strategy Part 7 - Portfolio Optimisation - Running a Traffic Test

[Whizzbang]

So you’ve just been to The Domain Conference in Fort Lauderdale and a monetisation company convinces you that you should run a test with them. Is there any real point and what is the best way to do this?

If you’ve ever moved your domains between parking companies by changing the nameservers then you will very quickly realise that there are so many variables to consider that the test becomes meaningless. For example, by routing the traffic at different periods of time, traffic volumes and domain market verticals all contribute to distorting the results.

In addition, if you move all of your domains across to the new company then from the previous article in this series we now know the best case scenario is they will win 35% of the time. Remember this number is for a properly optimised domain portfolio. If they win more than that then it’s only because your portfolio has not been looked after.
http://whizzbangsblog.com/index.php...portfolio-optimisation-running-a-traffic-test
Read more >

 

[NitroJunkie]

Hi Michael,

I've read all the articles you've posted recently. I liked the 6th one most. Very interesting stuff...

Especially, "65% of the time, in the best case, domains stay idle" part was shocking for me. I still don't get it... Since all the domain parking companies are using the "same" feed provider, how could this ratio vary from one parking company to another?

Let's say this is the case, then it all boils down to this: "The feed provider is treating every parking company in a different way. Therefore, there must be some parameters the feed provider is using in rating the parking companies." I guess this is the point where number and quality of the domains each parking company manage matters in the eyes of the feed provider, if I'm not wrong (pls correct me).

Anyway, I'm just trying to understand this biz in more detail.

Today, I applied to PL to run a test drive for a domain that I had an issue recently. Seems like you're pretty picky on the applications ... waiting for the approval.

Rgrds

 

[Whizzbang]

Hi Michael,

I've read all the articles you've posted recently. I liked the 6th one most. Very interesting stuff...

Especially, "65% of the time, in the best case, domains stay idle" part was shocking for me. I still don't get it... Since all the domain parking companies are using the "same" feed provider, how could this ratio vary from one parking company to another?

Let's say this is the case, then it all boils down to this: "The feed provider is treating every parking company in a different way. Therefore, there must be some parameters the feed provider is using in rating the parking companies." I guess this is the point where number and quality of the domains each parking company manage matters in the eyes of the feed provider, if I'm not wrong (pls correct me).

Anyway, I'm just trying to understand this biz in more detail.

Today, I applied to PL to run a test drive for a domain that I had an issue recently. Seems like you're pretty picky on the applications ... waiting for the approval.

Rgrds

First of all, let me thank you for your kind comments.....!

The reasons for the payment discrepencies are multitude, everything from:
1. Different landers
2. Different algorithms that allocate the funds earned.
3. Each parking provider has an overall performance tag and sub-tags. These tags influence the amounts you are paid.
4. Obviously, revenue share
The list goes on.......

Yes, we are quite picky with account applications at ParkLogic. The reason why is we get some many spam applications. If you think that your application has been falsely flagged as spam then reach out to me on my blog and I can sort it out for you.

 

[Hypersot]

Hello Michael,

I've been following your blog via feeds for quite some time now and I have to say I enjoy your writing (especially since you're a scifi fan) although I can't say I agree on many things regarding parking

having said that and since there are way too many things I'd like to talk about parking, I decided I should shrink everything to just one question:
Why choose you instead of Above.com?

one more thing,
have you changed the feed URL? Part 7 still hasn't appeared in the feed I mentioned. This is the url I currently use:
http://feeds.feedburner.com/Whizzbangsblogcom-DomainingNews?format=xml
I used to get all the posts of the blog except part 7 (that's actually the last feed I got)

thank you

 

[Whizzbang]

Hello Michael,

I've been following your blog via feeds for quite some time now and I have to say I enjoy your writing (especially since you're a scifi fan) although I can't say I agree on many things regarding parking

having said that and since there are way too many things I'd like to talk about parking, I decided I should shrink everything to just one question:
Why choose you instead of Above.com?

one more thing,
have you changed the feed URL? Part 7 still hasn't appeared in the feed I mentioned. This is the url I currently use:
http://feeds.feedburner.com/Whizzbangsblogcom-DomainingNews?format=xml
I used to get all the posts of the blog except part 7 (that's actually the last feed I got)

thank you

Really great question and I will highlight the differences between Above and ParkLogic not by disparaging the competition but by highlighting what we do.
1. At ParkLogic we manage all of your monetisation accounts and this means you get our revenue shares not your own. This will provide an immediate lift.
2. You do not have to do any of the setup of your monetisation accounts....we do it all.
3. We care a LOT about things such as which revenue number to use for normalised RPM calculations (there are about 7 different numbers). We ended up spending 6 months working on this problem to ensure we had the correct algorithms which routed the traffic to the right destinations.
4. We are not a self-serve business but we work with clients and each of our clients has direct access to myself for helping them make decisions.
5. We are very concerned about how to reduce our clients risk and are constantly exploring other revenue options that may even be completely outside the advertising and PPC channel.
6. We spent 6 months working out the best way to integrate zero-click solutions as they are essentially a spot price compared to an average price scenario. If you get this answer wrong then you end up selling your traffic for way too little.
7. We are completely transparent in how much we charge....it's published each month in your account.
8. At ParkLogic we typically work with larger sophisticated domain portfolio investors as they appreciate a lot of the analysis we conduct and also having a support team to help them out at a moments notice.
9. If the domain owner wants a completely outsourced model then we can provide this. This means we can manage all revenue, sales, first port of call for any legal issues and domain renewals against a set of agreed business rules. In some cases, we even provide board reports for our clients.
10. What keeps me awake at night is not incremental growth but revolutionary growth. In other words, what is going to massively impact the revenue line of customers....not just what will increase it by a few percent.

I hope that provides a few interesting differences.

In terms of my feedburner feed.....not sure what went wrong there! I'll have to investigate it.

 

[Hypersot]

Thank you for the quick response

I've been pondering of 'tasting' parklogic for more than a year now but it takes way too much work on my side since I have a solid setup part-above.com-automated / part setup by me, spread among 5 different parking services ... without mentioning all the URL re-directions from within Above.com to various marketplaces.

So, as you can probably imagine, I wanted some more info than what it's been shown on parklogic.com (which honestly, is not much)

again, thank you for your time, I'll carefully consider the points you made in your post

 

[Whizzbang]

Thank you for the quick response

I've been pondering of 'tasting' parklogic for more than a year now but it takes way too much work on my side since I have a solid setup part-above.com-automated / part setup by me, spread among 5 different parking services ... without mentioning all the URL re-directions from within Above.com to various marketplaces.

So, as you can probably imagine, I wanted some more info than what it's been shown on parklogic.com (which honestly, is not much)

again, thank you for your time, I'll carefully consider the points you made in your post

My pleasure.....feel free to reach out to me if you have any specific questions.

 

[thegeekguy]

Hi Michael, I think what your company is doing is great. I have some questions regarding Parklogic.

1. If we want to try out Parklogic how much minimum traffic do you recommend we send to see good results?
2. Regarding Zeroclick can we get around $5 RPM?

 

[Whizzbang]

Hi Michael, I think what your company is doing is great. I have some questions regarding Parklogic.

1. If we want to try out Parklogic how much minimum traffic do you recommend we send to see good results?
2. Regarding Zeroclick can we get around $5 RPM?

The minimum levels of traffic really depend upon what you believe is a significant statistical sample. What you don't want to do is run a test, get a result and find you still can't work out it is any good.
I'm not sure what you mean by the $5 RPM question....sorry.

 

[thegeekguy]

I usually get $4 to $5 RPM for my mix of blocked domains I send to Parkingcrew or Bodis

 

[Whizzbang]

I usually get $4 to $5 RPM for my mix of blocked domains I send to Parkingcrew or Bodis

For Google blocked domains the amount that is being paid really depends upon the end advertiser. This can vary greatly from market vertical to market vertical. We've seen literally hundreds of dollars RPM down to $1.

 

[decades80]

To be really honest, its the nRPM you are talking since years. Its just the statistical parameter on the other hand. Those 10 points could also stand for above.com. How do we know, you are not using above.com intermediate and sell it as a pure gold? There have been rumours you actually do that. The guy was making a test on his blog using your company and above.com and there was no significant statistical difference between you two.

 

[Whizzbang]

To be really honest, its the nRPM you are talking since years. Its just the statistical parameter on the other hand. Those 10 points could also stand for above.com. How do we know, you are not using above.com intermediate and sell it as a pure gold? There have been rumours you actually do that. The guy was making a test on his blog using your company and above.com and there was no significant statistical difference between you two.

Interesting questions.....hopefully I can answer them for you.
1. Yes, what's important is the normalised RPM.
2. We were around before Above.com (we actually managed the founders portfolio)
3. We have never sent traffic through to Above.com as it would be sub-optimal.
4. Would love to know who did the comparison on their blog. We are a high touch solution where we have a team that manages client's portfolios versus a self-managed solution on Above. We actually provide a very different service to them. If the blogger treated ParkLogic like Above then the result would definitely be similar.

 

[decades80]

We are not talking quantum physics here. Your algorithm could only make assumptions on the PAST data. No argue here. You cant predict the future about domain revenue. And thats exactly what above is doing, involving the history of domain name into the game. So no need for a fortune-teller here. Its just the simple fact, that some domains perform best when they are 100% of the time parked with the same parking provider and other category of names do better when they are moved around each couple of days. The moving could even be random or connected to the history. So no need to talk here about complicated algorithms. Nowdays the things are like that the actual revenue is the final figure of merit. Maybe the nRPM by other company is not statistically comparable to the other company. This discrepancy could cause greater nRPM and lower revenue at the end of the day.

ps: About the blog, that was already 3 years ago and the guy had XXX USD daily traffic and couldnt see the difference between the 2. Maybe it would help if you show testimonials with actual screen-shots how the revenue improved when using PL?

 

[Whizzbang]

We are not talking quantum physics here. Your algorithm could only make assumptions on the PAST data. No argue here. You cant predict the future about domain revenue. And thats exactly what above is doing, involving the history of domain name into the game. So no need for a fortune-teller here. Its just the simple fact, that some domains perform best when they are 100% of the time parked with the same parking provider and other category of names do better when they are moved around each couple of days. The moving could even be random or connected to the history. So no need to talk here about complicated algorithms. Nowdays the things are like that the actual revenue is the final figure of merit. Maybe the nRPM by other company is not statistically comparable to the other company. This discrepancy could cause greater nRPM and lower revenue at the end of the day.

ps: About the blog, that was already 3 years ago and the guy had XXX USD daily traffic and couldnt see the difference between the 2. Maybe it would help if you show testimonials with actual screen-shots how the revenue improved when using PL?

You raised a few interesting issues here which I hope to answer below:
1. It IS all about the algorithms and their sophistication. For example, there are around 7 different revenue numbers (eg. estimated, any account adjustments etc.) per domain that needs to be taken into consideration when routing traffic....get that calculation wrong and you route traffic incorrectly. That's the really simple side.....it does get more complex when you compare a spot price (ie. zero-click) versus an average price (eg. traditional parking).
2. The whole point of a nRPM being higher is that it removes the fluctuations in traffic levels (assuming statistical significance) and it DOES allow you to compare one monetisation source versus another. A large part of the algorithms is to get a statistically significant result for the minimum amount of traffic expended.
3. In doing a traffic test there are a lot of moving variables....some of them are external. For example, we had one client compare us just before there was a collapse in the Euro/US exchange rate.....this completely changed the results.
3. No idea who the client you are mentioning that blogged on us is.....loved to see the blog though!
4. Here is the results from a client who was using Above for their baseline revenue. The baseline data was from a large variety of parking companies. I am not prepared to indicate who the client is but we are in the process of developing a new website to go with the launch of ParkLogic Next that will have testimonials on it.

What the screen capture shows is if we saw at least 100% of the baseline traffic then we provided a 184% uplift in revenue. At least 80% of the traffic then we provided a 162% uplift until for all the traffic there was a 127% overall uplift in revenue versus the baseline. Essentially, if we see the similar levels of traffic then we knocked the results out of the ball park.....

Just for the record, there were some domains where the baseline outperformed our results. This could be for statistical variations across time etc. We would recommend to any client where the baseline outperformed ParkLogic to take those domains away from us and re-baseline them to ensure the results hold true. ParkLogic does NOT win for every domain ALL of the time (versus a baseline) but I can say that we DO send the traffic to the winning monetisation solutions at a particular point in time.

I hope that this answers your question....

 

[Hypersot]

Those graphs raise a lot of questions Michael... as in,
are you just saying that everyone that is not using ParkLogic, loses potentially nearly double their deserved revenue?

Even with the best optimisation formula, you can't just generate that much revenue difference compared to everything else out there...
either the above graphs exaggerate a bit or the rest of the parking companies are ..well, I don't know how to put it mildly, get a 'hefty' share of our income?

...or maybe I didn't quite get what those graphs are trying to say

 

[Whizzbang]

Those graphs raise a lot of questions Michael... as in,
are you just saying that everyone that is not using ParkLogic, loses potentially nearly double their deserved revenue?

Even with the best optimisation formula, you can't just generate that much revenue difference compared to everything else out there...
either the above graphs exaggerate a bit or the rest of the parking companies are ..well, I don't know how to put it mildly, get a 'hefty' share of our income?

...or maybe I didn't quite get what those graphs are trying to say

No....I'm not saying that. I am saying that this particular client was losing a lot of revenue. The graphs are a real part of the analysis that we conducted for a client.

Actually, you can generate a lot of additional revenue. We've worked with a number of companies that have their own Google feeds and improved the performance against it by as much as 32% overall. The parking companies are doing their best at what they do....I don't know any of them that are trying to "steal" money from domainers. I find that the parking companies on the whole are generally honourable to work with. In our case, if they tried to play with payouts then the traffic would just automatically route away from them anyway.

When we haven't performed as well what we have found is there is often a really good reason. For example, it could be that a few key domains aren't performing because of seasonality in the test versus the baseline data. The most important thing to understand is not that the numbers are up or down but WHY they are up or down.

For example, if we just look at the revenue we may be winning on a particular domain versus the baseline but when you look at the traffic we suddenly received twice as much. This is NOT a win....it's a lucky break and should really be discounted from the analysis.

Really understanding WHY a particular domain is winning or losing during a test is vital to extracting the full value from the traffic.

I would also like to add.....we've been doing this a LONG time and we understand this space intimately. To stay in business this long means we must be doing something right.

 

[thegeekguy]

Michael for a portfolio of 1000 domains how long does it take for your system to get the normalised RPM?

Im just rereading your post today I might need a 3rd go.

 

[Whizzbang]

Michael for a portfolio of 1000 domains how long does it take for your system to get the normalised RPM?

Im just rereading your post today I might need a 3rd go.

It really depends on the traffic levels per domain. The greater the traffic the less the time. I'm personally a risk averse person so depending upon the traffic/revenue levels I would recommend that you reach out to me and we can discuss the best strategy moving forward to minimize any risk.

 

[NitroJunkie]

First of all, let me thank you for your kind comments.....!

The reasons for the payment discrepencies are multitude, everything from:
1. Different landers
2. Different algorithms that allocate the funds earned.
3. Each parking provider has an overall performance tag and sub-tags. These tags influence the amounts you are paid.
4. Obviously, revenue share
The list goes on.......

Yes, we are quite picky with account applications at ParkLogic. The reason why is we get some many spam applications. If you think that your application has been falsely flagged as spam then reach out to me on my blog and I can sort it out for you.

Hi Michael thanks for the reply.

This answers my question, I guess: "Each parking provider has an overall performance tag and sub-tags. These tags influence the amounts you are paid."

Before going further down into the details of what influences the overall payouts, I want to clarify this "idle time" issue first. At this level, I'm leaving the discussion on the performance of the landers and the other "parking-provider level" parameters to another day.

In this regard, now I have a new question popped up after reading your another article titled: "What's Going On With PPC? - Part 1". In this article you mention about CAF: "Custom Ad Frame (CAF). Essentially this means that all parking pages are now largely served by Google and not the parking companies themselves.". So, if at the end of the day Google is serving the parking pages, why do Google need to rate the parking providers?

And again in connection with the same article, you finish that article by saying "I plan on unpacking this further in the next articles in this series as I explore what is actually going with domain traffic.". Have you wrote something about this or is it still on the backburner? This CTR / EPC debate is also very interesting.

Those articles are providing a more in-depth understanding of the parking business. Thanks for your time. I hope you're not finding my questions annoying. My curiosity comes from the opacity of this business. Thanks

 

[Whizzbang]

Great questions Nitro!
The reason why Google needs to rate the parking providers is because it keeps them focused on high quality traffic rather than high volume poor quality traffic.
I actually can't remember the exact article you are quoting but I know that I've written a lot on PPC CTR/EPC over the years. You can search these topics on my blog. It would seem that the current series needs an article on CTR/EPC.
No.....none of your questions are annoying LOL!

 

[decades80]

Those graphs are silly. They are made to one example and customer only. We need more than that, as you stated previously, the performance shouldnt be based on a single example :D And those graphs come on... They are looking like search traffic at Flippa. You could do better.
This is NOT how professionals present case study. You should know that. Those graphs from one account are on the 12 year old level. How we should trust the company who is based on nRPM high tech parameter and presents graphs like that?

 

[Whizzbang]

Hmmm....don't know how to respond to insults and I have no idea what you're talking about. I don't mind someone questioning but please keep it civil and without the sarcasm.

Someone asked for an example of performance, I pulled that out. The above graphs don't have anything to do with nRPM so I don't know what you are looking at....sorry about that. If what you were missing were the revenue and plRPM (ie. nRPM) trends then here they are below. I hope this helps.

If what you are saying is that you need more examples then I'm happy to oblige. In fact, due to this request our new website (to be launched soon) will have both a number of formal case studies and testimonials from clients. BTW - I'd love you to point me to a case study from someone else that you feel has been valuable to you.

I should mention that if you sign-up with ParkLogic and we have no idea who you are or anything about you then don't expect to get these results. Sadly, due to the nature of our industry there is a lot of fraud and we have to be really cautious with people that "just sign-up". What we really like is when we have a chance to speak or meet with the domain investor to understand their objectives and get to know them. This gives us the confidence to open more premium advertising networks that don't take too kindly to fraudulent activities.

 

[decades80]

Your tactic only works with average Joe, you know that, dont you? I am off this thread, but would advise everybody to deal with CAUTION.

ps: The company only has about 140.000 domains which is the size of very small domain fund portfolio.

 

[decades80]

I didnt want to be rude. Its just the fact that you didnt present any of the solid proof, how overall performance of your 140.000 domains improved on your platform. So without proof its 50% up and 50% down. Like roulette: black or red. And + if we substract from the result of 50% up your revenue share, it leaves us what? 25 - 30% up?