Get your catchy domain at it.com

discuss Cost of creating your own landing page

NameSilo

topdom

Top Contributor
Impact
1,365
Here is an idea
(not an original one).

I create a "one page" site.

People visit a domain, that domain is redirected to that site,
that page has the same content whatever the domain, except,
that domain name appears on the landing page and in the title, or maybe also in the url.
(better if I can customize further.)
No problem with resolving, and safe against ddos etc.

People can send me a message via email (form-mail)
(better also if I can also reply using the site)
they would provide their contact info, and I would also get their IP.
A chat option can be a bonus (online or offline).

Any such service? There must be lots of free services somewhere.
I prefer services which don't invest in domains.
No FB or linked in, or twitter.
No service requiring id, phone call, installing app etc.

(Here by cost, I don't mean money (only).
I mean, time, protection, avoiding headache,
looks, usefulness,..)
 
Last edited:

Duk

Established Member
Impact
10
Simple thing but consider the trust factor, I wouldn't share my details and buy from such an unknown, untrusted site. You need to build trust, or else the customer will buy that domain from GoDaddy or from another registrar in the Afternic network.
But what if you are offering reputable payment services such as paypal?
 

topdom

Top Contributor
Impact
1,365
Yes, trust factor is important. Although most people are trustable enough, I can't make sure in the short run, and I "have" people who can build such a thing for me.. I would prefer to experiment with a free service and then maybe upgrade. Payment for domain purchase: I would use a marketplace. .. So this thing would be for communication purposes only...So, why I'm asking,..for example, maybe there is a standard way of doing this. I see many domains with contact form only.
 

Charybdis

Established Member
Impact
496
But what if you are offering reputable payment services such as paypal?
I'd buy the domain from my favorite domain name registrar, even if the domain is a little bit more expensive there, because
1) I don't trust a random one-page site
2) My details and payment methods are already saved at my registrar
3) I need a proper tax invoice

So if you can't build trust, then don't use your own landing pages, but use a trusted marketplace's nameservers instead, such as GoDaddy/Afternic landers, Dan, Sedo, Squadhelp, BrandBucket etc.

By the way Squadhelp charges only 7.5% commission in case of standard listings, with landing pages, multiple playment methods, 24/7 live chat, installment payments, optional logos etc.
 
Last edited:

twiki

Top Contributor
Impact
20,059
The problem is cost.

It appears you haven't done a proper lander yet. You're just doing guesswork. I been there and reached 30k names max. So I know the drill.

Once you do it and use it live, you will see that at a larger number of names, perhaps some with significant traffic, you will need a serious server. And that costs. Some of the sites will drive a TON of bot traffic. You will have to handle millions of hits per day, and sometimes a crazy instantaneous load.

You'd need at least an say $150...$200 per month, good multi-CPU dedicated machine with to host anything above 30-40 k names in total. And that's IF you optimize everything very well. And yeah, it'll be easy to be DDoS'ed - piece of cake, actually. Also a great network connection, bandwidth etc.

To get this stuff done, you need an array of servers, or better an (expensive) cloud platform like AWS. You'll need to convince domainers to hand you $50 or more per month just to break even with your plain lander.

It's very hard to get this done and not be in the hole. You need to offer much more.

Actually I've seen plenty of similar services for other landing pages, such as A/B testing ones. They are all dead today. Obviously.

Edit: I had to use an 32-core server, dual CPU to handle the load of my names properly. That after optimizing the webserver, script, caching and everything. I'm not using the lander anymore cause I see no point for all this.
 
Last edited:

VadimK

Top Contributor
Impact
2,490
@twiki I agree totally.
There are lots of site constructors where you can do free or very cheap landing pages (zyro.com is the one I've been using, but for different purpose), but yes, there are first - all the problems twiki described, then trust issues, also cost for a payment gateway, - and above all - you need some sort of CMS to manage all these landing pages, to be able to control or change the prices or keywords for SEO or texts in bulk... You can't go to each landing page changing things if you have 1000 names (I don't know how many you have, just a random number).
At the end of the day your own landing pages are worth it only if you are very seriously in a full time domain business, with thousands of good names already and even then I'd think twice. You'd have to go to a custom coder to create not only the pages, but to build the whole CMS to operate them, and that's many-many $$$...
 

Duk

Established Member
Impact
10
@twiki I agree totally.
There are lots of site constructors where you can do free or very cheap landing pages (zyro.com is the one I've been using, but for different purpose), but yes, there are first - all the problems twiki described, then trust issues, also cost for a payment gateway, - and above all - you need some sort of CMS to manage all these landing pages, to be able to control or change the prices or keywords for SEO or texts in bulk... You can't go to each landing page changing things if you have 1000 names (I don't know how many you have, just a random number).
At the end of the day your own landing pages are worth it only if you are very seriously in a full time domain business, with thousands of good names already and even then I'd think twice. You'd have to go to a custom coder to create not only the pages, but to build the whole CMS to operate them, and that's many-many $$$...
As I said, this is actually not difficult to program. Very simple way: in the server information is usually included the information about the requesting domain; so the same script, in the same server, in the same IP, called by different domains, can adapt to the different domains and automatically change the name of everything you want in the output webpage.

You will have some file with the list of websites and the prices, and it will automatically retrieve them depending on the information of the request headers. It is something extremely easy to perform in PHP.

Instead of using any unnecessary webpage creator, for a 1 page site it is better to program the thing itself directly, you save a lot of unnecessary complications, and you have the power precisely to adapt to whatever you want.

It is neither something absurdly expensive nor something that requires any extreme computing power.
 

Donnyd

Established Member
Impact
105
Here is an idea
(not an original one).

I create a "one page" site.

People visit a domain, that domain is redirected to that site,
that page has the same content whatever the domain, except,
that domain name appears on the landing page and in the title, or maybe also in the url.
(better if I can customize further.)
No problem with resolving, and safe against ddos etc.

People can send me a message via email (form-mail)
(better also if I can also reply using the site)
they would provide their contact info, and I would also get their IP.
A chat option can be a bonus (online or offline).

Any such service? There must be lots of free services somewhere.
I prefer services which don't invest in domains.
No FB or linked in, or twitter.
No service requiring id, phone call, installing app etc.

(Here by cost, I don't mean money (only).
I mean, time, protection, avoiding headache,
looks, usefulness,..)
You can always use efty?
 

Duk

Established Member
Impact
10
The problem is cost.

It appears you haven't done a proper lander yet. You're just doing guesswork. I been there and reached 30k names max. So I know the drill.

Once you do it and use it live, you will see that at a larger number of names, perhaps some with significant traffic, you will need a serious server. And that costs. Some of the sites will drive a TON of bot traffic. You will have to handle millions of hits per day, and sometimes a crazy instantaneous load.

You'd need at least an say $150...$200 per month, good multi-CPU dedicated machine with to host anything above 30-40 k names in total. And that's IF you optimize everything very well. And yeah, it'll be easy to be DDoS'ed - piece of cake, actually. Also a great network connection, bandwidth etc.

To get this stuff done, you need an array of servers, or better an (expensive) cloud platform like AWS. You'll need to convince domainers to hand you $50 or more per month just to break even with your plain lander.

It's very hard to get this done and not be in the hole. You need to offer much more.

Actually I've seen plenty of similar services for other landing pages, such as A/B testing ones. They are all dead today. Obviously.

Edit: I had to use an 32-core server, dual CPU to handle the load of my names properly. That after optimizing the webserver, script, caching and everything. I'm not using the lander anymore cause I see no point for all this.
This is incorrect too. Any website could get a DDoS attack, and yet you see billions of websites there without the superpowerful server networks you say you require. Anyway, in case you want to prevent that you can use a service such as Cloudflare, which works as a middle server delivering all of what your true server has, protecting you from those attacks you mention automatically.

Now, to the significant traffic. Supposing you actually have a domain with 1 million daily hits, I can guarantee you that any decent shared hosting, or some cheap VPS could handle that, because you are serving only a very simple webpage, which does not require any enormous computing power, and that will mostly take time in starting the connections, rather than in sending or in processing the information to be sent.

Two examples:
1.- A node.js script can handle with ease, without any particularly special computing power, a few hundreds of concurrent connections (at the very same time). To be conservative let's say it can handle 100 concurrent connections: 100 users per second.
2.- A php script of that simplicity can be executed in a few milliseconds. To be conservative, let's say it takes 10 ms (actually something very slow for such tiny thing), and that your machine can only handle 1 thread at the same time; that is 100 executions per second: 100 users per second.

Mathematics:
1 million users in a day means, 1 million in 86400 seconds, which means around 11.57 users per second. Your poor conservative Node.js or PHP scripts can handle that average very easy, and up to more than 8 times that average, in case of a "rush second", without any issue.

So no, you won't need superexpensive infrastructure to serve a 1 page website.


On the hand, I am wondering about how you programmed your website that you required so much power to serve it. It does not make sense if it was a simple script. On the other hand, if you, for example, used some CMS or some other application you probably were requiring the extra processing power just for unnecessary procedures going on with each visitor. The simplest solution would have been to make a simple script, and if you see your hosting is not enough, use a cloud computing service to serve it: even though you would have millions of users, if your code is simple enough the costs should be small because you will use very little processing power and you will use very little bandwidth too; yet the cloud service would be ready to expand in case of a sudden explosion of users.
 
Last edited:

poweredbyme

Top Contributor
Impact
904
Cheap unmanaged dedicated servers with 8 or 16 GB ram would handle tons of traffic for 4k websites with 1-2 static pages each. Technical skills would be the biggest bottleneck.. Managed servers with paid control panels would cost the max. In order to minimize the costs and to have full control over almost everything, one needs to have all the required technical knowledge which would likely take 3+ years to obtain.
 

twiki

Top Contributor
Impact
20,059
To those who made estimations, try and do it yourself first. Then, only then please comment.

The bottleneck is not traffic, or RAM etc. It's processing/ cores / threads. You can't do this on a regular VPS 8 or 16 cores VPS. It just hangs when a burst of traffic comes, even if you serve a single static page via an efficient server such as NGINX. And yes, an optimized server.

Edit: Contention is very high on such a system. Not for regular time, but when bursts of traffic occur, which is each half hour or so (varies). Sometimes almost the whole day.

You need a dedi box. Don't make assumptions, do it in real life - nothing beats that.

I thought the same at first.

But when bursts of traffic come, you'll see. Also, you're gonna get a DDoS like each few hours if you're "lucky" like I was.

@topdom - we encourage you to try if you want. Just be aware it might not be that simple as it looks.
 
Last edited:

twiki

Top Contributor
Impact
20,059
I'm currently using a dedicated box for redirecting my smaller portfolio of 7500 names. SSD, 16 cores. It works well. However I still get DDoS'ed each other day. Also on spikes of traffic, it gets quite high in CPU.

But if you host ngTLDs, you're going to get much more of that DDoS flavor. Much, much more for some reason. I only have 99.9% coms.

Also, I only do a redirect. Which means I only send the header to the browser and no content. If you send just a bit of content there, things are much more complicated. (in performance needed)

CMS - I use a highly optimized proprietary CMS, that has been optimized for a decade. We have a DAL, plugins for everything you need. Optimized execution thread, designed for performance out of the box. The problem, however, is not in the scripting, you can basically do the same in a few lines of code (which will not be different from our CMS's simple/short execution path). The problem is with webserver and contention from a large number of clients.

Edit: I'm pretty sure you're going to get a lot of Layer7 attacks as well which are not mitigated by most providers, by the way it hangs and the spikes I see.

You have to also trim logs to minimum, tune the webserver a lot, work a bit with network parameters, cache everything etc etc etc. All done here.

It works, but with a dedi, not with a VPS which is NOT really optimal for that. Which brings you to the real cost of this.
 
Last edited:

topdom

Top Contributor
Impact
1,365
Thanks for great responses. Most are too high tech for me (so requires too much time). People mentioned Efty and Dnhat. Anything to say about these, and comparison, would be great.
I would prefer a nondomain service. For example FB offers free pages, but I just dislike FB, and not trust it
although I don't expect arbitrage there.

(all marketplaces are trying to be evil, instead of being competitive (applying bad side of capitalism, instead of good side of it)).
 
Last edited:

Charybdis

Established Member
Impact
496
Thanks for great responses. Most are too high tech for me (so requires too much time). People mentioned Efty and Dnhat. Anything to say about these, and comparison, would be great.
I would prefer a nondomain service. For example FB offers free pages, but I just dislike FB, and not trust it
although I don't expect arbitrage there.

(all marketplaces are trying to be evil, instead of being competitive (applying bad side of capitalism, instead of good side of it)).

So what's your problem with the Squadhelp standard listings for 7.5% commission (and no monthly fee) which includes the payment processing fees? Read more here: https://helpdesk.squadhelp.com/en/articles/4466451-what-are-standard-and-premium-listings

If you implemented this on your own, just the payment processing fees would cost you 3-5%, plus your other expenses.
 

topdom

Top Contributor
Impact
1,365
I prefer to complete sale through marketplace of my choice.. And the first step is establishing a communication channel with the potential buyer (independent of any service), and knowing something about the buyer; so future discussion will be possible, manipulations by others can be reversed, I can know if the buyer is serious..

I prefer to sell at registrar level, sell at epik, if domain is at epik, sell at dynadot if domain is at dynadot ..
(mls/dsl equally good, of the system is trustable enough; for example cancelling is not easy)
why, not because of commission...Afternic/Sedo/Dan could be great in theory, because our domains could be visible at many registrars, but when we add price, if domain is not sold until expiration, and "sells" after expiration, we will be responsible for consequences..and marketplaces would enjoy preventing sales, because they control the lead, and wait until domain expires, and dropcatch etc. Marketplaces would try to use other marketplaces for arbitrage purposes, and in the process transfer part of the lead.. Sellers would be victim of all these dirty games, but if sellers know who the buyer is, all such things can be avoided with reverse tricks.

And while registrars/marketplaces are applying their tricks, they can easily screw things up, and contradict themselves, and buyers would lose trust in this industry (registrar. marketplace, owner...) and decide not to buy at any cost.

There are infinitely many possibilities, but not knowing something about the buyer or the lead, would be
a lose-lose scenario for the seller in all cases.

I can invent a fair model for all, but I don't think fairness is a priortiy for marketplaces at all.
 
Last edited:
Top down