Google helped provide one form of CPM contraction when they changed what part of an AdSense ad block is clickable. Markus Friend said that knocked his CTR down by about 60%.

Markus Friend also highlighted that Google controls about 40% of the (heavily consolidated) ad market:

90% of Advertising revenues are made by the top 50 sites and the top 10 sites take 70% of that, with google taking 40% of all Online US advertising.

This CPM compression is going to cause many late movers to crank out content with more ads on it, further lowering their CPM and direct readership until they are financially insolvent. But then again, the web could use another bust cycle to clean up the meaning of the word "content". Shoddy intrusive ad networks like NetAudioAds should not be featured as the next big thing in the WSJ.

In a recent interview, Nick Carr said he expects the ad consolidation trend to continue

It’s the aggregators that are the big winners, at least in economic terms, not the legions of individual contributors.

Over the past 20 years, we’ve seen that the automation provided by computer systems has tended to concentrate wealth in the hands of a small slice of the population. I expect that trend will only accelerate in the years ahead. If you’re one of the digital elite, you’ve got it made. If not, the prospects are less bright.

What is the solution for publishers? How do prevent yourself from being absorbed by the commons? Develop meaningful relationships, be remarkable, and sell direct. Hugh on owning an idea:

Social Markers are a prime form of social shorthand, that people use to STAKE OUT the ecosystem they’re occupying. So why do I find this such a useful term for marketers? Because obviously, if your product is a Social Marker in your industry ecosystem [the way the iPhone is in the mobile world, or Starbucks is in the coffee world, or Amazon is the book world, or Google is in the search world, or Whole Foods is in the supermarket world, or Virgin is in the airline world, or English Cut in the bespoke world etc etc] you will have an AMAZING competitive advantage to call your own.

And if the product your company makes is not a Social Marker, I guess the first question would be, "Why the hell not?" Quit your job and start over.

One of the easiest ways to claim an idea is to turn its launch into an event, and differentiate it from everything else you are doing. Buy the matching domain name if you can.

Bonus cool link: Bill Slawski mentioned a Yahoo! patent about moving away from the random surfer model to a user sensitive PageRank. Now if they could only apply some good ideas in the SERPs. And no, this does not count.

Comments

Today I came across an AdWords ad for an automated ebook business model website. Their screenshot highlighting their Paypal account was:

• hosted on another site
• named powersell_paypal2.jpg
• did not show payments but showed withdrawals

Fake Business Statistics

A lot of (mis)marketing techniques are more covert though, through the use of

• false stats due to small sample size
• sponsored research
• syndicating bogus stats from third party research

Fake / misleading research is remarkable, so it is more likely to be cited, and recycled by people hyping similar business interests. A while ago MarketingExperiments did social media marketing research which was not really research, and yet those false stats promising social media goodness just appeared again:

Claiming that the above ROI is 1,427% higher is at best dishonest. You can teach the value of something without syndicating lies as truth.

Social Media Traffic Does Not Buy

Want to know the truth about most social media traffic? Its garbage. Some of my AdSense ad campaigns use an affiliate account to track ROI. Until I filtered them out for poor performance, MySpace and Digg were providing about 90% of my overall affiliate ad volume with 0 conversions, whereas some of my better affiliates make a sale a day or a sale every few days on far less traffic.

I know that was an isolated example and it would be unfair to judge the entire market on that, but consider this…those ads had a horrifically low clickthrough rate and still only cost a dime a click. If I was getting a lot of volume on a network that size while bidding next to nothing then that ad inventory is not worth that much. Simple as that.

Some Top Publishers Are Afraid Social Media Marketing

Some leading publishers are even worried about deflating their CPM by getting to much lousy social media traffic.

Comparing Social Media to Direct Navigation

Why do you think domainers make so much money without even needing to develop websites? For a person to end up on a parked page they have a lot of implied intent in their location. The same is true for a search result. If you just searched for something you have implied intent. Google is worth 200 billion and I am not.

For all the hype Facebook ad system has got, there is limited value in their user data:

Google actually knows all of that, and at least 10X more data about users than Facebook, but hasn’t seen the need to really mine the data yet, since search intent has proven to be worth about 100X more than that kind of data so far.

If social marketing gets you clean links it is great. If people recommend your product to their customers that is great. I get mentioned on Seth Godin’s blog and sales double. I make the front page of a social news site and nothing happens. Most of the social media hype is hollow and without value.

Put Social Media to the Test Today

Still don’t believe me that most social marketing traffic is worthless? Ask yourself why StumbleUpon only charges 5 cents a visitor for any category – including big money categories like daytrading, gambling, and financial planning.

If the successful secret marketing strategies that send 4 cent traffic are buzzworthy then they could at least have the decency to tell me I can get the same thing for a nickel with no effort.

 Comments

It’s a good post that needs to be read by every executive at major media companies today, and in many ways a lot of it seems like common sense to those who have been watching the transformation occur and understand what’s going on.

Specifically though I wanted to address Scott’s final principle:

Advertising must create value

Google turned search advertising into the most profitable media business on the web by following the basic principle that advertising must create value for consumers. Search advertising is so powerful because the ads are relevant and USEFUL.

The most successful new advertising models will be those that create huge value for consumers, not those that manipulate users or violate their privacy (i.e. be like Google, not Facebook)

As we browse the web today, display advertising is generally not as relevant and as useful as search advertising. Search has a big advantage in that it knows the specific direct intent of the user at that moment in time provided by a keyword. While a lot of the display ads we get are entirely untargeted and seem useless.

To be fair, there are a lot of display ads that are targeted now based on behavior, geography, and other factors that end up creating value for the user.

There is a lot of movement with technologies like Yahoo’s Smart Ads, better behavioral targeting, and other technologies that are going to continually move display advertising to be more useful and relevant to the user. As I posted recently, this should lead to higher click through rates, higher conversion rates, and lead to higher CPM rates for advertisers.

Unfortunately, part of this increased targeting technology effort needs to be focused on scale as the amount of web inventory continues to explode. User-generated content is inventory that has achieved the lowest CPM and response rates due to the engagement of users in the sites they’re on, as well as the high frequency of as impressions each user is seeing. How can improved targeting technologies and ad budgets continue to get more relevant while also scaling to meet the inventory explosion?

When that happens, advertisers, publishers, ad networks, ad exchanges, and most importantly USERS will benefit.

Comments

A report from Bear Stearns analyst Spencer Wang made the prediction based on advertising rates for banner ads.

The CPM rate they used as a comparison is $6. That is quite low, especially considering the demographics of the audience. as BusinessWeek says they are: “business-minded, college-educated professionals with significantly above-average wealth—the sort of audience that advertisers, particularly makers of luxury goods, want to reach.”

The site has an estimated 989,000 subscribers and subscriptions are $79/year (though not all paying that rate). Traffic to the site is about 122.4 million page views a month. It seems a far stretch to get 12x the traffic, but earlier estimates said that it would take much less than that.

The report compares the WSJ.com with nytimes.com, CNN Money, usatoday.com MarketWatch and Yahoo Finance (see graphic).

Comments

First, from Justine, I discovered a new video player from Viddler that has the cool feature of allowing you to tag and comment on segments of a video.

Second, Leo is getting $35 CPM from advertisers because they know his audience buys whatever he recommends.

Hmm, if there was only a way to bring both these items together…

Comments

Tag:

I’ve noticed three common complaints.  First, advertisers complain that Google’s algorithm tweaks do nothing to increase relevance and instead force them to raise their keyword bids.  Second, advertisers also state that Google is trying to learn about their business through CPA advertising and Google Analytics in order to take over their industry.  Third, privacy advocates complain that between search and browsing behavior, Google will know too much about their users.   

Theory #1 – Google adjusts algorithms to force higher bids

Google’s hybrid ranking method has always served two purposes, rewarding relevant ads and increasing their effective CPM (eCPM).  The basic hybrid auction model uses historical click-through rates and the advertisers’ maximum CPCs to calculate Google’s predicted revenue per impression.  The ads that generate the most revenue are ranked the highest.  Those ads also happen to be the ones that users typically prefer.

All the subsequent tweaks have similarly been aimed at increasing relevance and revenue, notably last year’s inclusion of landing pages in Quality Score.  Is Google an “evil” company for wanting to make money?  If they are, so is every ad network and publisher in the history of the world (excluding PBS and other non-profit organizations).  That’s what publishers and networks, (and other businesses, for that matter) do.  They charge their customers as much as they can get away with given product quality, demand, and market value.   

Theory #2 – Google is learning too much about businesses 

By charging advertisers on a cost-per-action basis and giving away Google Analytics for free, conspiracy theorists allege that Google is attempting to learn more about their business.  This empowers Google to enter into almost any industry as a serious contender.  But let’s look at the facts.  In order to learn enough about an industry to be able to start a business that could take over the industry, Google would have to be tracking a significant number of that industry’s minor and major players.  In general, major players choose not to use Google Analytics, and go to great lengths to avoid sharing their conversion data with anyone else.   

Regardless, could we one day see Google leverage their brand, resources, and trove of data to expand into different industries?  Could there be a Google Mortgage (in Beta, of course)?  Absolutely.  Would that be wrong?  Of course not.  It’s business.  If you’re an advertiser and you don’t like it, you’re free to not give Google access to your conversion data.  Or you can always pull an eBay. 

Theory #3 – Google is learning too much about individuals

 Personal users, at the behest of privacy advocates, are starting to get worried about Google collecting their search queries.  If the DoubleClick acquisition goes through, Google will also collect Web browsing behavior.  They will know a lot about what searches are conducted on a certain computer, approximately where that computer is located, and what Web pages that computer visits.  Sounds very Big Brother-ish. 

But what nefarious purpose are they going to use this for?  World domination?  Intergalactic invasion?  Nope, they’re going to use it to show Web browsers more relevant advertising.  I wouldn’t head to my nuclear shelter over that one.  Again, if you’re really afraid, you can stop using Google and/or stop accepting cookies.  Realize, however, that even if you accept no cookies, you will still see ads on the Web.  The ads just won’t be as good. 

The conspiracy theorists might ask what if someone hacks into Google’s system, or there’s an accidental data leak.  What if Google gets subpoenaed for information?  Well, of the search networks that got subpoenaed recently, only one refused to turn over its users’ data.  I’ll let you figure out which one that was.  And hacking?  The truth is, if you don’t know the password to someone else’s PC, you can’t hack into it.  Yet people are talking about accessing Google’s user data as though it’s the kind of thing you do on Sundays between washing your car and mowing your lawn.  A data leak, however, is a serious threat to privacy, one that I would hope Google is dealing with proactively. 

Perhaps Google made a mistake when it adopted the “don’t be evil” motto, which invited people to hold them to a higher standard.   In reality, they have shareholders who expect them to meet targets every quarter and every year.  It’s to be expected that they are going to charge as much as they can for their existing services, develop new products and services to keep their revenue up, and expand into other businesses.  Right now they’re just executing better than anyone else.  In fact, these conspiracy theories are probably less about double standards and more about sour grapes.

Marketing Pilgrim’s Andy Beal compares it to a mutual fund model, and Efficient Frontier’s latest "marketplace" approach. The advertiser sets the ROI and pricing goals and lets Google do the rest.

Instead of setting the maximum CPC or CPM (cost-per-impression), advertisers set the "preferred" CPC or CPM bid, representing the average price a marketer is willing to pay.

"For example," writes Vivian at InsideAdwords, "suppose you want to pay an average CPC of $0.50. Currently, you need to regularly monitor and adjust your maximum CPC bids to keep your costs at or around $0.50 per click.

"Using preferred cost bidding, you can simply tell us that you want your average CPC to be $0.50, and we’ll manage your bids to reach that goal."

The option comes to being through a phased rollout – available to some immediately, and to everyone else over the next few days.

Called an "advanced bidding option available for keyword- and site-targeted campaigns," the option is intended benefit advertisers who:

1.    Would rather target an average CPC or CPM than say the most they’re willing to pay
2.    Want more consistent AdWords advertising costs (no day-to-day) changes
3.    Don’t want to continually monitor and adjust keyword bids

Google says keyword status will not be affected by switching to the preferred CPC bidding, and that advertisers probably won’t see much change in ad position. Ad position may fluctuate, however, as the system adjusts to the targets specified, but position changes should be "relatively minor."

Here’s a big however, however. Position preference doesn’t work with preferred cost bidding. The AdWords system automatically adjusts ad positions as the preferred cost is reached.

The company says it takes at least four days for costs to stabilize, with actual costs differing from actual bids at any given moment as the system works to achieve the preferred bid over time.

Similar to the way the AdWords Discounter reduces the price of the top ad to the minimum needed to maintain the position, the preferred CPC model adjusts so that actual costs may be lower than specified bids.

Google also makes these guarantees:

·    For preferred CPC bids, the actual cost of any one click won’t be more than twice your bid.

·    For preferred CPC content bids, the actual cost of any one click won’t be more than four times your bid.

·    For preferred CPM bids, the actual cost of any one thousand impressions won’t be more than four times your bid. It is very rare that a click or thousand impressions actually costs as much as these limits. Our system allows this flexibility in order to reach your preferred average cost as effectively as possible. In addition, you can always control your costs by adjusting your campaign budget.

Preferred cost bidding works with any ad type – image, local, et cetera.

With more voters researching candidates online than ever before, smart candidates will allocate more of their marketing budget to online display campaigns. Since many candidates have formally entered the election earlier than ever before, they will need to find ways to advertise effectively while spending wisely. With retargeting, they can advertise only to specific voters, and show customized messages to different voters based on their site behavior.

Retargeting ad servers only serve impressions to users that the advertiser specifies. In this case, the candidate’s Web site would drop a cookie onto every computer that visited. Then, when that computer arrived at a publisher site within the ad servers’ network, the server recognizes the cookie and serves a highly-targeted message. Typically, retargeted impressions are charged on a dynamic CPM (cost per thousand impressions). This is a system in which the ad serving marketplace matches the advertiser with a publisher based on the advertiser’s campaign budget and the maximum CPM he’s willing to pay, and charges the advertiser the lowest possible price based on a competitive auction.

Perhaps a comparison to another advertising medium would help illustrate the efficiency of online retargeting. For a TV advertisement, the candidate would pay a CPM based on the networks’ estimated audience. However, the key word there is ‘estimate.’ The data to back up this estimate is about as hard as a mound of Jell-o. In reality, many TV watchers use commercial breaks to go to the bathroom, or to the kitchen to grab a beer, or even surf the Web. A large number of the TV impressions that the candidate pays for are wasted. By only serving impressions to specific users, retargeting eliminates those wasted impressions, ensuring that the advertiser gets exactly what he pays for. That’s an efficient use of a marketing budget.

However, it gets even better when you consider segmentation. Segmentation works by grouping users based on the specific action or actions they made on a site. The ad server then only serves specific ads to members of these groups. For example, let’s say a candidate’s site has an “online donation” function. Many voters will probably come close to using it, but stop during the checkout process. The candidate could then retarget this group of “almost donors” exclusively with messages urging them to complete their donation.

Then, once they’ve donated, the candidate could transfer them into a different voter segment, moving them into the group of “donors.” The candidate could then create a message targeting donors exclusively, perhaps urging them to sign up to be a volunteer, or buy a T-shirt or bumper sticker.

The possibilities here are literally endless. Candidates could even segment their audience by issue. For example, if a candidate had a Web page discussing his or her views on abortion, they could segment every user that visited that page. They could then retarget those users with a message urging them to read a transcript of the candidate’s latest speech on abortion. Deeper analysis of site behavior, like examining referring sites or search engine keywords, might help candidates market exclusively to groups like swing voters, or even voters that belong to the opposite political party.

The benefits of political candidates using retargeting are clear. With smart use of technology and killer analytics, political candidates can use retargeting to make sure that their online display ads reach the right audience at the right price at the right time. It’s simply impossible to do all this with a TV, print, or outdoor campaign. It’s time for candidates to recognize the shift that has occurred in voters’ media consumption habits and allocate their marketing budgets accordingly and efficiently.

Add to Del.icio.us | Digg | Reddit | Furl

Bookmark WebProNews:

Adscape is a video game advertising company whose AdverPlay product lets developers place dynamic ads right inside the game and Real Virtual Gateway product enables two-way text, audio and video communication via SMS Text or eMail.

Also, Google has added a column in its AdWords interface that shows an ads quality score, which determines how much advertisers must pay above the minimum bid if their ad isn’t particularly good. They are also making some changes in how the quality score is calculated.

Google has also taken Webmaster Central out of beta and added comments to its Webmaster Central blog. The blog is now the first official Google blog to have comments, as far as I know, probably due to a desire to have a more official place for public feedback than Matt Cutts blog.

Another change: Advertisers can now use site targeting for pay-per-click ads, not just CPM ads. I can imagine some AdSense publishers not liking this one, others thinking its sliced bread good.

Also, AdSense publishers got two 1099 tax forms from Google this year due to some sort of error. Google says they will not need to worry, as the forms may have been printed twice for publishers, but they were not sent twice to the IRS. Still, I’m worried, because my 1099 had my name misspelled! How do you misspell something when the computer that spits out my monthly checks from Google knows the right spelling? Did someone type up these forms by hand?

Finally, Ask.com is doing a funny thing: If you search for yahoo.com on Google, you might see an ad by Ask advertising their search engine. Yahoo is a pretty popular search term, because some idiots use Google for typing URLs instead of the always-there address bar, and Ask is hoping to catch their attention. Craziness.

Comments

Tag:

Add to Del.icio.us | Digg | Reddit | Furl

Bookmark WebProNews:

“The Google Display Advertising Network was created so Google can go after Fortune 1000 companies, which buy advertising to build a brand more than to sell a product…The goal being to sell these big companies display and video ads at a very high CPM – unlike the AdSense network, the display network is 100% CPM based.”

I recall hearing rumors – or maybe just speculating myself – that this network existed, but it’s interesting that Google is breaking with the tried and trusted CPC model. I wonder if John Battelle is aware that Google is not content with the “milk” of advertising (CPC) and is now going after the “cream” (CPM) – John’s analogy, not mine.

So how do you join the Google Display Advertising Network? Er, you don’t – they select you.

“Google won’t even acknowledge it exists. You won’t find anything written about it in any of Google’s web properties. The only way to get into the display network is if Google invites you, which is how I found out about it.”

Comments

Tag:

Add to Del.icio.us | Digg | Yahoo! My Web | Furl

Bookmark WebProNews:

Andy Beal is an internet marketing consultant and considered one of the world’s most respected and interactive search engine marketing experts. Andy has worked with many Fortune 1000 companies such as Motorola, CitiFinancial, Lowes, Alaska Air, DeWALT, NBC and Experian.

You can read his internet marketing blog at Marketing Pilgrim and reach him at andy.beal@gmail.com.