|
The other day I received the following post from a whizzbangsblog reader outlining some problems that they've experienced with Sedo over the past 12 months. Unless requested I will always protect the anonymity of readers who take the time and trouble to write about domain issues that have impacted them.
As it is an issue that relates directly to Sedo I forwarded the reader's experience onto them for comment as I believe that it is only fair that both sides of the argument are addressed . I have also made my own comments following the reply from Sedo.
If you would like to have your level of access to whizzbangsblog.com increased so that you can "Submit News" items then please do not hesitate to contact me.
---------------------------------------
Is there Something Screwing at Sedo?
Like many domainers, I have most of my domains with Sedo. However this may be all about to change. My average earnings per click (EPC) on my portfolio have continued a steady decline over the last 12 months, and the slide continues.
 Many are aware of the so called "Sedo screw" whereby Sedo progressively cranks back EPC over the course of the month. Sedo officially deny this exists - well to most people anyway - and cite their dynamic pricing policy with traffic being revaluated on a daily basis according to values set by Google. However when you analyse the data you see a very predicable trend of a progressive decline in any given month, with a little upward tweak in the last couple of days. While Google may set the keyword values, the percentage that is passed on to Sedo users is very much in Sedo's control.
While this disturbs me, I am most disturbed about the decline in my portfolio over the last 12 months. In August 2006 Sedo announced a brand new monetisation system that was guaranteed to benefit all users. And whammo -average EPC per month rocketed to €0.36. Much celebration and gratitude. But then the next month EPC declined to €0.32, then to €0.29, then to €0.26 and so on until it now back to September 2005 levels of €0.16- that's well before the new system, before Sedo pro membership and before any bonus levels. This is all in a fairly static portfolio (not so static now as I progressively spirit domains away from Sedo). Sound familiar, a steady cranking back of earnings? Sedo will of course cry out with their mantra - "optimise, optimise" - which is all good and well, but even poor optimisation should not cause such a clearly defined linear decline over time.
So I am left wondering, is there something screwy at Sedo?
-------------------------------------
Comment from David Thorpe (Sedo)
In reply to the post, it is difficult for me to make any detailed comments as I have not yet had chance to view this individual's account - the addition of poorly paying traffic or the removal of particular domain name could naturally lead to a drop in general EPC.
 Since the introduction of dynamic pricing in September 2006 there were indeed teething problems where prices were incorrectly calculated. This however has been long since amended and payouts relate directly to Google's value.
The value of a keyword is therefore crucial - this obviously depends on the number of bids that a particular term has received. We therefore recommend customers to optimise, i.e. change the allotted keyword on a regular basis.
Any keyword tweaking will cause our ad provider's system to reevaluate the per-domain pricing. This takes 24 hours. For the first 24 hours, it applies what it believes to be a fair price for the clicks. It updates it within 24 hours with the actual click price. All parking companies and Google's own programme do this - some do it like us, 24 hours afterwards and keep the first "estimated" click prices in place - this gives the client more transparency and means that our statistics are reliable. Others update 24 hours later and re-evaluate the new click price and also change the reported click price. Also, at the month's end, the customer may have chosen a keyword for which a new ad campaign had just begun, or a new budget allocation had just been freed up - this often happens. Either way, we have always advised keyword tweaking every 10-14 days to take advantage of the dynamic pricing. Every change causes a re-evaluation which can help ensure you're getting the most from your domains.
Under such circumstances you would expect a change in earnings. It is certainly not Sedo's policy to gradually reduce earnings, nor is this practised
I understand that it's concerning to see changing earnings, but the very nature of online advertising is built on trends of peaks and troughs as ad budgets peak and expire. Changes in click prices due to these trends are perfectly normal. Obviously without seeing the individual account, I am unable to comment, but in general I would recommend focusing heavily on getting account-wide CTR and RPM up and more stable by shedding badly-performing domains, tweaking keywords every 14 days and manually optimizing the related links of your top earners. That can make a big difference!
-------------------------------------
Comment from Whizzbang
Since last August we have seen a steady decline in the EPC rates domains and I have written a number of articles and spoken at both TRAFFIC and Domainfest on this point. I believe that it is the single most important issue for domain name owners and it has largely been ignored to date. The challenge is how to separate out whether it's the parking companies (in this case Sedo) or Google and Yahoo trying to support their share prices.
For those domainer's that are customer's of Sedo (I am) the bonus value in your Sedo Pro account is largely meaningless due to the additional factor being applied to your account behind the scenes. Evidence for the existence of this additional factor or "screw" as it has been called by the reader has been confirmed by many ex-Sedo employees. Is this a malevolent plan by Sedo, I don't believe so. The problem that Sedo has is that domains are all a part of the revenue pool and fluctuations in the overall revenue for the pool impact the profitability of individual portfolios. For example, it's not profitable for Sedo to have a portfolio earning 110% of what Google pays them. This is where it's important for domainers to negotiate agreements with Sedo based upon a percentage of Google and not a bonus rate.
Another interesting point that David raised was the fact that Sedo doesn't really know what the value of a keyword is for at least 24 hours. This means that if you are optimising a large proportion of your portfolio then Sedo is guessing what the payout would be. What we've noticed when a keyword is changed that quite often the payout is immediately high and then linearly slides down to a steady state. My guess is that this steady state represents what Google is paying and that Sedo "masks" the decline through the linear reduction to this point. This is actually to the benefit of the domainer as in reality the revenue drop should look more like an impulse rather than a decline.
The challenge for domainers with thousands of domains is that changing the keywords on a daily or even 14 day basis is a logistical nightmare. Unless you've built large systems to cope with the volumes of data and monitor keywords across years then it's unlikely that you'll achieve much more than marginal benefits across a whole portfolio. We've spent a lot of time doing this at ParkLogic and thousands of hours honing our systems.
Other than seasonality statistically speaking there should be a low level of fluctuation in the EPC for a reasonable number of domains (over 500) where no domain's revenue is greater than 5% of the total daily revenue. There may be fluctuations in EPC for different domains on particular keywords but this will even out across a portfolio as the individual peaks and troughs flatten each other out. It takes a huge amount of effort to increase or for that matter decrease the EPC value for a whole portfolio. The fact that fluctuations in EPC do happen across portfolios suggests that an algorithm is being applied to the EPC rates by either Google or Sedo.
I enquired with Sedo if it was the case at their end (other than the linear decline) and they indicated that the problem was at Google. I believe that the number one issue that Sedo or any parking company should do is have a team of people constantly monitoring their upstream advertising partner. I wouldn't be surprised if in the case of Sedo, Google is getting things wrong and this is flowing downward to the domainers as peaks and troughs in daily EPC rates. The other thing that can something impact the EPC is when a large volume of "fraudulent" or low paying traffic flood the parking companies network (lower EPC) and are then expelled from the network (increasing EPC). This is a case where David's suggestions that changing keywords just won’t have any affect.
In terms of a gradual reduction in revenue for domains that have just been moved to an account I invited David at Sedo to play with two of my accounts so that they could see this issue first hand. They haven't had a chance to take up my invitation as yet but I'll let you know of the results when they do.
I would like to express my appreciation to the reader and also to Sedo for replying to the comment. It's always good to debate issues so that we can all benefit from the analysis. Please feel free to raise any experiences that you may have with the parking companies so that we can all benefit from your experience.
Wiki: Sedo, EPC, ParkLogic
Trackback(0)
|
Article Archive 
TrackBack URI for this entry
Subscribe to this comment's feed
