurring manual job you're most likely to improve in the next few months. Don't improve it yet. Write down what it currently costs, in hours or in days, somewhere at least one other person will see it - a message to your manager, a line in a status update, a comment in the standing meeting.
That's the whole task. It takes a minute and it is the only cheap opportunity you will get, because the moment the job gets faster, the cost you're trying to describe stops being verifiable.
That closes out this thread. It started with a keyboard shortcut that made people think I was better at my job than I was, which is the same mechanism running in your favour: a visible moment doing more work than the underlying skill deserved. Then being good at Word, which nobody puts on a job description until the day it saves an eighty-page tender, where the skill has no name to be credited under. This one is the third shape of the same problem: work that is genuinely valuable, genuinely yours, and structurally unwitnessed. None of the three are solved by being better at the tool. They're solved by paying attention to who is in the room and what they can actually see.
Frequently Asked Questions
Isn't this just telling me to be better at self-promotion?
No, and the distinction matters. Self-promotion is describing your value after the fact to someone who has no way to check it. What's being suggested here is much narrower and much less uncomfortable: put a number in someone's head at the one moment when stating it costs you nothing, then let the result speak. That's evidence management, not personal branding, and it works for people who find self-promotion excruciating.
What if I've already automated the thing and never mentioned the before state?
You've missed the cheap opportunity, but not the only one. The next best move is the object, not the number: get the thing in front of the person, let them use it, let them notice how fast it is without being told. Failing that, the honest version works better than the retroactive claim - "this used to be a two-day job, which is why I'm suggesting we do the same to the quarterly one" points forward and gives them a reason to care that isn't your appraisal.
Does this mean the invisible work isn't worth doing?
It's worth doing, and it's worth doing knowingly. Forty hours a year is forty hours of your life back regardless of whether anyone claps. The mistake isn't doing invisible work, it's doing invisible work while quietly expecting it to be seen, then feeling resentful when it isn't. Decide which it is before you start. Work that is invisible because it just works is a whole category, and most of it is still worth building.
My manager says they want people who improve processes. Doesn't that contradict this?
They do want it, sincerely, and they still can't see it. Wanting a thing and being able to observe a thing are different capabilities. A manager can observe outputs, meetings, and things that arrive on their desk. A process improvement produces the absence of a problem, which is the one category of result that never arrives anywhere. Take the stated intent at face value and then do the work of making it observable anyway.
What if the visible moment never comes?
Then engineer proximity rather than the moment itself. The pivot happened because I was in a meeting where the question could be asked. The pattern generalises: the visible moments happen where the questions get asked, so being in the room, or being the person the question gets forwarded to, matters more than any individual demonstration. If you are structurally never in the room, that is the actual problem to solve, and no amount of Excel fixes it.
Is this different early in a career versus senior?
Early on, visible moments are worth more than they should be, because nobody has enough evidence about you yet and a single ninety-second demonstration fills a lot of empty space. Senior, the calculation inverts: you have a track record, and the risk is that you become the person who is fast at tools rather than the person who decides what the tools are for. The invisible work is more valuable at that stage and harder to get credit for, which is exactly why the baseline sentence matters more, not less.