Should I fake it before building it?

Generally, yes, you should build a prototype first.

It's important to understand why. There are numerous advantages to doing things this way, but I'll just name a few:

1. It will make your eventual product better. Your hypotheses about what people do/don't like are just guesses until you actually start interacting with customers.
2. It will save you time. Assuming you're learning from the aforementioned interactions, you'll spend more time on useful features and less time building things nobody wants.
3. It will keep you motivated. Many founders burn out before launching something real. The simpler your product and landing page are, the less time you have to lose motivation. Additionally, having actual people signing up and showing interest will boost your motivation much more than going into a solitary cave of development.

Validating with a landing page is better than nothing but I don't think it's good. At this point it seems to me like people are cargo-culting it.

If there's a way you can talk to your potential users then you should read "The Mom Test" book (my summary: https://blog.kowalczyk.info/dailynotes/note/b4u674cvj43jdajdsukg-summary-of-) and interview the users following the ideas in the book (alternatively, you can watch author's youtube talks, it's the same material).

Advice in "The Mom Test" is the only one that resonated with me as far as validating ideas before doing the implementation work and absent other strong signals that the idea is viable.

Putting a web page on the internet, waiting for people to stumble upon that page and then counting the number of people who sign up to be notified or press the "buy" button? It doesn't sound to me like it provides you with a lot of information.

If people don't sign up, is it because they have no interest in the product or is it because you didn't explain the product well?

Or maybe you botched design (e.g. a page is unreadable due to low contrast) ?

If conversion rate is low, is it because visitors are hair-dressers and your product is for bartenders? Or they are bartenders but you misjudged what bartenders need?

What kind of visitors/sign ups do you consider to be a success? A failure?

My point is that signup/no signup gives you very little information.

"The Mom Test" talks about how to talk to people so that you can answer such questions and understand people's need better.

But to apply it, you need to have a conversation, not a landing page.

What about a third way of validating ideas as described in this article: https://medium.com/@dk/how-i-validate-all-my-business-ideas-4d063d64feed. The author describes how he gathers email addresses of potential users as a first step (narrowly defining whom he thinks his target audience is, of course), then tries to engage them to get feedback on a potential idea. To be sure it's a lot of work as each mail must be personalized to some extent to avoid looking spammy or canned. This seems to be a bold way of approaching people, but then again the benefits of this approach may outweigh the costs.

I got so tired of landing pages for non-existing products that I immediately leave as soon as I realize it.

You should definitely prototype your app and build a landing page before building the app itself. I agree with most of the answers here and I want to share this great blog post with you https://landerapp.com/blog/how-to-use-a-landing-page-to-validate-your-startup-idea/

I believe it really depends on the product or idea. Some ideas are easy to test with landing pages only. For example, PhpFog (heroku for php and eventually acquired) followed the landing page approach and it worked... tons of early interest which proved already he was solving a major pain point. Some ideas are hard, like new innovative ones, unless the pain point is crystal clear. Like getting me to sign up for the early release of a better email app is going to be hard... it better be clear what my pain is with gmail before I give a service my email. The last in between approach is offering a freebie with the email gating. Maybe the most common technique for digital goods like ebooks. Also very effective as the user is getting something immediate on return. No different than that popup vendor stand offering you free coffee, as long as you sign up to their email list. Actually, in real life it was a vendor making free balloons for kids, sucking us parents in, then cleverly asking the parent to try some free coffee, at which point you feel guilty if you don't sign up because of all the freebies. Didn't like that approach though.

I think having just a landing page will not really be enough. At a minimum your page should have an interactive prototype that gives your users something to play with. It will also show potential users you're closer to launch.

It just so happens I run a SaaS app that allows you to do this 😉

https://hunchapp.io

Upload your designs and create an interactive prototype in minutes. You could then embed this onto your landing page.

For getting some initial interested users I would recommend getting it featured on https://betalist.com/. We had over 250 users sign up from there when we were in closed beta.

I would say yes.

I recently put up a fake landing page (and Adwords to get traffic) along with doing interviews with potential customers of the product. The fake landing page doesn't provide much answers, since, like someone else pointed out - the signups are silent. However, the fact that about 5% of visitors do sign up even though just for early access tells me that there are users out there that are searching for a solution to the problem I'm trying to solve. And, are willing to pay for it (I have a pricing model on the page). Also, it means that I know that I will be able to market the product via Adwords once it launches.

But - the interviews I do tell me more about actual product needs.

Update: Forgot to mention that I've used Hotjar to track visitors and have made several alterations to improve the site content which is really helpful because it helps you understand how to communicate the product.

From an indie hacker perspective I would say don't make a fake landing page (I assume what your meaning). Most of time, indie hackers focus on niche markets, which means, the service you're building is going to serve a specific group of people, so, a landing page sounds to me like you're trying to add one more layer to a funnel (marketing funnel analysis) to filter out more potential customers.

Yes I think you should fake it before building it. I would even recommend not building the prototype before it's necessary.
Setup a landing page and start collecting emails. That's what I did (https://getbitboard.com) and it gave quite a boost in my confidence to see people signing up before I had anything. It also gives a sense of pressure to actually deliver the end result.
But most important, it will let you move on to a new thing if people aren't interested (without having spend a massive amount of time).

So I would definitely say, build a landing page which forces you to think about your value proposition. And don't worry about it being perfect on the first try (it probably won't be).
If you're up for it consider running some ads for it (I did twitter ads which worked fine).

At least that's what I would do.