Traffic: Are click fraud numbers being exaggerated by the traditional media?
As you certainly noticed, a lot of traditional media has recently been focusing on click fraud. Is it as big of an issue as it is made out to be, compared to traditional advertising? Unfortunately Eytan Elbaz of Google will not answer this question with statistics, but he lets us know that Google has the problem under control.
Here are some notes based on the Click Fraud Session at the Targeted Traffic Conference in Hollywood, Florida last week.
How does traditional media provide viewer/reader metrics?
Traditional media does not really provide verifiable metrics for their viewer numbers and circulation statistics. While TV statistics are becoming more precise with the help of digital cable and bi-directional networks, circulation metrics for papers are still very much basic estimates.
Newspaper circulation is measured with the help of two main numbers: Paid and free circulation. How reliable are those numbers?
Let’s take a look at an example provided by lawyer John Berryhill: Some years ago, hotels started giving us “free” newspapers in the morning, that most of us do not read. These papers are included in the newspaper’s circulation numbers, which are used in order to calculate the advertising rates.
Since those papers in the hotels were originally free, they had to be included in the “free” circulation numbers, which of course meant less value to the advertiser than a “paid” copy.
So the hotels and paper publishers decided to include the free paper in your room price, and allow you to get a credit if you return the paper to the hotel front desk. It now was a paid copy, which could be counted towards the “paid” circulation numbers – even though the value is more likely to be comparable to the value of a “free” copy.
The traditional print media is running out of advertisers. Thus it is of course in their best interest to discredit their biggest opponent, online advertising. Could this possibly be the main explanation for the frequent coverage of the “click fraud” issues in print media?
Some advantages of Internet advertising vs. traditional media advertising
- you can tell which context the ad was viewed in (you cannot tell if the reader read a certain page of a newspaper)
- “circulation” numbers are more reliable
- conversation rates from ads can be tracked
- smart pricing for advertising, i.e. depending on conversion rates
An issue that might affect publisher payouts was brought up by Rick [of OffShoring.com]: ‘Shaving’ by advertisers, meaning the manipulation of conversion rates. Since the online advertising systems rely on the advertiser to track the ‘action/conversion’ (by adding some tracking code to their “sale complete” page), they are in a position to manipulate the number of closed sales.
Another comment brought up by Joe [of CentralNIC] was that additional transparency, (i.e. publishing click fraud numbers), could be used to counteract the perception that clickfraud is a big issue.
[...] Are click fraud numbers being exaggerated by the traditional media? [...]
This post has also been featured over at CircleID: Are Click Fraud Numbers being Exaggerated?
[...] It’s easy to be scared by it since the fraud is so apparently simple to carry out. But, on closer examination, it’s probably not hurting most online advertisors significantly [...]
Some related articles
NY Times: Marketers demanding better count of the clicks
Yahoo! News: Daily Circulation falls at U.S. papers
Optimized Domains: Google gets serious about click fraud
[...] Marketing Pilgrim brings us some exclusive Google information, which confirms my theory that click fraud numbers are generally being misrepresented. [...]
Frank, I think Google’s click fraud rate is pretty low. However, for individual advertisers, click fraud can be spectacularly high. I usually see this when Google lumps in parked domains with their Search network. I think this is a poor practice on Google’s part. Parked domains are not search engines. Advertisers think they’re buying search engine advertising when they opt into this Search network (I’m not referring to the Content network), but in some cases, their ads are being displayed on parked domains.
For these advertisers, either click fraud is a huge problem or they perceive it as a huge problem. They see a spike in traffic from low quality sites that do not convert. Google tells them that this is not click fraud because parked domains can convert as well as search traffic. I don’t buy this argument. I think Google should create a separate ad distribution network, call it Domain network or something, and let advertisers choose whether or not their ads display on parked domains. If the traffic really does convert, advertisers will opt in.
What do you think? Incidentally, I believe that if Google took this step, the perception of click fraud would drop significantly.
Richard, thank you for your comment – I am actually also about to post an article about your DomainSponsor/information.com post here.
Do you always see a co-relation between click-fraud and parked domains? How do you determine what is click fraud? Is it a non-converting click?
I’ll comment on the inclusion of parked domains in “search traffic” in a separate post.
Traffic from parked domains being delivered by Google AdWords when the Content network is turned off is, by definition, click fraud. An AdWords account that’s configured to only distribute ads on Google and the Search network is an account that is choosing to only purchase search engine advertising. Any clicks that are not search engine advertising are fraudulent. Of course, Google doesn’t see it that way. What do you think?
The particular instance that drove me to start blogging about this mess was a client account where more than 50% of clicks for a single, exact match keyword were coming from parked domains, via searchportal.information.com. When I see odd patterns like this in a client’s account, I investigate. I also look at the visits themselves, both in terms of the click path after the ad click and to check for conversions.
I’ll be curious to read what your take is on these parked domains – and information.com in particular.