AdMoolah News and Views

AOL recently released a document on the Internet containing query data for about 20 million searches from 658,000 different users.

This has been seen as a huge mistake on AOL part and raises some serious privacy concerns. The AOL user IDs in the data had been replaced by random numbers, but there was still quite a bit of private data in the search queries such as Social Security Numbers and credit card numbers. See Elliot Back’s post about the privacy issues.

The reason this is of interest to publishers is that it is a very extensive source of information about keywords and search behavior. Mining the data can reveal which keyword phrases are popular and what alternate phrases people use when one phrase does not show the desired results. This data is much more detailed than anything obtained from sources such as Wordtracker or Overature keyword listings.

I expect to see people’s analysis of this data popping up soon.

See this Digital Point thread for more discussion.

Update: Here are some websites which you can use to access the data:
http://www.aolsearchdatabase.com/ – Users can query by user id, keyword, date or clicked website
http://fuaol.com/ – A list of the top 5000 keywords
http://simplifiedsec.com/KeywordDigger.html – Search by keyword

Update 2: http://dontdelete.com/ This looks like a good, quick interface to the data.

The equation for calculating how much revenue is generated with a CPC program is:

pageviews * CTR (click-through-rate) * RPC (revenue per click)

So far, this blog has focused on CPC and RPC. Lately I have become more interested in the pageviews part of this equation – basically increasing traffic. This is were the real growth opportunity is. CTR and RPC will always be limited, but traffic is almost limitless. Even giants like Yahoo and Google worry about increasing page views.

So, from here on in you’ll start seeing posts about things like marketing, SEO (search engine optimization), link building and buying advertising. I hope you enjoy exploring these topics with me.

I will still be posting about the advertising related topics I always have, so if you like that part of the blog, please do stick around.

One things that has bothered me about YPN is that they are very slow to report the previous day’s earnings in performance reports. I often want to check my balance for the day before I turn in, which is often just after midnight Pacific Time, and it never has the figures for yesterday’s revenue at that point. Today I realized that there is an easy way to calculate the previous day’s earnings.

First, note the “pending balance”, which is equal to the previous month’s earnings plus the earnings accumulated so far this month. Then run a performance report with a date range that include the previous month up to the current day. Subtract the “Report Totals” figure from the “pending balance” and you end up with yesterday’s revenue.

For example, today is August 3rd just after midnight Pacific Time. Let’s say my pending balance shows $500.00 (not the real figure). I run a report with a date range of July 1st to August 2nd. The report does not show August 2nd’s revenue and the total shows $450.00. I now subtract $450.00 from $500.00 and I know that I have earned $50.00 for August 2nd.

Now I can go to bed satisfied that I know how much I earned from YPN the day before.

As reported by ZDNet, Google’s CEO Eric Schmidt had this to day about the click fraud problem:

According to Schmidt, Google’s auction-based pay-per-click advertising model is inherently self-correcting. Schmidt’s scenario for what would happen if Google did not police click fraud and it was “rampant”:

Eventually, the price that the advertiser is willing to pay for the conversion will decline, because the advertiser will realize that these are bad clicks, in other words, the value of the ad declines, so over some amount of time, the system is in-fact, self-correcting. In fact, there is a perfect economic solution which is to let it happen.

This type of attitude is bad news for publishers. If this were too happen the advertisers wouldn’t mind because they are paying less per click for fewer real clicks, so their ROI stays the same. Google wouldn’t mind because they are still getting paid for every click whether fraudulent or not. The click fraudsters would love it because they would be making lots of easy money. The only losers ion the equation would be the legitimate publishers. They would be earning less per click.

Note that this was a “what if” scenario, not something that Schmidt acknowledged was going to actually happen. It’s still scary that he thinks this is a “perfect economic solution” when the fraudsters are getting rich and the real publishers are losing out. Doesn’t sound perfect to me!

There has been a lot of talk recently about how click fraud is killing Cost-Per-Click (CPC) . Reports of bot nets generating fraudulent clicks have been appearing for a few months in places like ClickZ and JenSense . In early June, Mark Cuban expressed his opinion in a piece called Why I think ClickFraud is far greater than imagined.

More recently Outsell, Inc. published a report that click fraud is costing advertisers $1.3 Billion and that fraudulent click account for 14.6% of all clicks. This led to many mainstream news source such as The Financial Times and BusinessWeek to report these problems with click fraud and question the long term health of the CPC market.

Something else that has been seen as a sign that CPC is in trouble is that Google has recently started to experiment with CPA ads. Because of this some people concluded that Google itself worried about CPC in the future.

Are CPC programs such as AdSense and YPN really dying? Is click fraud going to kill them? Many people are predicting that CPC market will collapse the way the CPM market did several years ago. But, I think there are some important difference between the collapse of the CPM market and the current CPC market.

1. The CPM market reached its heights along with the Internet bubble. Web sites were willing to pay extraordinary amount just to get “eyeballs”. The more eyeballs they had the bigger there IPO would be. The CPC market is not in such a bubble.
2. One of the major problems with CPM advertising was that it was difficult to measure the effectiveness of it. With CPC advertising and good analytics, it is much easier to measure the exact ROI (Return on Investment) advertisers are getting for their money. If CPC ads were not working for the advertisers they would stop using them. Click fraud does happen and will continue to, but as long as the overall ROI of a CPC advertising campaign is positive, advertisers will continue to use them.
3. Google’s experiments with CPA does not automatically mean they are giving up on CPC. Google is constantly experimenting with new product and new ways of doing things. They are willing to try anything, see if it works, and when it does, they will introduce it as a full fledged product. Google generally does not introduce new products in a reaction to something, they are continually trying out new ideas.
4. There will always be advertisers that will want to use CPC instead of CPA. For example, car manufactures want people to visit their web sites and learn about their cars. They are not expecting to actually sell a car when someone clicks through to their web site. It is difficult to tie a visitor to a major car manufacturer’s web site to a specific action to charge via a CPA program.

All that being said, I do thing CPC is coming under some pressure. For certain industries, CPA ads can provide a better model for advertisers. Smart advertisers will try CPM, CPC and CPA campaigns and see which has the best ROI. Sometimes CPA will beat out CPC programs. Click fraud is another danger. If Google and Yahoo cannot keep the click fraud under control advertisers will need to drop their bid per click to maintain the same ROI. CPA ads can be harder to game than CPC ads, so they may have an advantage in this area.

I do believe that CPC ads are in there heyday right now, and prices may fall. Smart publishers will experiment with all type of advertising, CPM, CPC, CPA, Fixed price, or any other arrangement they can think of. Each niche and each website will benefit from different combinations of these types of programs, the publishers who make the most money will be the ones who maximize their revenue, regardless of the type of program used.

Google announced today that they are making things tougher for Made-For-AdSense (MFA) sites who use AdWords for advertising.

Google had this to day:

As you may recall, we began incorporating advertiser landing page quality into the Quality Score back in December 2005. Following that change, advertisers who are not providing useful landing pages to our users will have lower Quality Scores that in turn result in higher minimum bid requirements for their keywords. We realize that some minimum bids may be too high to be cost-effective — indeed, these high minimum bids are our way of motivating advertisers to either improve their landing pages or to simply stop using AdWords for those pages, while still giving some control over which keywords to advertise on. Although it is counter-intuitive to some who hear it, we’d rather show one less ad than to show an ad which leads to a poor user experience — since long-term user trust in AdWords is of overarching importance.

From time-to-time, we improve our algorithms for evaluating landing page quality (often based on feedback from our end-users), and next week we’re launching another such improvement. Thus, over the coming days a small number of advertisers who are providing a low quality user experience on their landing pages will see increases in their minimum bids.

I see this as a direct attack against MFA sites. MFA sites are sites which display AdSense ads and provide no useful content. Usually the content is simply copied from search results. These MFA sites target high paying keywords while using very low bids on the AdWords system to get traffic. Since there is little useful content on the pages, users end up often clicking on the ads. Many AdSense publisher despise these sites because they drive down the cost of an average click and do not provide users with any useful information.

The Google AdWords Landing Page and Site Quality Guidelines clearly state that MFA type sites are not considered quality sites:

• In general, build pages that provide substantial and useful information to the end-user. If your ad does link to a page consisting of mostly ads or general search results (such as a directory or catalog page), provide additional information beyond what the user may have seen in your ad or on the page prior to clicking on your ad.
• You should have unique content (should not be similar or nearly identical in appearance to another site).

If this works out the way Google intends, it will be good for everyone in the AdSense ecosystem: publishers, advertisers and users. The only ones who will be hurt will be the MFA producers.

According to Seeking Alpha, Google’s AdSense has started to test CPA (Cost-Per-Action) ads, in addition to the CPC (Cost-Per-Click) and CPM (Cost-Per-Thousand-Impressions) they already offer.

Publishers are able to choose which ads they want to show on their site, and the ads will not compete with regular AdSense for Content ads.

Some interesting quote from the letter Google sent to select publishers:

How can I promote the CPA ad unit?

Since this is a test and these CPA ads are not regular ad units, we are giving you more flexibility in saying things like “I recommend this product” or “Try JetBlue today” next to the CPA ad unit. However, you should still not incite someone to click on the ad, so saying “Click Here” is not ok.

Google has never allowed this type of promotion with referral ads, so it will be interesting to see if this is allowed when the CPA ads come out of beta.

Will I be able to see reports within my account?

When the test begins, you will receive weekly email reports of conversions you have accrued and your total revenue within the CPA test.

Google already has reporting features for their own referral products, so I’m surprised that they won’t have online reporting. This means people testing the ads will have to wait at least a full week between changes to see what works.

This is a major addition to the AdSense offering and puts them in direct competition with networks such as Commission Junction. This could be a huge new revenue stream for Google and its publishers.

Found via John Battelle’s Searchblog

Yahoo has introduced some new features to publishers.

First, they replaced the old “Content Setup” tab with a “Publisher Services” tab. It is subdivided into three sub-tabs – “Drive Traffic”, “Enhance Your Site” and “Build Your Site”. Each sub-tab has forms and links for various Yahoo! offerings that might be of interest to publishers. The “Drive Traffic” sub-tab has forms for site submission, RSS feed submission and site suggestion. It also has links to Yahoo! services such as podcast and video submission pages, Yahoo! Site Explorer and Search Submit Express. The “Enhance Your Site” sub-tab aggregates many Yahoo service that you can add to your site such as search boxes, Flickr badges, and Yahoo videos. The last sub-tab “Build Your Site” includes links to Yahoo! domain registration and hosting services, and Yahoo! Small Business.

I can see why Yahoo! added the Publisher Services tab. It promotes their own services by making it easier for publisher to find out about them and use them. It will be a convenience for publisher who regularly use these features.

The next change is a new help center. Yahoo! has started creating how-to guides for setting up and managing accounts. Several FAQs are included as well.

My favorite change is a new column on the reports – Revenue Per Click. This shows total revenue divided by number of clicks. I’m not sure why Yahoo! or AdSense have not included this column on their reports before. I’m sure its something that most publishers calculate on a regular basis.

The final change is bulk URL Reporting upload. Instead of adding reporting URLs one at a time, you can now cut and past up to 100 URLs into the form at once. The maximum number of active reporting URLs allowed is still 100.

It’s good to see Yahoo! adding features and helping to raise the bar. Now I hope AdSense follows suite and adds a Revenue-Per-Click column into their reports!

Last week Google added a new Site Diagnostics page to the AdSense console. It has since disappeared. Darren over at ProBlogger wrote to the AdSense team to find out what was going on, and Google wrote back that it was just a temporary test.

This seems a little strange to me. My guess is that this was put onto publishers pages by mistake, and Google quickly realized that it wasn’t supposed to be there and removed it.

eBay is going to launch their own contextual ad network, which means they will be competing with AdSense, YPN, Chitika, and eventually MSN’s adCenter. Their “AdContext” product will offer publishers a chance to put ads to eBay auctions on their sites. Much like AdSense, an eBay robot will scan pages for context and then display relevant ads. Publishers will get 35 percent to 60 percent of sales, depending on volumes. One major difference between AdContext and AdSense is that AdContext will pay the publisher only when an actual sale takes place, not on a per-impression or per-click basis.

Because this is a contextual program, AdContext ads will not be able to be placed on the same page as AdSense or YPN ads. They can however, be placed on the same site.

I have a feeling that this will be a good program for certain types of websites that deal with consumer goods. Because of this, the program might be a bigger threat to Chitika then AdSense.