Data normalisation is something that is often overlooked when trying to compare one parking companies stats versus another but it's absolutely critical if a domainer is going to make an informed decision.

What is data normalisation? It's what you need to do in order to compare apples with apples and oranges with oranges. For example, let's imagine that you want to compare data from two parking companies; parking company A and parking company B then the first thing that you need to do is understand the defintions that they are applying to the data.

A unique at one company may be counted in an entirely different manner compared to a unique at another company. This can drastically affect the RPM (Revenue Per Thousand Uniques) for this reason:

RPM = revenue / (total number of uniques) * 1000

If we have the same revenue (eg. $10) Parking Company A may say there was 1 unique while parking company B says there were 10 uniques. The RPMs for the two companies would then be:

Parking Company A   RPM = 10 / 1 * 1,000 = 10,000
Parking Company B   RPM = 10 / 10 * 1,000 = 1,000

So do parking companies have different defintions of uniques? Absolutely they do. For example, the last definition of a unique at Sedo was a unique IP address every 18 hours while Domain Sponsor was a unique IP address every 24 hours. This defintion difference would greatly influence the RPM for domains.

What's so important about the RPM? The RPM is often used for optimisation of domains when you're monitoring the impact that a keyword change has made. For example, if I have a keyword of "music" for a week on a domain generating an RPM of 10 and I then decided to change the keyword to "online music" which gave an RPM of 5 then "music" by iteself would be the better keyword.

Now let's imagine that I'm comparing a change in keyword for a domain across multiple parking companies. Measuring the RPM is not necessarily a good idea until we first of all normalise the uniques. The reason why you may want to do this is because you may be trying to work out whether to put all of your music traffic at Domain Sponsor rather than Sedo or vice versa.

The problem with just looking at the revenue generated by each keyword is that there could have been a spike in revenue that is distorting the figures and causing you to make an inaccurate decision. By dividing the revenue by the traffic tends to smooth out revenue spikes and assuming that the behaviour of the traffic is consistent there should be a norm that the RPM trends towards.

I hope that this was interesting reading. Feel free ask any questions that you would like about this topic as it can sometimes be a little confusing.

TrackBack URI for this entry

Comments (0)

Subscribe to this comment's feed

Name

Email

Website

Title

Comment

smaller | bigger

Subscribe via email (registered users only)

I have read and agree to the Terms of Usage.

Add Comment Preview