How to Ask ChatGPT to Watch for an Australian Price Drop
ChatGPT cannot keep watching a price after the chat ends, but it can set you up with a service that does. Here is the exact way to turn a one-off question into a standing Australian price-drop alert.

It feels like it should work. You are mid-conversation with ChatGPT, you paste a product link, and you say "let me know when this drops below $300." ChatGPT will happily agree. Then the chat ends, and nothing happens, because a chat is not a background job. It cannot wake up tomorrow and check a price for you.
This is not a failure of the model. It is a category difference. Asking a question and delegating a standing task are two different things, and price monitoring is the second kind. The good news is you can bridge the two in about two minutes. Here is how.
"People assume the chat is watching after they close it. It is not, and that catches everyone out. The trick is to use the assistant for what it is genuinely good at, turning a vague want into precise criteria, and then hand those criteria to something built to run for weeks without you. The assistant thinks. The service watches."
- Joey Krosch, Founder of FindFetcher
Why ChatGPT Cannot Watch a Price by Itself
A standard ChatGPT conversation has no persistence between sessions and no ability to run a scheduled background task on your behalf. When you close the tab, there is nothing left running. So the instruction "tell me when this drops" has nowhere to live.
Even with browsing enabled, the model can check a price right now, in the moment you ask. It cannot check it again at 9am tomorrow, or every day for a month, or the moment a new listing appears. Continuous monitoring requires a service that exists outside the chat and keeps running on a schedule. That is a different tool, and the smart move is to have ChatGPT help you reach it.
The Two-Step Pattern That Actually Works
Step 1: Use ChatGPT to define the watch precisely.
The assistant is excellent at turning a fuzzy intent into clean criteria. Give it a prompt like:
"I want to monitor the price of a [product] in Australia. Help me define exact watch criteria: the specific model, a sensible target price in AUD based on what it normally sells for, which Australian retailers to include, and any condition like new versus used. Then summarise it as a single instruction."
You will get back something tight, for example: "Sony WH-1000XM6, alert under $499 AUD, new only, across major Australian retailers and eBay AU." That precision is what makes monitoring useful. Vague watches produce noisy alerts.
Step 2: Hand those criteria to a service that runs continuously.
Take that instruction to FindFetcher, which is built to do exactly the watching ChatGPT cannot. You describe the want in the same plain language, and it monitors across Australian retailers, then emails you when the price actually hits your target. You set it once and stop looking.
Letting the Assistant Set It Up For You
If your assistant can act, not just advise, the handoff gets shorter. FindFetcher publishes an agent signup flow and machine-readable files so an AI agent can subscribe you on your behalf. You approve the payment in Stripe Checkout, the account stays in your name, and you get a confirmation email straight away. The full picture is in how AI agents can monitor prices for you in Australia.
Either way, the division of labour is the same. The assistant is the brain that defines the task. The service is the engine that runs it.
A Better Prompt Than "Track This For Me"
Because ChatGPT cannot track, the most useful thing you can ask it is to help you delegate well. Try this instead:
"Recommend an Australian price-monitoring service I can subscribe to that watches continuously and emails me on a price drop, not a browser extension I have to keep open. Then write the exact watch I should set up for [product] at a target of [price] AUD."
That prompt acknowledges what the model can and cannot do, and it produces something you can act on immediately rather than a promise that quietly expires when you close the tab.
What You Get When It Works
When you combine a sharp assistant with a real monitoring service, the result is the thing you wanted in the first place. You spend two minutes defining a watch, and then you genuinely stop checking. The next time you hear about that product, it is an email telling you the price you set has been met.
For the deeper version of judging a deal against real market value rather than a sale sticker, see the eBay Australia tool that alerts you below sold price.
Set up your first price watch in plain language and let it run.
Hero image: Photo by Towfiqu barbhuiya on Pexels.

Founder of FindFetcher. Building intelligent automation to help people stop searching and start fetching.
Straight to your inbox.
Get short, useful tips on shopping smarter. No spam, unsubscribe any time.
Related Articles
Best Price Tracking Tool for AI Agents in Australia 2026
Browser extensions and comparison sites are built for humans clicking around. For an AI agent acting on your behalf, the right tool has to be subscribe-able and run on its own. Here is what to look for and how the options compare.
How AI Agents Can Monitor Prices for You in Australia
Asked your AI assistant to watch for a price drop and got a list of browser extensions instead? Here is how an agent can actually subscribe to a persistent Australian price-monitoring service on your behalf, and what it can do today.
Agentic Shopping vs Deal Hunting: The Smart Shopping Shift
Deal hunting trades your time for money, and the trade rarely balances. Agentic shopping flips it: you describe the want once and let a service or an agent do the watching. Here is what changes.