All of it can be done by hand, for nothing. Reading forums, booking calls, writing a promise, building a page, running a small campaign — none of it requires a subscription.
So the question is not which tool is best. It is which part of the job you want to stop doing yourself, because what these tools sell is time.
There is a second thing worth saying before any list: no tool makes the decision. Everything here either gathers evidence or puts something in front of people. The judgement at the end is yours, and a tool that offers to score your idea out of ten is selling you the one thing it cannot honestly supply.
The tool you have already tried
Most people reading this have already described their idea to ChatGPT or Claude and asked what it thinks. That is a reasonable instinct and it is worth being precise about what you get back.
They are genuinely good at the work around the edges. Turning a messy idea into a clear sentence. Drafting questions for a customer call. Summarising forty pages of notes you paste in. Arguing the other side when you ask it to. For all of that, use them.
They are weak at the judgement itself, for two specific reasons.
They give you the average answer. They are trained on what has already been written, so what comes back is roughly what most people already think. That is exactly the wrong place to look for an opportunity — the whole point of the previous article in this sequence is that the gap is where the market's shared assumption is wrong, and the average opinion is that shared assumption, written down.
They agree with you. Describe your idea and ask for an opinion and you will usually get encouragement. Not because it is true, but because agreeing is the behaviour these systems have been trained into. You have to work quite hard to get an honest negative out of one, and if you have to work hard to get it, you cannot trust the positive.
The honest summary: an LLM is a good thinking partner and a bad witness. It does not know what your market said last week unless you go and get it, and going and getting it is the actual work.
The four jobs, and what speeds each one up
| The job | Doing it by hand | What speeds it up |
|---|---|---|
| Find where they gather | Search, read, take notes | PainOnSocial, Draper |
| Hear it in their own words | Read threads, book calls | PainOnSocial, Draper |
| Describe it and put it up | Write it, build one page | Lovable, Carrd, Framer, Draper |
| Get strangers in front of it | Set up and run ads | Meta, X and Reddit ad managers, Draper |
| Read it and decide | Look at two numbers | Nothing. This one is yours |
1. Find out whether the group is real, and where they are
Before anything else, you need to know that the people you are aiming at exist as a group, gather somewhere you can name, and describe the problem without being prompted. This costs nothing but time, and it is the cheapest job on the list to do badly.
By hand it is search and patience. Find the subreddit, the Facebook group, the Discord, the trade forum. Read six months back. Write down the phrases people use for the problem, because those phrases are what you will put in your ad and on your page.
PainOnSocial does the Reddit half faster, surfacing recurring problems out of discussions rather than making you read everything yourself. It is Reddit-only, which is fine if your market is there and useless if it is not. Draper does Reddit, social media, and the extended internet, to surface exactly what your future customers are saying.
2. Hear the problem in their own words
Reading gets you the language. Talking to people gets you the reasoning behind it, which is the thing you cannot get any other way.
By hand this is a calendar link and fifteen-minute calls. You do not need a research platform to run it, and adding one mostly adds admin.
The hard part is not the conversation. It is finding people who genuinely have the problem and are not simply being nice to you. Friends, family, anyone invested in your success and anyone you have already pitched will soften what they say. What you want is strangers who are currently paying to solve this some other way — in money, or in an afternoon a week. Communities are where you find them, which is why this job follows the last one.
Ask about what they did, not what they would do. "Tell me about the last time this happened" is worth ten answers to "would you use something like this."
3. Describe it, and put it somewhere it can be judged
Now the idea needs to exist outside your head, as one page with one promise and one thing to do.
By hand means writing the promise yourself and building a single page. Lovable will generate and deploy one from a description. Carrd and Framer will get a page live in an afternoon. All of them are more than sufficient — the page is not the hard part and it should not take you more than a day.
The catch is what comes next. A page is easy to build and difficult to get anyone onto, and this is where most validation attempts quietly stop: the page goes up, a handful of friends look at it, and nothing conclusive ever happens. Decide how you are getting traffic before you spend a day on the page, not after.
4. Get strangers in front of it
By hand means the ad managers themselves — Meta for most consumer products, X and Reddit as real alternatives depending on what you sell and what you can spend. A few hundred dollars over a week is enough for a directional answer.
This is the job where inexperience costs actual money. The platforms are built to spend your budget efficiently for their definition of efficient, and the default settings will quietly reshape your test — optimising for conversions you will never get enough of, or widening your audience past the people you meant to reach. It is entirely learnable, and the mechanics are worth reading before you turn anything on, but going in blind is the most reliable way to spend three hundred dollars and learn nothing.
Where Draper fits
Draper does these jobs with you, in one place: finding where your customers gather and what they are complaining about, working out how to describe what you are making, building the page on-brand, and running the traffic to it.
It is the same work described above. The difference is that you are not assembling four tools and the gaps between them, and you are not doing the reading yourself. What it does not do — what nothing on this page does — is tell you whether to go ahead. That part stays with you, which is as it should be.
Before you pay for anything
When a tool is the wrong purchase
When it offers you a score. Anything that rates your idea out of a hundred has converted a judgement into a number, and the number has no basis outside the tool. It feels like progress and it is the opposite: you now have a figure you cannot interrogate standing where your own reasoning should be.
When choosing tools has become the work. Comparing subscriptions is comfortable, and it produces no evidence at all. If you have spent longer picking than you would have spent doing the first job by hand, the tool was never the bottleneck.
Related questions
If you are still working out whether anyone wants this, the sequence those tools serve is the whole demand test, end to end. The first job in it is checking that your target market is real, and if you have found competitors already serving these people, that is good news and it needs an angle.
Draper runs the research, the page and the traffic with you, and leaves the decision where it belongs. Try Draper free →
