Every year I rebuild my ops stack on paper, even when I change nothing. I list every tool I pay for, what job it does, and what I’d feel if it disappeared tomorrow. Anything that gets a shrug gets cancelled. This January the exercise killed four subscriptions and confirmed seven keepers, and enough people asked about the resulting list that I’m writing it up.
Context matters for stack advice, so here’s mine: I run a one-person digital business. Products, services, a newsletter, a few hundred customers, no employees, occasional contractors. If you’re a funded startup, your constraints are different and this list will look quaint. If you’re a solo operator, your constraint is the same as mine: every tool has to either make money, save meaningful time, or prevent a disaster, because the person paying for it and the person maintaining it are the same tired individual.
The stack, job by job
Money in: Stripe. Payments, subscriptions, invoicing for the service side. The fees sting and I grumble every month, but the alternative is being my own billing department, and I have watched founders lose entire weekends to reconciling payments from cheaper, jankier setups. Disaster prevention, purchased monthly.
Money tracked: a real bookkeeping tool, from day one. I use one of the mainstream small-business accounting apps; which one matters less than the habit. The founders who “do the books at tax time” are doing forensic accounting, not bookkeeping. Thirty minutes a week, categorized as it happens, and January stops being frightening.
Words out: a newsletter platform with automation. The newsletter is the business’s front door, so this is one of the two places I pay for the good tier without flinching. Sequences, segmentation, and a clean editor. The lesson learned the hard way: the platform is replaceable, the list is not, so export subscribers monthly to cold storage.
Everything scheduled: a plain calendar plus a booking link. No standing meetings with myself, no productivity-system cosplay. A booking page killed the “what time works for you” email chain, which I conservatively estimate was 5 percent of my working life.
Operations: one grid to rule the pipeline. Here’s the spine of the whole thing. Client work, product launches, content calendar, contractor tasks, renewal dates: it all lives in Wisegrid, one sheet per area, one dashboard-ish overview sheet at the top. I chose it for a reason that sounds unserious but is actually the entire decision: it works like a spreadsheet. I have twenty years of spreadsheet reflexes and zero desire to learn a proprietary metaphor with boards and pods and pulses. The grid does what my reflexes expect, and then adds the parts a spreadsheet was never going to do for me.
Two of those parts earn its seat. First, the AI formulas. My inbound leads land in a sheet, and a =CLASSIFY formula tags each one as product, services, partnership, or spam before I ever read it; a =SUMMARIZE column turns long inquiry emails into one-line cells I can scan over coffee. It’s the closest thing I have to an assistant. The AI usage is metered in dollars with a hard ceiling I set, which is exactly the pricing shape a solo founder needs, because “surprise usage bill” is a phrase that belongs in horror fiction. Second, the seat math: $19 for my editor seat, and the contractors and clients who just need to see status get free viewer access. I’m not paying $19 every time I want to show someone a sheet. There’s a 7-day trial if you want to kick the tires; I moved my lead sheet in first, since that’s where the AI formulas prove themselves fastest.
Files: boring cloud storage, rigorously named. Nothing interesting to report, which is the goal. A naming convention beats a search feature.
Passwords and access: a password manager, non-negotiable. The disaster this prevents is not hypothetical. Solo founders are one phished login away from losing the business, and we have no IT department standing behind us. This is the other tier I pay for without flinching.
The rules that keep the stack small
The list above is seven jobs. The discipline is in what’s not on it, and after years of subscribing and cancelling, I’ve compressed that discipline into four buying rules.
Rule 1: One tool per job, one job per tool. The moment two tools overlap, one of them is already dying; cancel it deliberately instead of letting it linger at $23 a month. The moment one tool does three jobs, you’re one pricing change away from a hostage situation.
Rule 2: The tool must survive the Tuesday test. Not “is this impressive,” but “will I actually open this on a random Tuesday in month four?” Most productivity purchases fail this test by week three. I now trial everything against a real week of work before the card goes in, and I read reviewers who actually test things rather than reword press releases; sites like Digital Filter Sales do hands-on writeups of business software, which beats vendor landing pages for figuring out what a tool is like on that Tuesday.
Rule 3: Data must be exportable, and I must have actually tried the export. Every tool in the stack holds business-critical data, and every vendor is one acquisition away from a “sunset announcement.” I do a quarterly export drill: subscribers, bookkeeping data, the ops sheets, everything, into a dated folder. It has saved me once, which pays for every drill I’ll ever run.
Rule 4: Total stack cost gets reviewed as one number. Individually, every subscription is defensible; that’s how they get you. Summed, my stack is a line item I compare against revenue every quarter. The 2026 number is under $200 a month for everything above, which for a business that runs on it is the cheapest employee I will ever have.
What I dropped this year
For honesty’s sake, the four cancellations: a social scheduling tool (posting manually made me post more thoughtfully and, oddly, more often), a fancy analytics dashboard (the built-in analytics on my platforms answered every question I actually asked), an AI writing subscription (the AI I use now lives inside tools I already pay for, like those formula cells, instead of being a destination), and a project management tool I’d kept out of guilt because I’d paid annually. That last one hurt to admit. Sunk cost is a subscription’s best friend.
The meta-lesson after years of this: a solo founder’s stack is not a collection of tools, it’s a set of decisions you’ve made once so you don’t make them daily. Every job on the list is something that used to interrupt me and now doesn’t. That’s the only metric that matters. Build the smallest stack that makes your Tuesdays quiet, review it once a year on paper, and spend the attention you save on the actual business.