Twelve subscriptions. Half of them overlap. None of them talk to each other. You're not sure what two of them are for, and you're afraid to cancel them in case something breaks. That isn't a stack. It's a pile.
A soul-led business has a small number of real jobs. The tools should match the jobs, and stop there. This is how you get there from the pile.
Your stack is not the tools you've collected. It's the smallest set of tools that covers the jobs your business has. Everything past that is cost without return.
The pile happens for one reason. People buy tools before naming jobs. A post recommends a thing, it looks useful, the trial is free, and now it's a line on the card and a tab in the browser, doing a job that another tool was already doing.
Reverse the order and the problem mostly dissolves. Name the handful of jobs the business genuinely has. Then, and only then, pick one tool for each. The stack falls out of the jobs. It was never supposed to come first.
Tools serve the work. The moment the work is bending around the tools, the stack is running you.
One place that's yours, where the work lives and people can find it and act on it. A website. Not a website plus three separate landing-page tools plus a link-in-bio service all showing slightly different versions of you.
A way to reach the people who said yes, that doesn't depend on an algorithm letting you. Email. This is the one asset you should own outright, because it's the one that survives every platform you'll outgrow.
Money in, cleanly, without a tangle. One payment path that works. This job needs boring and reliable, not clever and feature-rich.
The actual handing-over of the work, plus a way to book it if the work is one-to-one. What this needs depends entirely on the offer, which is exactly why it's the job people most over-tool.
One tool for each job. The skill is not finding the best tool. It's holding the line at one.
One site that's yours. Owned beats rented over any timeline that matters, and that case is made in full in the website-after-templates guide. The owned build is the destination.
What to use until it's built, narrowed to ones you can leave with your work intact: Ghost if the work is writing-led, where every post exports as a clean file and the whole site is portable. Webflow if it's design-led and you want the markup to become yours later. Both beat the all-in-one builders that hold your pages hostage the day you outgrow them.
What you don't need: a website, plus a separate funnel builder, plus a landing-page tool, plus a link-in-bio page, each a subscription, each a slightly different front door. One door.
One platform that sends your email and holds your list. You own the list even when you rent the sender, which means you can export every contact and leave without losing them.
Three that hold up. Flodesk if you want it to look like you without fighting a builder, at a flat price that doesn't punish you for growing. Kit, formerly ConvertKit, if automations and selling from email matter more to you than design. Buttondown if you want it lightweight and your data plainly yours. All three let you export every contact and walk. That is the test, not the feature list.
What you don't need: an email tool, plus a separate newsletter tool, plus a CRM, plus a second list somewhere you forgot about. One list. One sender.
One way money comes in. Stripe, directly, for almost everyone. Payment Links and Checkout cover most offers with nothing extra layered on top, and most other tools you'd reach for sit on Stripe anyway, whether you see it or not. Add PayPal only if your audience genuinely won't pay without it. Keep this part boring on purpose. A second checkout system is a second place for money to break.
Match the tool to the offer, not the offer to the tool. A Google Doc or a PDF delivers more than people admit. A digital product or a small course: Gumroad or Podia before you reach for Kajabi, because the lock-in scales with the price you're paying. Don't buy a learning management system to deliver something that is, honestly, an email and a PDF.
Scheduling belongs here too, and only if the work is one-to-one. Cal.com if you want to own it, Calendly if you want it handled. One scheduler. If you don't do one-to-one, you've been paying for a tool you don't use.
A stack you could copy today and not resent in a year:
The custom build is the destination. Ghost or Webflow as the stopgap, because you can leave both with your work intact.
Flodesk or Kit. Buttondown if you want it lighter. Every one of them exportable on day one.
Directly. Payment Links and Checkout cover most offers, with nothing layered on top of it.
Email and a file for most things. Cal.com or Calendly only if the work is one to one.
Claude or ChatGPT. A single paid seat under the whole stack, not twelve bolt-ons across it.
Every tool you add is a tool you maintain. Add like it costs you something. It does.
For every subscription you pay for: what job does this do, and is anything else already doing it.
Go down the card statement, line by line. Most stacks have two or three tools doing a job a tool you already pay for also does. The overlap is invisible until you ask the question on purpose.
Cut the doublers. Keep the one that does the job best for how you work, not the one with the most features. Most piles shrink by a third on the first honest pass, and nothing breaks, because the job was always covered.
A tool you can't name a job for is not a tool. It's a habit with a monthly price.
Not every tool is equal. Some hold the things your business is made of. Those are the ones to think hardest about.
The home should be yours, not space you rent inside someone else's platform on their terms and their price increases. This is the highest-value thing to own outright, because everything else points at it.
You will rent the sending tool. Fine. What matters is that the list itself is yours: exportable, portable, leaveable. If a tool makes it hard to take your people with you, that difficulty is part of the price, and it's a high one.
Before you adopt anything that holds your content, your audience, or your money, ask how you would leave it. A tool you can't leave isn't a tool you chose. It's one that chose you, and the bill goes up from there.
Every tool in the pile now has an AI feature, and there's a new standalone AI tool launched since you started reading this. Most of it is the same plumbing with a different label.
One capable assistant, learned well, used across the unglamorous middle of the business: drafting, admin, the work behind the work. Claude or ChatGPT, a single paid seat. That beats twelve AI point-tools, each doing a sliver, each a subscription, none of them yours.
AI belongs underneath the jobs, helping you do them faster. It does not replace the judgment, the voice, or the stand. The AI & Voice guide covers where it helps and where it flattens you. Use it as plumbing. Stay the one deciding what runs through it.
There is always a new one, and it is always being called essential by someone who hasn't run a business on it for a year. The model is rarely your bottleneck. The pile is. Adding to it to feel current makes it worse.
One login that does everything sounds like simplicity. It is one company owning your site, your list, your payments, and your content at once, on terms you don't set, with an exit you can't afford. Simplicity you can't leave is a trap with good marketing.
Reorganizing the system is not doing the work, though it feels close enough to fool you for an afternoon. The best system is the smallest one you'll keep using without thinking about it.
The whole pile is this one mistake, repeated. If you can't say in a sentence what job a tool does that nothing else covers, you don't need it yet, and possibly at all.
Here is the whole thing, shorter than the pile. A business has a few real jobs. A tool for each. Own the ones that hold what the business is made of. Cut everything that can't name a job.
The pull is always toward more: more tools, more features, more tabs, more sense of being equipped. The work pulls the other way. The strongest stack is the one you stop noticing, because it's small enough to disappear behind the work.
You don't need the best tools. You need few of them, doing clear jobs, that you could walk away from if you had to.
Name the jobs first. Home, list, getting paid, delivering. The stack falls out of that.
One tool per job. Hold the line at one. The second one is almost always overlap.
The honest pass. What job, and is anything else already doing it. Cut the doublers.
Own the website and the list. Rent senders if you must. Never rent the people.
AI is plumbing. One capable tool underneath the jobs. Not twelve, not the product.