What four years of agency life and one Sunday with the credit card statement actually look like.

$127.38. That’s what my Mac costs me every month. I didn’t know that number until last Sunday, which is the embarrassing part.

I’d been sitting at the kitchen table with the second coffee, half-watching the rain, and I pulled up two years of statements because I’d decided to “tidy my finances.” The phrase “tidy my finances” is doing a lot of work in that sentence. What I actually did was discover that three of the apps charging me every month had not been opened on this laptop in over a year. One of them I’d genuinely forgotten existed.

So I made a list. All of it. Every single piece of software my MacBook Pro pays for, organised by how often I actually open it.

16 apps. $127.38 a month. $1,528.56 a year.

Here’s what I learned, what I’m keeping, and what I’m cancelling on Sunday. Maybe.

TLDR

• 16 active app subscriptions or licences on my Mac in May 2026
• Total monthly spend: $127.38 (roughly $1,528 a year)
• Six apps I actually use daily, four most weeks, three I pay for and barely open, and three I genuinely forgot about
• The three I forgot cost me $371.76 a year between them, which is an okay weekend in Lisbon I didn’t take
• I work in marketing at an agency, manage 12 client accounts, and take four or five client calls most days, so my stack skews toward email, calendar, screenshots, and notes more than design or dev tools
• I pay for everything at full price out of my own pocket. No discount codes, no comps, no affiliate links in this piece.

A note before we start

I’m not affiliated with any of the apps below. Nobody paid me to write this, nobody sent me a free year, and I don’t have referral links to share. I bought every one of these subscriptions with my own card, mostly while half-listening to a client on Zoom. If I recommend something it’s because it has earned a place in my Sunday-morning audit and survived. If I tear into something, it’s because I’m still annoyed I’m paying for it.

USD throughout, because that’s how I get charged.

The daily drivers: 6 apps, $26.91/month

These are the apps I open before I’ve finished my first coffee. If you took them off my Mac on a Monday morning, I’d have a problem by 9:15.

Slack ($0, agency-paid, four years)

Slack is free for me because the agency pays for it. I’d never volunteer for this. I open Slack 74 times a day. I counted last Tuesday because I wanted to know, and I have not been the same since.

The reason it’s daily and not weekly is that 12 client accounts means 12 Slack Connect channels, plus internal teams for each account, plus the general “what is everyone watching tonight” room I mute and then unmute and then mute again. I don’t have a productive relationship with Slack. I have a marriage with Slack.

If your job involves talking to people who aren’t in your building, this is the cost of admission. The cost just isn’t on my card.

Mimestream ($4.17/month, billed annually, two years)

Mimestream is a native macOS email client built specifically for Gmail. I switched off the Gmail web app two years ago and have never opened it on my Mac since.

Why this one matters: I get roughly 90 emails before 9:30am most weekdays, and 312 client emails came in last month, 90 of them follow-ups. The native app is faster, keyboard-driven, and doesn’t sit in a browser tab fighting for my attention with seven other browser tabs. It supports Gmail labels and filters properly, unlike most third-party clients, which is the entire reason I pay for it.

Annual licence is $49.99, which works out to $4.17 a month. It’s the smallest line item in my daily stack and the one I’d defend hardest.

Fantastical Premium ($4.75/month, year 4)

Fantastical is the calendar I should probably have stopped paying for. I have a complicated relationship with it. Some weeks I think it’s the best calendar app on the Mac, and some weeks I think Apple Calendar would do the same job for free if I gave it a chance.

What it does that Apple Calendar still doesn’t: natural language event entry that doesn’t trip over British date formats, decent integration with the conferencing dropdown (Zoom, Google Meet, Teams), and a menu bar widget that shows me my next three things without my having to click anything. For four or five client calls a day, that menu bar widget alone is worth keeping.

$56.99 a year, which is $4.75 a month. Year four of paying for this. I am not optimistic I will cancel in year five.

Raycast Pro ($8/month, eight months)

Raycast is the only paid app on this list that’s genuinely changed how I use my Mac. I open it 40+ times a day. It’s replaced Spotlight, my clipboard manager, my window manager, my emoji picker, my snippet tool, and the calculator. The Pro tier gives me the AI commands, which I use mostly to draft three-sentence Slack replies that I then rewrite to sound less like a robot.

Before this, I used Alfred for six years. Switching wasn’t dramatic. It was just that Raycast’s extensions ecosystem had quietly become better than Alfred’s workflows for the things I actually do. I’m not in the Raycast vs Alfred war (the people in that war on Reddit are intense). I just defected.

$8 a month, on a monthly plan because I wasn’t sure I’d stick with it. Now I’m sure. I should switch to annual and stop paying the premium. Sunday’s problem.

CleanShot X Cloud ($9.99/month, two and a half years)

CleanShot is a screenshot app, which sounds like the most boring possible thing to pay $9.99 a month for. I take roughly 20 screenshots a day. Most of them are of client comments, design feedback, or “see attached” replies where the attachment didn’t come through.

The cloud bit is the part I pay for. I capture a screenshot, it uploads automatically, I get a short URL in my clipboard, I paste it into Slack or an email. That whole loop is about three seconds. Without it, the loop involves a Finder window, a Drive upload, a share permission dialogue, and a sigh.

Annual is cheaper at $96/year (about $8/month), which I keep meaning to switch to. I’m currently paying the monthly tax because I never quite get round to it. Sunday’s problem, also.

Apple Notes ($0, since I bought the Mac)

Apple Notes is the only app on my Mac that’s never asked me for anything. I love it the way you love a colleague who just gets on with it.

I use it badly. I have 411 notes, most of them untitled. I search rather than organise. I have a note that’s just been called “Notes” since 2022 that contains 38 unrelated lines of text. I pay nothing for this. I keep meaning to “fix my note-taking” by paying for Bear or Notion or Obsidian, and I’ll get to that in the theoretical bucket.

The weekly stack: 4 apps, $36.99/month

These are the apps I open most weeks but not every day. The line between this and the daily bucket is genuinely blurry on some apps and clear on others.

Keynote ($0, came with the Mac)

I build roughly three client decks a week in Keynote. It’s free. It opens fast. It exports to PDF cleanly. It doesn’t have the corporate veneer of PowerPoint or the layout-fight of Google Slides.

I considered Pitch and Tome and Beautiful.ai over the past two years. I tried Pitch for a month and it was fine. None of them solved a problem Keynote wasn’t already solving for me, so I stopped trying.

1Password Individual ($2.99/month, four years)

Password manager. I have 287 entries. It works. There is nothing more boring or more obviously worth $2.99 a month than a password manager that works.

I switched from LastPass after their breach in late 2022, like roughly half the internet. I haven’t thought about it since, which is the highest compliment I can pay a piece of software.

ChatGPT Plus ($20/month, two years)

ChatGPT is my reluctant default. I use it for first drafts of cold outreach, restructuring messy meeting notes into proper recaps, summarising 40-message Slack threads I missed while I was on a call, and asking it to find the diplomatic way to say “no” to a client request that’s been creeping for three weeks.

I am not a power user. I don’t have a dozen custom GPTs. I don’t have a project-based workflow. I open the app, paste in the thing, ask the question, copy the answer, paste it somewhere else, and rewrite about 40% of it so it sounds like me.

$20 a month. The most reliable line item on this list. Also the one I feel the most ambient guilt about, which I’ll come back to.

Granola Business ($14/month, six months)

Granola is the meeting note tool that finally stuck after I’d tried four others over the past three years. It runs in the background during a call, transcribes, and produces structured notes after. It doesn’t join the meeting as a visible bot, which my clients appreciate more than they’ll ever say out loud.

The pricing oddity worth noting: their Business plan at $14 per user per month is actually cheaper than the Individual plan at $18. I’m a one-person “team” because I noticed this in the signup flow and refused to pay more for fewer features. (Granola, if you’re reading this, you know what you did.)

Is it perfect? No. I edit the structured summary fairly heavily every time, and the action items often need rewording before I’d send them to a client. But it’s the only meeting tool I’ve used where the friction of having it on is lower than the friction of not having it on. That’s the bar.

The “theoretically” bucket: 3 apps, $32.50/month

These are apps I pay for, open occasionally, and keep around because I genuinely believe I will use them properly one day. I have believed this for somewhere between four months and two years per app. The evidence is not on my side.

Bear Pro ($2.50/month, billed annually, eighteen months)

Bear is the markdown notes app I bought because Apple Notes felt undignified. I open Bear about once every two weeks, write a long, well-formatted note in it, feel briefly virtuous, and then forget it exists for the next twelve days.

$29.99 a year. I am not cancelling it because the moment I cancel it I’ll need it. This is exactly the irrationality these monthly subscriptions are designed to harvest.

Notion Plus ($10/month, four months)

I joined Notion in January to build a second brain. I’m not sure my first one was ready for the upgrade.

What I have in Notion: one personal CRM with three contacts in it, one habit tracker I updated for nine days in February, one half-built reading list, and one page called “ideas” with a single bullet that just says “podcast?” with a question mark.

What I’m getting for $10 a month: optimism. I’m not cancelling this either. By June it’ll either be a working system or a more expensive failure. The optionality feels worth $10.

Claude Pro ($20/month, three months)

I pay for both ChatGPT and Claude. I open ChatGPT first. I’m working on it.

The honest answer for why I keep Claude is that it’s better at the writing-shaped parts of my job. When I need to take a four-page client brief and turn it into a one-page strategy summary, Claude’s drafts need less rewriting than ChatGPT’s. When I need a quick caption or a punchy line, ChatGPT gets there faster. So I keep both, and resent paying for both, and then keep paying for both.

This is the line item on this list I’m least sure about. Ask me again in three months.

“I just found out I’m still paying for this”: 3 apps, $30.98/month

Here’s where the audit got embarrassing. These three I genuinely did not realise were still on my card until Sunday.

Dropbox Plus ($11.99/month, last opened 14 months ago)

I have not opened Dropbox in 14 months. Dropbox has opened my credit card every month since.

I migrated to Google Drive around the start of 2025 because the agency standardised on it. I never went back to clean out my Dropbox or downgrade the plan. The annual renewal hit in March 2025, and again in March 2026, and I just didn’t notice either time. That’s $287.76 for two years of storing files I have not looked at.

This is going on Sunday’s cancellation list. I’m sure of this one.

Toggl Track Premium ($9/month, never used)

This one is the punchline. I have never tracked an hour in Toggl. My agency runs on Harvest. I signed up for Toggl in 2023 because I was going to “track my own time properly” outside the agency’s system, and I have not opened the app on any device.

$108 a year for a productivity tool I have produced precisely zero data with. There is a podcast somewhere about this exact pattern. I should listen to it. After I cancel Toggl.

Adobe Lightroom ($9.99/month, one personal project in 2024)

In summer 2024 I went through a phase where I was going to “get back into photography.” I edited maybe 30 photos. The phase lasted six weeks. The subscription has lasted twenty months.

Adobe makes cancelling deliberately tedious, which is its own argument for never starting an Adobe subscription in the first place. I am promising myself I’ll cancel it on Sunday. Whether you believe me is up to you.

The math

Annual: $1,528.56.

The forgotten bucket alone is $371.76 a year, or roughly the cost of a Friday-to-Monday in Lisbon I keep telling my partner we should book.

The honest read on the rest: the daily six and the weekly four (ten apps, $63.90/month) are doing real work for real money. The theoretical three are an aspiration tax I’m choosing to pay. The forgotten three are just a slow leak I let happen, which I cannot in good conscience defend.

Who this is for

You will probably recognise this stack if:

• You work in a client-facing role with a calendar full of recurring meetings
• You live mostly in email, Slack, and slides, not in code or design files
• Your Mac is your primary work device and not a secondary machine
• You have, at some point in the last six months, paid for an app you no longer remember signing up for

Who should skip

This is not the stack for you if:

• You’re a designer, developer, or video editor whose paid apps would be a different (and probably more expensive) list entirely
• You’re a Windows or Linux user who landed here from a search result, in which case sorry about that
• You’re looking for a curated “best apps” list rather than a real one with admissions of waste

FAQ

How did you actually pull this list together? Two years of credit card statements in a spreadsheet, sorted by merchant. I cross-referenced anything that looked like a software charge against my Applications folder and my browser bookmarks. The whole thing took me about ninety minutes, including being annoyed at past me for thirty of those minutes.

Is $127 a month a lot for one person’s Mac apps? Honestly, no. The average knowledge worker pays for somewhere between eight and fifteen active subscriptions across all categories, and software is usually the largest one. What’s striking isn’t the total. It’s that 24% of it (the forgotten bucket) was buying me nothing. I’d rather pay $96 a month for ten apps I use than $127 for sixteen apps I half-remember.

Why pay for both ChatGPT and Claude? Different strengths. ChatGPT for fast, transactional asks. Claude for longer-form writing where I want the draft to need less editing. Whether that’s worth $40 a month between them is a real question and one I haven’t fully answered yet.

Why not just use Setapp? I’ve considered Setapp roughly once a year for three years. Several of the apps I’d want from it (CleanShot X, Bear) are on it. Several of the apps I rely on most (Raycast, Fantastical, Mimestream) are not. The math doesn’t quite shake out for my specific stack, but it’s a fair option if your daily drivers happen to be on the Setapp roster.

Did you actually cancel the forgotten subscriptions? At the time of publishing, no. I’ve blocked out Sunday morning. I’ll add an update to this piece in a couple of weeks once I’ve done it. If I haven’t, you’re allowed to be disappointed in me.

What about Reminders, Things, or Todoist? You didn’t mention a task manager. I use Apple Reminders for the small stuff and a shared Notion-adjacent system inside the agency for actual project tracking. I tried Things 3 ($50, one-time) twice. Both times I drifted off it within a month. I’m not against it. My brain just doesn’t seem to want one in this slot right now.

Anything you’d add if money weren’t a factor? Maybe Arc or Dia for browsing, but I haven’t found a strong enough reason to leave Chrome and the 14 logged-in client tools I have open in it. Maybe a proper read-it-later setup. Mostly though, the answer is no. The instinct to add another app is usually the instinct to avoid the work the existing apps could already do.

The point of the audit wasn’t to brag about how well-curated my stack is. It clearly isn’t. The point was to put a number on it so I’d have to look at the number.

$127.38 a month. $1,528.56 a year. $371.76 of that is buying me literally nothing. The rest is doing real work, most of it well.

I’ll let you know what happens on Sunday.