As well as the regular suite of charts I went I’d trawl through the archives and pull some data on a monthly basis from July 2019 until today. This really showed a number of the major trends and the fact that domain monetization investors have benefited from an uplift in normalized RPM over the last few months.
I hope you find the information that I’m sharing each week of benefit for your business. Don’t forget to like, comment or subscribe so you don’t miss out on future updates.
Please click on the below link to view the first three parts of A History of RPM:
Part 1 - A History of RPM
Part 2 - A History of RPM
Part 3 - A History of RPM
Some people have mistakenly thought that if another player is involved in the chain (eg. someone like ParkLogic) then the margin that they take will consume any benefits. This couldn’t be further from the truth as companies such as my own play more than just a revenue increase, we also manage downside risk.
To understand this hidden benefit, we need to return to the RPM formula from Part 1 in this series. At ParkLogic we effectively have an RPM for every domain, at every monetisation company at every point in time. In fact, many years ago I coined the term normalised RPM (nRPM) as the unit of measurement for every monetisation source.
This means we actually know who is paying the most for any domain at any point in time. We then route the traffic to those destinations so that ParkLogic customers achieve a greater yield for their traffic.
As a by-product of this process we also manage downside risk. For example, if a domain is being paid an nRPM of $10 at one provider and $9 at another we would rightly route the traffic to the $10 provider. Suddenly the $10 provider loses a key advertiser and is now paying $2. If you were leaving the traffic at a single provider, then you would get the $2 but with ParkLogic the traffic will automatically flow to the $9 provider. This dramatically smooths out these types of market disruptions and reduces the risk for the domain investor.
Now here’s the problem with all of this history and discussions about RPM. When you deal with direct advertising networks, they have another type of RPM altogether….and it’s a difficult one to crack and it highlights the difference between a spot price and an average price.
This is the third in the series on running a traffic test. The first two articles in the series can be viewed at:
Part 1 – Traffic Test – Baseline Data
Part 2 – Traffic Test – Focusing on Performance
As we discussed in Part 2, what’s vital in running a traffic test is understanding the performance, both positive and negative. The data is the data but interpreting what it’s telling you is far more important if you are wanting to develop long-term value from your domain traffic.
So how is the real live test portfolio progressing so far? The following two charts show the overall trend lines are all heading in the right direction. Despite the time of year and the typical overall decline in advertising spend over the northern hemisphere summer, revenue is continuing to trend upwards.
The second chart clearly shows the reason for this is due to the normalised RPM (or plRPM as we call it at ParkLogic) is also trending upwards as the optimisation algorithms continue their work. For those of you that are unaware, RPM stands for Revenue Per Thousand - more of per thousand what shortly. It’s at this point that I’m going to take a little detour to explain a few things about plRPM and how it differs from RPM.
I don’t know why it’s taken me a little while to update some of my analysis but there has been a big reporting change by Google which got me a little excited. In the past Google only reported an overall aggregate Traffic Acquisition Cost (TAC) but now this has all changed.
Since Q1 2015 Google has broken out the numbers for the network members from the distribution partners. You may ask, what is the difference and where does the domain industry sit?
Distribution partners are people like Apple, who charge Google for real-estate on iPhones etc. Google has hidden the amount that companies like Apple take by publishing the amount of revenue as a percentage of the overall Google website revenues (an odd metric actually). This means it’s impossible to reverse engineer the financial relationship…..which actually makes commercial sense.
On the other hand, for the first time ever Google has released the TAC to network members and this is the segment that the domain industry sits within. According to the reports, Google is paying out between 68-70% of the advertising revenue to this channel.
This begs a really simple question, if Google is paying out this amount then why have many domain investors experienced drops in their revenue?
To answer this question, we need to appreciate that the domain channel is one small part of the overall Google Network. This means that although on average Google is paying out between 68-70% of the advertising revenue the domain channel may be getting more or less of this figure.
So it’s time to do a dive into the data. The first stop was to pull out the average Google normalised RPM (nRPM) figures for a VERY large domain portfolio for each Google reporting quarter. I then graphed the quarterly percentage change over time (see the blue line on the graph below) and I then compared this to the payout figures that Google released for the network (the yellow line).
What we can clearly see is that the two lines nearly mirror each other. When Google pays out more, then the normalised RPM increases and when it pays out less then it immediately drops. But the huge swings in nRPM don’t really make sense when you compare it to the small swings in the Google average payout…..unless there is something else going on.
The only reason this would occur is if there is an artificial gain control for the overall domain industry that is being triggered by Google. In fact, one of these exists and it’s called smart pricing. In other words, a small change in the overall Google payout results in a massive change in what the domain industry receives. I think this has less to do about quality and more to do about Google making its numbers.
So here’s the really fun part. Since we now have a percentage change in the nRPM and also the % paid out by Google for a quarter we can multiply these two figures together and get a pretty good read on what the industry is actually paid out as a percentage of the gross advertising revenue!
Now here's the catch. We also have to assume that the Domain Industry was initially paid in Q1 2015 about what the average for the whole Google Network. There could be some scaler at work here but I tend to doubt it.
The below graph shows that in the recent 2016 quarter the domain industry received 91.4% of what Google received from advertisers. In fact, we’ve been on an upward trend since Q2 2015.
I know that many of you will cite specific examples where you’ve seen advertisers paying $50 for a click and yet you only get pennies. These specific perturbations in the data will continue to occur because what I’m looking at is averages and trends.
So right now we should be living in the golden age of Google payouts so get ready for a sharp trend down.....unless there is some corporate reason why Google would keep the payout rates high.
Now here’s where it gets REALLY interesting. Armed with this data the domain industry knows a big chunk of the Google quarterly report BEFORE anyone else does. In fact, with a little more effort I wouldn’t be surprised if we could pull just about the whole report apart!
This really got me thinking about whether we as an industry actually sit on a gold mine of data that would allow us to predict the movements in the Google share price……now that would be interesting!
I should say that if you plan on using this data for investment purposes you should first of all get some professional advice.
