A/B testing with/without afternic

[davidc1]

Hello all,

As I see hugedomains delisted their domains from afternic, i want to make a try...

Just starting a split test on my portfolio, trying not listing domains on afternic.
Half of my portfolio is as usual dan landers + domains listed on afternic, whereas the other half is dan landers but domains not listed on afternic.
These days my sales are 2/3 afternic - 1/3 landers , 5 years ago it was the contrary...

Let's see results, i think i'll have enough data in less than a week...

 

[phaethon]

Let's see results, i think i'll have enough data in less than a week...

Are you serious? So soon? Given the randomness of domain sales, you would need years and still those 'results' will have little importance. Is your portfolio as large as HD? If you have millions of domains, you could try it, but with a few hundreds or even thousands, I doubt the results will have any importance at all..

 

[AEProgram]

Split testing does not make sense if the products are different products. Each domain is its own unique product.

If someone has their names listed on Afternic but they sell it through their own lander (even a dan or namepros lander) that sale most likely did not come from the person searching godaddy. It came from a typin, someone wanted to check if the name is available.

GoDaddys power is the fact that most people know their name and go there when looking to buy a domain. Nobody is about to take that from them and they know that.

A company looking to upgrade their current name from a .io to .com is going to the domain directly. A person that already made up their mind for a name or has a few names they like will go to those domains directly or check whois.

 

[davidc1]

Are you serious? So soon? Given the randomness of domain sales, you would need years and still those 'results' will have little importance. Is your portfolio as large as HD? If you have millions of domains, you could try it, but with a few thousands, I doubt the results will have any importance at all..

The key number is not the number of domains but the number of sales I dont have millions of domains but enough domains to have results in a short time...

 

[johnn]

Selling domain names is not like selling pizza.
If you have crap names then you can wait for 100 years and nothing will be sold.
Good names are always sold.
Statistics are not working in this business

 

[davidc1]

Split testing does not make sense if the products are different products. Each domain is its own unique product.

If someone has their names listed on Afternic but they sell it through their own lander (even a dan or namepros lander) that sale most likely did not come from the person searching godaddy. It came from a typin, someone wanted to check if the name is available.

GoDaddys power is the fact that most people know their name and go there when looking to buy a domain. Nobody is about to take that from them and they know that.

A company looking to upgrade their current name from a .io to .com is going to the domain directly. A person that already made up their mind for a name or has a few names they like will go to those domains directly or check whois.

Every domains are different, but i believe in stats and numbers... it's the key IMO

GoDaddys power is the fact that most people know their name and go there when looking to buy a domain. Nobody is about to take that from them and they know that.
Possible, but all portfolio are different, explain me how HD could not lose millions by delisting their domains ?

 

[davidc1]

Selling domain names is not like selling pizza.
If you have crap names then you can wait for 100 years and nothing will be sold.
Good names are always sold.
Statistics are not working in this business

I sell them like pizza, several time a day !

 

[AEProgram]

Every domains are different, but i believe in stats and numbers... it's the key IMO

GoDaddys power is the fact that most people know their name and go there when looking to buy a domain. Nobody is about to take that from them and they know that.
Possible, but all portfolio are different, explain me how HD could not lose millions by delisting their domains ?

HD has millions of domains names. They know their numbers better than anyone. They know what their fees are with AN, they know how many sales came to their landers. Maybe they figured out that if it says abc.com is taken, it triggers people to go check by who by visiting the site and people like to buy things they were just told is taken? I doubt this is why they removed it, but they have data that does not compare to what we have.

They also did not remove all names. VFlights .com is a HD name and you see it come up on godaddy when you search for it. So there must me a lot more to this.

 

[davidc1]

HD has millions of domains names. They know their numbers better than anyone. They know what their fees are with AN, they know how many sales came to their landers. Maybe they figured out that if it says abc.com is taken, it triggers people to go check by who by visiting the site and people like to buy things they were just told is taken? I doubt this is why they removed it, but they have data that does not compare to what we have.

They also did not remove all names. VFlights .com is a HD name and you see it come up on godaddy when you search for it. So there must me a lot more to this.

Which data would they have that we dont have ?
I know what's my AN fees are. I know which % of my domains are sold through landers.

 

[davidc1]

Just checked some domains of HD, some of them are still listed.
Some are listed using "make offer", some others BIN. Maybe they let some domains as they already received offers over afternic.

 

[davidc1]

First 2 sales through landers, but it doesnt count a lot, US is sleeping (sales from Madagascar and UK)

 

[Wisher]

So are you going to also add the Dan only lander ones to SEDO MLS (Like HD does).
Otherwise of course you will be losing out on other Reg path sales. Not a like for like if you don't do this.

 

[davidc1]

So are you going to also add the Dan only lander ones to SEDO MLS (Like HD does).
Otherwise of course you will be losing out on other Reg path sales. Not a like for like if you don't do this.

Not for this test. To me it's very strange that HD do that, except if they have good conditions with SEDO MLS and that is not my case.
SEDO vs Afternic is around 20/80 in term of power as 80% of domains sold through afternic goes to godaddy (which is not a sedo partner unfortunately )

 

[Wisher]

Not for this test. To me it's very strange that HD do that, except if they have good conditions with SEDO MLS and that is not my case.
SEDO vs Afternic is around 20/80 in term of power as 80% of domains sold through afternic goes to godaddy (which is not a sedo partner unfortunately )

Whats the point of the test if it does not represent what someone would do in the real world?
You would not cut off Reg path sales completely by only using landers as you would obviously lose out on a lot of sales.

 

[davidc1]

Whats the point of the test if it does not represent what someone would do in the real world?
You would not cut off Reg path sales completely by only using landers as you would obviously lose out on a lot of sales.

The test is to see how much I lose if i dont use afternic only.
Do you think in a real world that someone seeing "domain taken" on godaddy, will go to namecheap to search the domain ?

I know exactly what i lose not using sedo as I decided to stop listing on sedomls few months ago as they are unable to manage VAT.

Anyway, I conclude that i dont lose money delisting domains on afternic, i'll probably relist domains on sedomls as there is no penalty with them if i dont use their dns.

 

[Wisher]

The test is to see how much I lose if i dont use afternic only.
Do you think in a real world that someone seeing "domain taken" on godaddy, will go to namecheap to search the domain ?

I know exactly what i lose not using sedo as I decided to stop listing on sedomls few months ago as they are unable to manage VAT.

Anyway, I conclude that i dont lose money delisting domains on afternic, i'll probably relist domains on sedomls as there is no penalty with them if i dont use their dns.

No maybe not. But in the real world if they are using NameCheap or others to start their search they will not see any of your domains and so you will lose out on sales. If you had SEDO MLS like HD does you would'nt be losing out on those sales.

So your test is not a true reflection on the amount of sales you lose from not using Afternic as some of those Afternic sales would have been from reg path (not from GoDaddy).

And it is not comparable to what Huge Domains does.

 

[davidc1]

No maybe not. But in the real world if they are using NameCheap or others to start their search they will not see any of your domains and so you will lose out on sales. If you had SEDO MLS like HD does you would'nt be losing out on those sales.

So your test is not a true reflection on the amount of sales you lose from not using Afternic as some of those Afternic sales would have been from reg path (not from GoDaddy).

And it is not comparable to what Huge Domains does.

Yes you are true.
Previously I used sedo, but i prefered to stop as it's impossible to manage VAT with them, i need to collect it but they dont, i know it's not a big loss when domains are afternic listed.

 

[Dane]

I sell them like pizza, several time a day !

What price-point seems to be most attractive to buyers? Meaning, have you noticed a particular price that causes buyers to purchase the "pizza" () more-so than with other prices?

 

[davidc1]

What price-point seems to be most attractive to buyers? Meaning, have you noticed a particular price that causes buyers to purchase the "pizza" () more-so than with other prices?

As near to everything in this world... If you want to sell more, sell low as possible.

 

[Recons.Com]

Hello all,

As I see hugedomains delisted their domains from afternic, i want to make a try...

Just starting a split test on my portfolio, trying not listing domains on afternic.
Half of my portfolio is as usual dan landers + domains listed on afternic, whereas the other half is dan landers but domains not listed on afternic.
These days my sales are 2/3 afternic - 1/3 landers , 5 years ago it was the contrary...

Let's see results, i think i'll have enough data in less than a week...

I think, for the experiment, you should do 1/2 names listed with AN, 1/2 not listed, but all of them with Sedo (MLS if eligible).

That way, your name will still get exposure via distribution network which often duplicates Afternic's with the exception of Godaddy itself.

if your results show that the afternic unlisted bucket makes more (probably less in pre-commission and more post commission calculation) after the commission, then it would be something that all of us should use.

E.g. 50k Afternic sales with 25% average commission (boost, non-boost, afternic lander, other lander mix) will net you 37.5k

While 45k sales with own landers + Sedo network with average of 15% (mix of 3% cc charge with no commission to 20% Sedo partner sales to 15% Sedo own sale to 10% with Sedo lander) will give you 38.2K.

 

[davidc1]

I think, for the experiment, you should do 1/2 names listed with AN, 1/2 not listed, but all of them with Sedo (MLS if eligible).

That way, your name will still get exposure via distribution network which often duplicates Afternic's with the exception of Godaddy itself.

if your results show that the afternic unlisted bucket makes more (probably less in pre-commission and more post commission calculation) after the commission, then it would be something that all of us should use.

E.g. 50k Afternic sales with 25% average commission (boost, non-boost, afternic lander, other lander mix) will net you 37.5k

While 45k sales with own landers + Sedo network with average of 15% (mix of 3% cc charge with no commission to 20% Sedo partner sales to 15% Sedo own sale to 10% with Sedo lander) will give you 38.2K.

I stop this test now, i have enough results to see that, low prices domains needs to be afternic listed, otherwise i lose sales (30%+). As some said, could be different for high priced domains, but that kind of test would need much more time to be made... and could cost me a lot if the ratio of missed sales is the same what i see today.

 

[Recons.Com]

I stop this test now, i have enough results to see that, low prices domains needs to be afternic listed, otherwise i lose sales (30%+). As some said, could be different for high priced domains, but that kind of test would need much more time to be made... and could cost me a lot if the ratio of missed sales is the same what i see today.

Wow. I envy your velocity. You had enough results in one day to conclude your test? ))

 

[davidc1]

Wow. I envy your velocity. You had enough results in one day to conclude your test? ))

Yes, around 15 sales (most low prices +2 lto)

 

[Dane]

Yes, around 15 sales (most low prices +2 lto)

How low is "low"?

 

[davidc1]

How low is "low"?

from 100, so maybe the test should be done with higher priced domains only, 2K min , but even at this price not sure not to loose 10-20% of sales....