Viktor.comAn AI coworker that actually does the work
Promoted
Subscription Incinerator
Burn the subscriptions you forgot you had
21 followers
Burn the subscriptions you forgot you had
21 followers
Free tier tracks up to 10 subscriptions manually, forever, no credit card. Premium ($9/mo or $59/yr) unlocks Gmail auto-scan, bank statement import, and the full cancellation assistant. Would love feedback from this community — especially on what other services you'd want cancellation guides for. 🔥
Subscription Incinerator
@andrew_dainty Just gave it an upvote. I like that you solved a problem almost everyone has experienced but rarely addresses until money starts disappearing. Curious, how are you handling merchants that use inconsistent billing names across different payment methods?
Subscription Incinerator
@spotch_founder Thank you for the upvote. We use a two-layer system — a pattern-matching alias database for known merchants (handles things like NETFLIX.COM 800-123 vs NETFLIX*ABC123 mapping to the same service), and an AI fallback for anything unrecognized, which normalizes the name and scores its own confidence. The alias table grows over time as new variants get caught by the AI layer.
Went through the site before commenting, mostly the cancel guides since that is what you asked about. Marking Adobe as hard to cancel is accurate, their early termination fee catches everyone at least once.
Guides I would add next, based on where I see clients bleed money: LinkedIn Premium since the annual renewal ambushes people, Semrush whose cancellation flow is genuinely infamous in the SEO world, and GoDaddy domain plus hosting auto-renewals. A generic template for local gym style memberships would help too even without a brand attached.
One thing to fix quickly while launch traffic is flowing: your PH description says Gmail auto-scan is a Premium feature at 9 dollars a month, but the bottom of the cancel guides page says the Gmail and bank scan surfaces charges automatically for free. I could not tell which copy is right, and that is exactly the moment where someone decides whether to sign up.
Upvoted. Free manual tier forever with no card is the right trust move for a product that literally handles people's money leaks.
Subscription Incinerator
@abdullah_javaid3 Appreciate the thorough pass, especially catching that pricing contradiction — you're right, that's exactly the moment someone bounces. Just fixed it: cancel guides page now says $9/mo, matching the PH description. Good catch, thank you.
Adobe's early termination fee is brutal, glad the "Hard" flag reads as earned rather than lazy. LinkedIn Premium, Semrush, and GoDaddy are going on the guide list — Semrush's cancellation flow having a reputation in the SEO world specifically is a great signal that it's worth prioritizing.
The generic gym-membership template is a good call too, that's a whole category of "confusing local business subscription" that a single guide could cover well.
And yeah — free tier with no card required forever isn't a growth-hack trial, it's the actual point of the product. Thanks for the upvote.
@andrew_dainty Good, that was quick. The thing I would watch next is timing.
Right now the Gmail scan catches the receipt, which arrives after the money has already left. The email that actually saves someone is the one before it, the "your trial ends in 3 days" or "your plan renews on the 14th" notice. Every one of those has a deadline in it. If you parse that date and put a single reminder in front of someone while cancelling is still free, you move the product from telling people what they lost to stopping the loss.
That also fixes the AI tool pile you mentioned to Varun. Nobody forgets a 20 dollar charge because it is large, they forget because the trial email landed weeks before the invoice and nothing connected the two.
Question on the alias table, since you built the AI fallback to grow it: when the AI normalizes a merchant nobody has seen before, does that alias become available to every user straight away, or only after you review it? The first is a real network effect where every user makes the matcher better for the next one. Worth saying out loud on the landing page if it is true.
Subscription Incinerator
@abdullah_javaid3, This is a sharp reframe, thank you. You're right that the receipt is the wrong moment — by the time it lands, the decision's already made and all we can do is narrate it. The trial-end and renewal notices are the ones with actual leverage in them, because they arrive while cancelling is still free. Parsing that date and surfacing one reminder ahead of it moves the product from "here's what you lost" to "here's your last free exit" — that's a real difference in what the product is for, not just a UX tweak.
It's the same reason the AI-tool pile keeps happening: nobody misses a $20 charge because it's small, they miss it because the trial email and the invoice landed weeks apart with nothing linking them. Putting a reminder in that gap is the fix, not a bigger/louder receipt.
On the alias question: yes, the first case, and it's true today, not aspirational. When the AI normalizes a merchant it hasn't seen before, that mapping gets written straight into the shared table everyone's imports there's no review step and it isn't scoped to your account. So the next person whose statement has a similarly-formatted line item benefits immediately, without an AI call even running for them. You're right that it's worth saying plainly on the landing page — "every merchant we learn, we learn for everyone" is a real claim here, not marketing language.
Appreciate you digging into this — the pre-charge angle is going on the roadmap.
@andrew_dainty Good, and thanks for answering that straight rather than aspirationally.
One risk worth naming, since you said there is no review step and it is not scoped to the account. That cuts both ways. A correct inference reaches everyone instantly, which is the network effect. A wrong one reaches everyone instantly too, and nobody downstream has any way to know the mapping is wrong. One bad normalization that merges two genuinely different merchants shows up as a subscription someone does not have, or hides one they do.
You probably do not need a review queue to handle it. The confidence score is already there, so the cheap version is to only write to the shared table above a threshold and keep everything below it account local until a second independent inference agrees. Same speed for the common case, and one bad call cannot reach the whole user base.
Worth doing because this particular failure is silent. A missed subscription does not file a support ticket, it just keeps charging.
The cancellation walkthroughs for stuff like Adobe and that gym membership I forgot about were surprisingly thorough, saved me a solid hour of digging through help docs. Would love one for SiriusXM next.
Subscription Incinerator
@can4ojy Adobe's early-termination maze is exactly the kind of thing worth writing up in detail, glad it actually saved you the digging.
SiriusXM's a good add to the list — their retention flow (the multiple "wait, one more offer" transfers) is exactly the kind of thing a step-by-step guide is built for. Noting it for the next batch of guides.
Subscription Incinerator
@can4ojy Just added it! SiriusXM's now up at /cancel/siriusxm — covers the retention-chat dance and what to do if they won't let you cancel online. Thanks for the nudge.
Honestly the Gmail auto-scan sounds great but have you thought about adding a shared household mode? like my partner and i both pay for stuff we forget about, and it'd be super helpful to see one combined view of everything we're burning money on together.
Subscription Incinerator
@alperenkerepmm That's a real gap — right now everything's scoped to one account, so you and your partner would each need to track separately and mentally merge the totals, which defeats half the point if the goal is seeing the combined bleed.
Shared/household view would mean rethinking the data model (subscriptions currently belong to exactly one user, no grouping), so it's a bigger lift than the other requests in this thread, but it's a legitimate one — a lot of money leaks specifically because two people each assume the other is tracking it. Adding it to the list as a real feature ask, not just a nice-to-have.
Subscription Incinerator
@alperenkerepmm An amazing request, we just added Household sharing 🙌
You can invite a partner (or roommate, family member, whoever) from Settings, and once they accept:
- You both see one combined dashboard of all your subscriptions — no more "wait, are we still paying for that?"
- If you're on Premium, they get Premium too, at no extra cost
- Everyone keeps their own login, their own data — it just merges into one shared view
Takes about 30 seconds to set up. Let us know if you run into anything!
The name alone is genuinely funny, but the fact that you built a working free tier for manual tracking before gating the Gmail scanner shows you actually understand your early adopters. Curious how the cancellation assistant handles services that require phone calls though.
Subscription Incinerator
@yeim7obx Thanks, glad the name lands. On sequencing — yeah, manual tracking free-forever came first on purpose, gating the scanner felt like the right line since that's the part that costs money to run.
Phone-call cancellations: honestly, the assistant doesn't place or script the call for you. What it does is give you step-by-step guide text (what to say, what they'll push back with, what to ask for) plus a backup email draft where the service accepts one.
Anything that's call-only — Adobe's early termination path is the classic example — just gets flagged "Hard" difficulty with a reason so you know what you're walking into before you dial. So: documents it, doesn't automate it. That's a real gap, not oversold as solved — automating the call itself is on the list but not built yet.
Funny enough I've got a separate project, everyring.ai, that's an AI voice agent — though it's built the other direction, answering missed calls for businesses rather than placing outbound ones. Different problem, same underlying voice-agent tech, so there's a real path from "documents the call" to "makes the call" eventually — just not conflating the two before it's actually built.
NINA
The hardest part about subscriptions isn't paying for them, it's forgetting they exist. I like that you're focusing on helping people actually cancel services instead of just showing another dashboard full of recurring charges.
Subscription Incinerator
@varun1jan That's exactly the failure mode we're built around — a dashboard just restates the problem back at you, it doesn't fix it. The whole point was making "I found this, now what" not be a dead end: here's the guide, here's what to say, here's the email if you'd rather not call.
Forgetting they exist is the actual disease; a list is just a symptom tracker.
The pattern I see constantly right now is AI tool trials — everyone's stacked up 3-4 of them (an LLM chat app, a coding assistant, an image generator, whatever) from a free-trial-turned-$20/mo moment, and completely lost track since they don't show up on a bank statement the way "NETFLIX" does. Half the point of the Gmail/bank scan is catching those before they've quietly renewed five times.
One more piece in that "now what" direction: when you cancel something, we also surface an open-source/self-hosted alternative if one exists — Spotify: Navidrome or Funkwhale, Notion: AppFlowy, Adobe Creative Cloud: GIMP/Inkscape/Kdenlive, that kind of thing. Not for everyone, but for the subset of people cancelling because they resent the subscription model itself rather than the price, it's the actual off-ramp.
One thing I'd love is the ability to detect duplicate or overlapping subscriptions (like multiple cloud storage plans) and flag the ones I'm barely using. A monthly "you only used this twice in 30 days" nudge would make this way more valuable than just listing charges.
Subscription Incinerator
@rojin34058 Duplicate/overlap detection — building it right now, this pushed it up the list. Since we already group subscriptions by category (cloud storage, streaming, etc.) from the Gmail/bank scan, flagging "you have 3 things in Cloud Storage" doesn't need any new permissions, just a query against data we already have. Should ship shortly.
Usage nudges are the harder one — we don't have API access into Spotify/Netflix/Adobe to see if you actually opened the app, and building that per-service would mean asking for a bunch more account access, which cuts against the whole "read-only, minimal trust footprint" thing the free tier is built on.
Realistic first version is self-reported: a lightweight monthly "did you actually use this?" check-in instead of pulling real usage data. Not as sharp as "twice in 30 days," but doesn't require you to hand over more keys than you already have to. What do you think?