A Day in the Life of an Operations Manager
The job description said "drive operational excellence and own cross-functional process improvement." You read that line at the offer stage and pictured something with a whiteboard and a Gantt chart. Eighteen months in, the actual work looks more like this: you are the human Zapier between Finance, Legal, IT, and whoever just Slacked you "quick question." You broker. You chase. You write docs that nobody reads, and then you read them out loud to the people who didn't read them. That is the job.
This is a real Tuesday. Not the curated one from the careers page. The one that ends with 22 tabs open and a Notion page you started writing at 2pm and never finished.
8:14am — The Ticket Queue
You open Asana. Or Monday. Or Linear if your engineering team shoved you onto theirs because "ops requests are basically tickets, right?" Fourteen tasks landed overnight. You sort them in roughly this order:
- Three are urgent. A sales AE locked out of HubSpot. A new hire whose laptop didn't ship. A finance lead asking for a vendor's W-9 because a payment is on hold.
- Five are routine. Access requests, expense reimbursements, the "can you add me to this Notion page" stack.
- Four are people asking where the Notion page is. It is pinned to the channel. It has been pinned to the channel for six months.
- Two are actual problems disguised as questions. One is "hey can we get a new tool for X" (that is a $14,000 procurement decision pretending to be a tag-along ask). The other is "is there a process for offboarding contractors?" There is. You wrote it, and the fact that someone is asking means they didn't find it, which means the process is broken.
The first 45 minutes of your day is triage. The signal-to-noise problem is the actual job here. If you start replying to every ticket as it lands, you become the team helpdesk and your week is gone. The discipline is to sort first, batch second, and protect at least one block of focus time before noon. You set status to "do not disturb" in Slack. Three people ping you anyway. One starts with "sorry to bother you."
You unblock the urgent three, ack the routine five with a one-line "in queue, end of day," and flag the two real problems for deeper work later. The "where is the Notion page" group gets a single Slack message in the channel with the link, pinned again, with a tiny note: "if you can't find this, that's on me. Let me know what you searched for so I can fix the title." That last sentence is the difference between an ops person who lasts and one who burns out blaming the team for not reading.
9:30am — Vendor Email Triage
You open the inbox and there it is. The renewal email from a vendor your team spends $48,000/year on. Auto-renew is in 11 days. Nobody flagged it. The sales AE on the other side has already booked a "quick chat" on your calendar for 2pm. Your stomach does the small sideways thing it does whenever you realize you are about to negotiate from a losing position.
This is not a one-off. This is most Tuesdays. Vendor renewals are the ops calendar that nobody told you you owned, and somehow you do anyway. Here's what's stacked up:
- Vendr or Tropic queue: four renewals in the next 60 days, two of them software you're not sure the team still uses. Tropic is showing you that one tool has 11 active seats out of 40 paid. That's $14,200 of shelfware.
- Legal redline: a DPA from a new analytics vendor has been sitting in your Legal counterpart's queue for two weeks. The procurement is blocked until it clears. Your Head of Marketing is asking for a status every other day.
- The "quick chat": the AE wants to upsell you to the enterprise tier. You know this because the email subject line was "Q2 partnership opportunity" and that phrase only ever means one thing.
You spend 25 minutes pulling usage data from the tool's admin panel, cross-referencing it with the seat list in your HRIS, and writing a one-page renewal brief. Tools used, seats active, dollar value, three negotiation levers (downsizing, multi-year discount, removing the modules nobody uses). You ship it to your CFO before the 2pm call so the conversation is "here's our position" instead of "let me get back to you."
This is the unglamorous reality of vendor management. It is not strategic sourcing. It is being the person who notices the auto-renew date on a contract that someone signed before you were even hired. The reason ops owns the vendor calendar is that nobody else's job description includes "remember when things expire." Yours does, even if it doesn't say so.
11:50am — Mid-Day Process Review
You have a 1:1 with the new hire on the customer success team. Three weeks in. She is running through the onboarding doc you wrote in February, the one you were proud of, the one that took a whole weekend.
She gets stuck on step 4.
Step 4 is "create your support inbox in Zendesk." Except in March, IT migrated the team from Zendesk to Intercom, and nobody told you, and the doc still says Zendesk. Step 5 references a Slack channel that was renamed in April. Step 7 has a screenshot of an admin panel that hasn't existed since the SaaS vendor pushed a UI refresh in January.
This is what process docs do. They rot. You write them once, ship them, feel good about it, and 90 days later they are quietly lying to every new hire who reads them. The new hire is too polite to tell you the doc is broken. She just says "oh, I figured it out, it's fine." It is not fine. The next person will hit the same wall and silently lose 40 minutes.
The honest reckoning here, the one no JD prepares you for: the doc is not the deliverable. The review cadence is the deliverable. A process doc with no owner and no quarterly review is a liability dressed up as documentation. You either build a review rhythm (quarterly, named owner, last-reviewed-on date in the header), or you stop pretending the doc exists.
You add three things to your week:
- A calendar block on the last Friday of every quarter labeled "process doc review" with the four onboarding docs linked.
- A Zapier rule that pings the doc owner in Slack 60 days after the last edit, asking "is this still accurate?" If they don't respond in 5 days, it pings their manager.
- A "last reviewed" line at the top of every doc. Not "last edited" (that's the wiki). Reviewed means a human read it end to end and confirmed the steps still work.
You make the changes to step 4 right there in the 1:1. The new hire watches you do it. That is also the job: letting people see that the doc is a living thing, not a monument.
1:15pm — Async with Finance, Legal, and IT
You eat a sandwich at your desk and look at three Slack threads, each one a different async conversation where you are the connective tissue.
Finance wants a PO for the analytics vendor. They cannot cut the PO until Legal clears the DPA. You have already poked Legal twice. You poke them a third time, this time with a clear deadline: "Marketing needs this live by May 6, can we land the redline by Thursday?" Specific deadlines move things. Vague urgency does not.
Legal has questions about a new vendor's data residency. They want to know if the vendor stores customer data in the EU. You don't know. You go to the vendor's trust portal, find the answer (Frankfurt and Dublin), screenshot it, paste it into the thread. Legal acks. The DPA moves to position one in their queue. This took you four minutes. It would have taken Legal 20 minutes because they don't know which vendor portals exist, which is why ops owns the brokering.
IT Slacks you at 1:47pm: "hey, we ran out of Notion licenses. All-hands is at 2pm and three people in marketing can't access the deck."
This is the license panic. Every ops person has lived through it. The pattern is always the same: someone added a batch of contractors to a tool last quarter, the seat count drifted past the paid tier, the bill went up by $200/month, nobody noticed, and now you have 13 minutes to fix it before a company-wide meeting.
You do the thing. Open the Notion admin, downgrade four dormant accounts (the contractors who left in March and somehow still have seats), free up the licenses, ping the three marketing people with the link to the deck. It works. You did not write a strategic memo. You did not drive operational excellence. You unblocked an all-hands. That, also, is the job.
The async pattern that makes this work, when it works: be the person who shortens the loop. Don't relay messages. Pull the data, paste the screenshot, propose the deadline, name the owner. Every message you send should reduce the number of messages someone else has to send. If you are forwarding without adding value, you are slowing the system down, not helping it.
3:40pm — The Scope Creep Meeting
A product manager pulls you into a 30-minute "quick sync" that turns into 50. He has a new initiative. He needs help with "the operational side." Twenty minutes in, you realize the scope: he wants you to own vendor selection for a new tool, build the rollout plan, train the team, write the SOP, and run the first 90 days of usage reporting. None of this was in the original ask. None of this has a budget owner. None of this has a deadline. He just genuinely thinks you have the time.
This is scope creep, and naming it out loud is the only way to handle it. You don't say no. You say: "happy to help with the rollout plan and the SOP. The vendor selection should sit with you and procurement, and I'll review the shortlist. Training and 90-day usage reporting are a separate project; let's scope that as a Q3 ask with its own budget." You write down what you committed to in the meeting notes. You send him the notes within an hour. If it is not in writing, it did not happen.
The Ops people who burn out are the ones who absorb every adjacent ask without bounding it. The ones who last are the ones who say yes to the right slice and put the rest in a Q3 backlog with a real owner.
5:20pm — End-of-Day Prioritization
You close the laptop with 22 things still open. This is normal. The instinct is to stay an extra hour and crush three more. The discipline is to stop.
Before you stop, you run a 10-minute filter: of the 22 open items, which three actually move the company forward tomorrow? Not which three are loudest. Not which three feel most urgent. Which three, if they don't get done, will cost the company time, money, or trust?
Today the three are:
- The vendor renewal brief for the CFO (deadline: 2pm tomorrow).
- The DPA chase email to Legal (every day this slips, marketing's launch slips).
- The onboarding doc fix you started at noon and didn't finish.
The other 19 items go into a "tomorrow" list in Asana with a one-line note each. Not because they don't matter. Because tomorrow's brain will have a cleaner read on them than tonight's tired one.
The reason ops people burn out is not the workload. It is the absence of a closing ritual. Every other function has a natural off-switch. A deal closes, a sprint ends, a campaign ships. Ops work has no edges. It just keeps going. So you have to build the edge yourself, and the Friday-afternoon habit that prevents Monday-morning panic is this: 30 minutes, every Friday, to walk the queue, kill the dead tickets, write the three things that matter most for next week, and close the laptop. That is not a productivity hack. It is the only way to do this job for more than two years.
What This Job Actually Is
This job is not operational excellence. It is operational adequacy held together by a person who notices things. The vendor renewal nobody else flagged. The doc that quietly stopped being true. The license panic at 1:47pm. The Slack thread where three teams are talking past each other and need a translator.
If you are reading this and feeling a little seen, a little tired, and a little defensive — good. You are not bad at ops. The job is genuinely this fragmented, and the people who pretend otherwise either don't actually do it or just got hired into something that hasn't broken yet.
The reason good ops people are impossible to replace is not that the work is hard in any one moment. It is that the work is composed of 200 small noticings, and the institutional knowledge of which noticings matter takes 12 to 18 months to build in any given company. You are not a Zapier flow. You are the part of the company that holds the seams together while everyone else builds the next thing. That is unglamorous. It is also why, when you leave, three different functions will quietly fall apart for a quarter.
Tomorrow morning the queue will have 14 new tickets in it, and you will sort them, and you will broker between Legal and Finance, and somewhere around 2pm somebody will Slack you "quick question." That is the job. Welcome to it.
