← Back to all posts Smart buying

How to use AI to find exactly the right product for what you need

Buying something new used to mean trawling through review sites, reading articles written for everyone in general and nobody in particular, and eventually making a guess. The reviews might be excellent, but they rarely account for your exact situation — your budget, your space, your preferences, the specific use you have in mind.

AI changes this completely. Instead of searching for "best Bluetooth speaker", you can describe exactly what you are trying to achieve and let AI ask you the questions that narrow it down to the right thing for you. The result is something much closer to getting advice from a knowledgeable friend than anything a review site can offer.

The key is starting with what you want to achieve, not what you want to buy.

Tell AI what you are trying to achieve

Most people type product searches the same way they would type them into Google: "best Bluetooth speaker under £50" or "rechargeable speaker with good battery life". These are fine starting points, but they produce generic answers.

The better approach is to describe the situation. Where will this be used? What matters most to you? What have you had before that worked or did not work? What is your budget? What problem are you actually trying to solve?

When you give AI this kind of context, the answer it comes back with is completely different — and considerably more useful.

Ask AI to question you

Here is a technique that makes a real difference, and one that most people would never think to try. Instead of giving AI all the information upfront, invite it to ask you questions first.

You can simply say: "I am looking to buy [type of product]. Before you make any recommendations, can you ask me a few questions to make sure you suggest the right thing for my situation?"

AI will then ask you things you might not have thought to mention — what you will mainly use it for, whether portability matters, whether you have strong preferences on brand or size, what you did not like about whatever you had before. Each question helps it build a clearer picture of what will actually suit you.

This is exactly how a good salesperson in a specialist shop would approach it. The difference is that AI is available any time, asks patiently without any pressure to buy, and has no interest in selling you the most expensive option on the shelf.

A real example — finding a replacement for a discontinued speaker

I use a small garden shed for hobbies and I like to have music on while I am in there. For years I had a little travel Bluetooth speaker — a compact, rechargeable unit that did not need to be plugged into the mains. It was a marketing product called the Ditti, given away at events rather than sold in shops. After years of heavy use, the battery finally gave out and I could not find a replacement anywhere. Marketing products like that often disappear completely once the promotion is over.

So I asked ChatGPT. Rather than just typing "small Bluetooth speaker", I described the situation properly:

"I am looking for a small rechargeable Bluetooth speaker for a garden shed where I do my hobbies. I had a speaker called the Ditti — a compact travel speaker that ran off its own battery without needing to be plugged in — but I cannot find it anywhere as it was a marketing product. I would like something similar: small, rechargeable with decent longevity, a reasonable sound for the size, and under £50. Ask me any further questions you need before making recommendations."

ChatGPT came back with a couple of clarifying questions — did I want it to be waterproof or splashproof? Would I also use it outdoors or just in the shed? How important was the length of battery life compared to sound quality at that price point?

Once I had answered those, it gave me a shortlist of three or four options, each with a clear explanation of why it suited my situation, what the trade-offs were and roughly what I should expect to pay. It also flagged which ones had a strong track record for battery longevity specifically — which had been my main concern after losing the Ditti to a worn-out battery.

I came away with a clear recommendation, a realistic price expectation and the confidence that I was not going to buy the wrong thing. That is more than any review site had given me in an hour of searching.

What a good AI recommendation looks like

When you describe your situation well, AI does not just give you a single product name. It typically gives you a shortlist with an explanation of why each one suits your needs, the key differences between them and any trade-offs worth knowing about. It will tell you if one option is better value, if another has a stronger reputation for durability, or if a particular model has a known weakness you should be aware of.

You can push further at any point. "Which of these would you recommend if longevity is my top priority?" or "Are there any that are widely available in the UK from well-known retailers?" or "Is there a version of this that comes in under £40?" Each follow-up question sharpens the answer.

The conversation builds towards the right choice rather than landing on a single answer from the start.

This works for almost anything you are buying

The approach works wherever you have specific requirements that a generic review would not account for.

Choosing a new vacuum cleaner? Tell AI about your home — the flooring types, whether you have pets, whether stairs are involved, whether you want cordless. It will ask anything it needs and come back with options matched to your house, not a hypothetical average household.

Replacing a kitchen appliance? Describe how often you use it, how many people you are cooking for, what you found frustrating about the old one. You will get recommendations that reflect how you actually cook, not how the manufacturer assumes you do.

Buying a gift for someone else? Describe the person — their interests, their age, what they already have, what they tend to use versus what sits in a drawer. AI is surprisingly good at this and will often suggest something you would not have found on your own.

The common thread is always the same: describe the situation, not just the product.

A few useful phrases to get you started

These work well as opening lines for almost any buying conversation with AI:

"I am looking to buy [product]. Before you make any recommendations, can you ask me a few questions about how I will use it?"

"I need something for [describe the situation]. My budget is around [amount]. What would you recommend and why?"

"I had [previous product] and it worked well apart from [problem]. What would give me the same benefits but fix that issue?"

"Here are my priorities in order: [list them]. What would you suggest and what would I be giving up compared to a more expensive option?"

None of these require any technical knowledge. They just require you to describe your situation honestly — which is something anyone can do.