The Subscription Cull: Score Every Recurring Charge and Reclaim Your Money
Let Claude weigh every subscription on what you actually get from it β then tell you what to cancel, keep, or quietly downgrade, with the yearly saving in dollars.
Subscriptions are designed to be forgotten. That's the whole business model: a small enough monthly charge that it never quite crosses the threshold of "I should deal with this", multiplied across a dozen services, running quietly in the background of your bank account for years. You're almost certainly paying for something right now you'd struggle to name.
The reason it never gets fixed isn't laziness β it's that "should I keep this?" is a genuinely annoying question to answer for fifteen different services. Is it worth it? Compared to what? Am I on the wrong plan? Doing that maths for every charge, one by one, is exactly the kind of tedious task that never makes it to the top of a Sunday.
So hand it over. Claude will score every recurring charge on how much value you actually get per dollar, and sort them into keep, cancel, and downgrade β the last one being the category everyone forgets exists. You're not just cutting; you're right-sizing. Same life, smaller bill.
What You Need
- A free or Pro account at claude.ai β the free tier is plenty for this.
- A list of your recurring charges. Either export a few months of statements (CSV or PDF) so Claude can find them, or jot down what you know you pay for.
- 15 minutes, and honesty about what you actually use.
- Recommended: keep this in your "Money" Claude Project so you can re-run the cull every few months.
Step 1: Find every recurring charge (including the ones you've forgotten)
If you have statements, upload them and let Claude hunt. This catches the sneaky ones β the annual renewal, the free trial that quietly converted, the thing you signed up for in 2023.
You are a sharp financial analyst helping me audit my subscriptions. I've uploaded [NUMBER] months of statements. 1. Scan every transaction and identify all recurring or subscription-style charges β monthly, quarterly, and annual. Include the easy-to-miss ones: annual renewals, free trials that started charging, app-store subscriptions, and near-identical repeat charges. 2. For each, list: name, amount per charge, frequency, and the annualised cost. 3. Flag anything that looks like a duplicate, an overlap (e.g. two services that do the same job), or a charge you might not recognise. 4. Give me the total I'm spending on subscriptions per month and per year β I want the number that makes me wince. If you can't tell whether something is a subscription, list it separately and ask me. Use my real currency and exact figures.
What just happened: the invisible has a number now. Most people are genuinely shocked by their total annual subscription spend β because you never see it added up, only drip-fed a few dollars at a time. That wince is the whole point; it's what makes the next step worth doing.
Step 2: Score each one on value-per-use
This is the part that makes cancelling easy: not "do I like this?" but "what am I actually getting per dollar?". Answer Claude's questions honestly and let it score.
Now help me judge each subscription on real value, not vibes. For each recurring charge, I'll tell you honestly how often I actually use it and how much I'd miss it. [For each: NAME β how often I really use it (e.g. daily / a few times a month / haven't opened it in months) β how much I'd genuinely miss it (a lot / a bit / not at all).] Then: 1. Score each one on value-for-money using my usage and the cost β a simple High / Medium / Low. 2. Sort them into three buckets: KEEP (clearly worth it), CANCEL (low value or unused), and DOWNGRADE/RETHINK (I use it, but I may be on too expensive a plan, paying monthly instead of annually, or duplicating another service). 3. For the CANCEL bucket, total the annual saving. 4. Be honest with me β if I'm clearly keeping something out of guilt or "I might use it someday", say so plainly.
What just happened: every charge now has a verdict grounded in your real usage, not a guilty guess. And you've been shown the third option most people miss β downgrade. You don't have to cancel the thing you love; you might just be on the $20 plan when the $10 one does everything you need.
Step 3: Turn it into a five-minute action list
A verdict you don't act on saves nothing. Get the literal steps.
Turn my cull into a short action list I can knock out in one sitting. 1. For the CANCEL bucket: list each service and the literal first step to cancel it (where to go, what to click). Order them biggest-saving first so I feel the win early. 2. For the DOWNGRADE bucket: state the specific change (switch to annual billing, drop to the cheaper tier, etc.) and the saving for each. 3. Give me the grand total I'll save per year if I do everything on this list. 4. Add one sentence framing what that annual saving could go toward instead β as a goal, not a guilt trip. Keep it tight enough that I'll actually do it today.
What just happened: you've gone from a fuzzy sense that you're overpaying to a ranked, do-it-now list with a dollar figure at the bottom. Cancelling the biggest-saving one first means you feel the payoff immediately β which is what carries you through the rest of the list.
Bonus: The "worth it at renewal?" gatekeeper
The best time to judge a subscription is the moment before it renews. Set up a reusable check.
I'm about to renew (or was just charged for) [SERVICE] at [PRICE per period]. Play devil's advocate: ask me 3 quick questions about how I've actually used it since I last paid, then give me a straight verdict β renew, downgrade, or cancel β and your reasoning. If there's a cheaper plan or an annual option that'd save me money for the same use, tell me. Don't let me renew on autopilot.
Payoff: subscriptions leak money precisely because renewal is frictionless and thoughtless. Dropping a 60-second gate in front of each renewal turns autopilot spending into a real decision β the single most effective habit for keeping this from creeping back.
Bonus: The overlap detector
We pay twice for the same thing more often than we'd admit. Hunt the redundancy.
Look at my full subscription list and find redundancy β places where two or more services do overlapping jobs (streaming, cloud storage, music, productivity apps, news, fitness). For each overlap, explain what they duplicate, which one I seem to use more, and what I'd save by consolidating to one. Suggest the consolidation but leave the final call to me.
Payoff: overlap is invisible when you sign up for services one at a time over years. Seeing them side by side β "you're paying for three ways to store files" β surfaces savings that a service-by-service review completely misses.
Claude didn't decide what you value β you told it what you actually use and what you'd miss, and it did the tedious scoring and maths so the decisions got easy. You choose what stays, what goes, and what to downgrade; it just made the choice a two-minute one instead of a chore you keep avoiding. That's the idea: offload the friction, keep the judgement. Your money, your calls β now with the numbers in front of you instead of hidden in a dozen forgotten charges.
Ready to go deeper?
Build your own AI assistant in one afternoon.
A step-by-step guide to building a custom AI assistant that actually fits your life and work β no coding, no overwhelm, just a system you can use from day one.
Get the guide β $47Daily Stack
The tools behind the work.
The tools that have earned a permanent spot in my workflow.
Claude
Anthropic's flagship product and the tool that almost every resource on this site is designed for.
ManyChat
The easiest way to build automated Instagram DM flows, comment replies and lead capture without coding.
Planoly
My go-to for planning, scheduling and visually mapping out social content across platforms.