Seven ways to use ChatGPT as a shopping check before Black Friday

Shopping help before the rush

As Tom's Guide reported, writer Elton Jones is treating ChatGPT as a shopping assistant for this year's Black Friday. His reasoning is that sales now start weeks before the actual day, and the big retailers, including Best Buy, Amazon, Walmart and Target, are already posting their biggest discounts.

The article is a list of seven prompts, which are the instructions you type into a chatbot. Each has blanks in brackets for your own details, such as the product, the budget or the price. The pitch is that a well-built prompt makes the chatbot do some research and ask you some hard questions before you spend anything.

None of this is a new technology. It is a practical way of using one that many readers already have open in a browser tab.

Four prompts that test the deal itself

The first prompt tackles a wishlist. You give ChatGPT your list, your total budget and what you already own. It asks you a set number of questions about what you need, then sorts everything into buy, consider and skip, with reasons.

The second checks whether a sale is real. You give it the exact model, the advertised price and the supposed original price. It then looks for documented price history, says whether the cut is unusual or routine, cites dated sources and admits what it could not verify. That last request matters, because it pushes the chatbot to show where its answer is weak.

The third looks for the cheapest option that still does the job. You state the product, what you will use it for and your maximum spend. ChatGPT picks out the must-have specifications and compares three options, including a cheaper or previous-generation model. It also says what you would give up by choosing the cheaper one. Jones says he is trying to stay under $200 this year, so this is the one he is aimed at.

The fourth adds up total cost. Printers, coffee makers, consoles and phones often need subscriptions, accessories, replacement parts, delivery or installation. The prompt asks for first-year and three-year costs, separates necessities from optional extras and labels estimates as estimates.

Three prompts about fine print and willpower

The fifth prompt goes after reviews. Jones says he likes reading Reddit and Amazon reviews, but the prompt asks ChatGPT to gather independent reviews and owner feedback for an exact model. It should pull out recurring complaints, long-term reliability worries and the strongest positives, and it should separate measured test results from anecdotes. It also asks who should avoid the product.

The sixth is about the listing. A price that looks too good can hide a third-party seller, a refurbished item or a harsh return policy, even on a major retailer's site. You paste in the listing and its policies, and the chatbot lists who sells and ships it, its condition, warranty, return deadline, return shipping costs and restocking fees. It also flags anything unclear to confirm before paying.

The seventh is the most unusual. You give ChatGPT your whole cart, your budget and your priorities, and ask it to argue against every purchase. It should flag duplicate functions, compatibility problems and unnecessary extras, then recommend a final cart that stays in budget. The prompt even allows it to recommend buying nothing.

That last idea is the most valuable one in the piece. Most shopping tools push you toward buying, and this one is built to resist.

Where the chatbot can let you down

Jones closes by telling readers to treat the chatbot's output as evidence to check, not a verdict. His advice is to do some research of your own to confirm its claims are accurate before you commit.

That caution deserves emphasis. Chatbots can state prices, stock levels and discounts that are out of date or simply wrong, and they can point to deal links you cannot confirm. The safest approach is to use ChatGPT to build a shortlist and a list of questions, then confirm the real price, return terms and availability on the retailer's own page.

The price-history and review prompts depend on the chatbot finding and citing real sources. Treat each citation as a lead, and open the page it points to before relying on it.

It also helps to match the prompt to the purchase. For a laptop or a piece of software, where specifications matter as much as price, the cheaper-alternative and total-cost prompts will do the most. For a small impulse buy, the final cart audit may be all you need.