Creo este es un articulo muy bueno sobre las estadisticas reales en parking y los datos que te dan las compañias de parking. Son tres partes, agrego las 2 primeras.

What is Traffic?
Michael Gilmour de ParkLogic.com

Part 1

I've found that many people believe that they know what domain traffic is. They'll tell me, "It's simple, how many unique users going to a website." I hate to spoil the illusion but that is only a small part of the answer. A simple test can be conducted by parking a domain with a consistent level of traffic at two different parking companies. What you will find is that two different levels of traffic are reported. You may say, "This doesn't make sense? Traffic is traffic therefore the parking company reporting the lower level of traffic is ripping me off". That may be the case but is actually quite unlikely. Let's get a little deeper into what traffic actually is and how we can exploit traffic.

In order to understand what traffic is we need to understand what a parking companies definition of traffic is, afterall, they are the ones that are doing the measuring for us. The first part of any parking definition is that one of the jobs of a parking company is to eliminate as much of the fraud from the traffic as practical. One of the major ways of accomplishing this is to filter out traffic coming from the same IP address. This helps in ridding the world of fraudulent domainers who have people click on their advertisements thus increasing the revenue and traffic for a domain. I should state right now that this is FRAUD and all good domainers should report any such activity that they discover to their parking company.

The next part of the definition is to work out what time frame an IP address is "refreshed". For example, if you have two unique users coming from a single IP address then only one is counted if the IP address is viewed by the parking company within a particular time window.

Now let's look at a real world example. If we were to direct traffic from a domain name to Sedo one week and Domain Sponsor the next we would find that Domain Sponsor would report approximately 75% of the traffic of Sedo. The reason for this is that Sedo counts traffic as a unique IP address going to a parked page every 18 hours while Domain Sponsor counts a unique IP address going to a domain every 24 hours.

This is a HUGE issue. For a start this means that every Internet user that is routed through NAT (Network Address Translation) risks in not being counted as a user as someone else being routed through the gateway IP has already been counted. For example, Australia's largest company (BHP Biliton) has over 100,000 employees but only uses 5 gateway IP addresses. This means that over an 18-24 hour period (depending upon whether domain sponsor or sedo is used for parking) a maximum of 5 users will be counted going to a parked page. This isn't so good because you may have a domain that is industry specific (like mining) and generates a lot of traffic from single gateway IPs - you'll never actually know the true traffic levels for your domain.

Part 2

In part one we just went through the problems that domainers have with counting unique users and how it depends upon the definitions that each parking program has for what is a unique visitor. This lack of standard between parking programs has a number of different impacts upon optimisation of a domain.

1. If you are determining whether a domain has traffic (eg. bulk registration method) by viewing parking statistics then you won't get the correct answer and may be throwing away perfectly good domains.
2. Optimisation often relies on optimising a domain name to the best RPM. Let's think about this in terms of the example of Sedo and Domain Sponsor from part 1. Let's imagine that both companies have identical revenue then since Domain Sponsor has a lower traffic (ie. counts users as IP addresses over 24 hours rather than 18 hours for Sedo) then it will have an artificially higher RPM value compared to Sedo. This means that Domain Sponsor could claim that it produces a higher RPM than Sedo but since the traffic is commensurately lower the claim is irrelevant.
3.Because we are unsure of the TRUE traffic values for a domain we may decide not to invest further effort into the domain so that it can be monetized in other way.

So what do we do to get around the problem of traffic not being counted, or worse, we aren't getting paid for traffic? One way could be to route all traffic through your own webservers, record the IP addresses and route any traffic from the same IP address within a 18 hour window to a different parking company from the first time that the IP address appeared. This sounds simple but ends up becoming quite complex when you begin to aggregate up all the data from all the parking companies (let me assure you that just getting the data is pretty difficult!) and then normalise the data for the differing time zones etc. But, it works! One of the things you could also do is build a report that displays the actual traffic data that you sent to the parking companies for the domain.

We've been doing this for a long time now for our own domains and have built a pretty large and complex system for tracking all this data. Here's the shameless plug.....about six weeks ago we made a decision to provide access for domainers to a system that does this called Park Logic. It beats the heck out of building it all again! :-)

Here's another problem..... What happens when a single IP address clicks on a link and then clicks back to click on another advertiser's link on a park page. This will be discussed in part 3.

Leer parte 3: http://whizzbang.typepad.com/whizzbangs_blog/2007/04/index.htmlDominiosIDN.com ||

Muy interesante DOminero, gracias, creo que nos confirma muchos de los rumores ...
Salu2

Este dominiero escribe muy bonito pero sus dominios son terribles.Domains for sale: BtcArg.com FondoBTC.com TravelDupes.com NubePrivada.com

Buena información dominero, se pueden sacar buenas conclusiones de todo ello.