ChatGPT Can Identify $300 in Potential Annual Subscription Savings
Ethan Bloch’s examples show ChatGPT helping users assess a purchase, identify recurring charges to reconsider and test budget changes, while leaving the choices to them. The workflows turn product listings, subscription records and spending data into comparisons or proposed adjustments, rather than final buying, cancellation or budgeting decisions.

A product comparison helps decide what to investigate before buying
A product screenshot gives ChatGPT a concrete item to compare. In Ethan Bloch’s example, the listing is for a Nookvale Supply Alder Mesh Desk Chair priced at $199.99. He asks whether it is a good deal and whether the same or a similar product is available for less.
The response names two alternatives: a Norwood Flex Mesh Office Chair for $169.99 and an Everwell Comfort Pro Chair for $174.99. That comparison gives the buyer options to consider, but the next question addresses a different decision: what might be worth checking about the original chair before purchasing it? ChatGPT lists armrest hinges, lumbar support and mesh durability.
Bloch says this kind of follow-up could surface common complaints or concerns about how well a product holds up over time. The point is not only to find a lower price, but also to identify issues a buyer may want to investigate when deciding whether something is well-made or likely to last. He says the approach could be useful whether buying for oneself or as a gift.
The example therefore separates two questions that are easy to conflate: whether another option costs less, and what to watch out for in the item already under consideration. The demonstration shows a product image as the starting point and a follow-up question about potential drawbacks. It does not present the alternatives or concerns as a final buying decision; they are information for the user to review.
A subscription search helps decide which charges to keep
Recurring charges call for a different kind of review. Rather than comparing one purchase with alternatives, the task is to assemble charges from information that may be in bank transactions or email. Bloch describes two routes: securely connect a bank account through ChatGPT Finances, or connect email so ChatGPT can search for subscription receipts and renewal notices.
A request such as “list all my recurring subscriptions” can turn the available information into a list to review. In the demonstration, that list includes a streaming service at $16 a month, a music service at $11, a fitness app at $12, a productivity app at $13 and cloud storage at $3. The total is $55 a month, or $660 a year.
That total is a starting point, not a recommendation to cancel everything. Bloch’s example narrows the question to the fitness app and productivity app, then asks how much canceling those two would save. Using transactions it finds, ChatGPT calculates a combined cost of $25 a month, or $300 a year. The user can then go to each service to cancel the subscriptions they no longer need.
The useful decision here is which recurring charges are no longer worth paying for. The list makes the charges visible; the follow-up calculation puts a possible cancellation in annual terms. The choice of what to keep remains with the user.
A budget revision makes the cost of a savings target visible
A budget request shifts the focus from identifying particular products or charges to balancing spending categories. Bloch describes asking ChatGPT Finances to look at his spending and help build a monthly budget that saves more while leaving room for dinner with friends or a hobby.
The suggested budget in the demonstration starts with monthly income of $4,200 and expenses of $3,650, leaving potential savings of $550. Its expense categories are rent at $1,800, utilities at $180, groceries at $450, transportation at $220, dining out at $300, subscriptions at $50, shopping and hobbies at $250, and other expenses at $400. Bloch says the budget is based on his actual spending and on priorities ChatGPT knows are important to him.
The follow-up makes the trade-off more specific. Rather than asking for a general reduction, Bloch asks, “Find me another $50 in my budget.” ChatGPT proposes reducing dining out from $300 to $275 and shopping and hobbies from $250 to $225. Those two adjustments make up the requested $50. The revised total expenses shown are $3,600.
This example treats a budget as adjustable rather than as a fixed set of cuts. The user can ask for a savings target, see which categories the suggested changes affect, and decide whether those changes fit their routine. Bloch describes the aim as working toward saving more with a plan that feels realistic.
The three workflows support different kinds of decisions: what to investigate before buying, which recurring charges may be worth canceling, and where a budget might make room for more savings. In each case, ChatGPT provides information or a proposed adjustment for the user to review and act on.