My Opera web browser, normally the king of stability, crashed and completely screwed up my saved tabs, so I’m posting everything old right now, in order to set things right.
Google has tweaked the algorithm that determines what advertisement gets to be at the top of the stack in Google AdWords. Previously, they’d push to the top the advertiser who was paying the most, based on the actual CPC. Now, they promote the ad with the highest potential CPC. In other words, if you bid up to $3.00, but only wind up paying 25 cents, you’ll be ranked higher than someone who is paying 30 cents, if that person has only bid up to $2.00.
Google is testing a program to put ads in Google Gadgets on websites. The Gadget would basically be a non-traditional ad unit, one with interactive parts as opposed to a pitch and a link. Advertisers would bid on getting in the Gadget ad much like they do for any Google ad, by click or by thousand impressions. Google unveiled these ads (which have been rumored and spotted for a little while) at an auto industry marketing summit, but they will be available in all categories when the program launches this summer.
The Gadget ads, as HTML containers that can have Flash, video, real-time information and even the ability to purchase from inside the Gadget, are obviously very attractive for advertisers. However, due to the nature of these ads, and the possibility of annoying punch-the-monkey games and irrelevant interactive content (which is more damaging to sites than an easilly ignored non-interactive ad), Google should consider releasing these ads like their referral ads program, with all ads being opt-in and chosen by websites.
I don’t want advertisers sticking pointless stuff on my site to distract and piss off you guys, but some sort of unit that adds information to the site, like current stock prices, while offering you the chance to sign up and trade those stocks, that is something very useful for everybody. I’d want to choose that, I don’t want it to be the exception.
(via Niall Kennedy > Steve Rubel)
Michael Arrington has an article about companies already in this space.
Google has agreed to acquire Adscape, a company that puts ads inside of video games, for $23 million. Google had missed out on Massive, a company Microsoft picked up for $200-400 million about a year ago, and is going to have to settle for the much smaller Adscape, which it will have to build into a bigger player. Judging by Google’s great success with dMarc, I’d assume nothing at this point.
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 betaand 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.
Google is giving a second go advertising, running ads from AdWords advertisers in newspapers around the country. One advertiser, Chefs.com, has the results of their ad analyzed at PPC Discussions. The results: The ad generated just under 1,000 visitors, visitors who spent, on average, 30% less time and generated 70% less pageviews than usual pay-per-click visitors. They’ve got a PDF of the ad, which, curiously, does not sport an “Ads by Google” line like the first test did.
So, I guess it all depends on how much the ad cost. If you are dealing in an industry where cost-per-click is $5 a click, 1,000 visitors will cost you $5,000. Compare that with an ad in the Chicago Tribune (PDF rate card link), where this ad ran, which gets as high as $755 for a column inch of ad space, the newspaper ad seems like a good deal.
I never thought of it before, but The New York Times reports on a survey by DoubleClick, which claims that twice as many people look at online ads and visit them later, as opposed to those who click on it. What it means:
- Of visitors who see ads on your website, 30 percent admit to sometimes clicking on an ad.
- Meanwhile, 61 percent look at the ad, process it, and visit the advertiser later, without clicking on the ad.
- With Pay-Per-Click advertising getting enormously popular, that means that publishers who display PPC ads never get paid, even though their ads are effective for that 61%.
- The 61% is a complete validation of the idea of brand advertising, that it is indeed twice as useful to get your name in the minds of visitors, as it is to get them to click.
- Google has ads that pay out for just being shown, but it has no type of ad that combines payments for clicks with payments for showing the ad.
If three times as many people visit an advertised site as those who are actually reported on clicking on it, then publishers are losing a ton of money. What can be done?
Well, one option is for Google to stop showing the advertiser’s URL in ads, or at least let publishers turn it off optionally. Advertisers should have to pay a more complicated structure, with ads that include a website URL costing more money than those who don’t. Google should also introduce a combo ad, one that combines CPM (pay for ad impressions) with CPC (pay for ad clicks), charging a complex formula that charges less for impressions and more for clicks.
A combo ad wouldn’t work everywhere, as there are a lot of cheats in the AdSense system, but high-level trusted publishers should have access to it. Advertisers should have the option of getting their ads in at a lower cost-per-click than rival bidders, if they agree to pay for impressions. As an example, an ad that bid $10 per click would be considered equal to an ad that bid $3 for 1,000 impressions as well as $9 per click.
Am I crazy? Would a third type of ad payment kill Google’s ad system and overcomplicate things, or would advertisers welcome an opportunity to get lower per-click ads in the system by paying up front? No one can argue that PPC is perfect, and the Times article makes that clear. I’d like a combo ad as an option that might make everyone happy.
(via Amit Agarwal, who suggests a different AdSense change that might make the ads more user-friendly)
Amazon.com has announced ClickRiver, a PPC service letting advertisers buy text link ads on their search result pages. While ClickRiver will make money just like Google AdWords does, Amazon sells products, unlike Google, which means the ads will be competing with their own sales, and earning far less perclick than Amazon earns on sales. Still, Amazon already runs sponsored links, and the idea of competing with yourself seems silly.
Maybe a smarter idea would involve something more along the lines of Snap.com’s cost-per-action ads, where advertisers only pay when they get a sale. In that case, instead of Amazon making money from the ads, it would come from the sales of their advertisers, just like a commission, and could be competitive with the money Amazon would have made if it had completed the sale itself.
Yahoo is no longer serving contextual text link ads on ESPN’s website, losing the publisher to rival Quigo, who scored a multi-year agreement. ESPN is such a big customer that it effectively doubles Quigo’s ad inventory. Quigo will enable customers to buy ads directly from ESPN.com, and target their bids by site, section, keyword or topic. It must be assumed that ESPN was dissatisfied with Yahoo’s offering, either due to notoriously below-average contextual targeting, low payouts, or some unknown factor.
More coverage by ClickZ, via Andy Beal. Read Andy’s post, which is cheesy-chock-full of sports metaphors.
The Google AdWords blog has announced that they are fulling rolling out “click-to-play” ads. We’ve seen these ads previously, appearing as Flash ads in the style of embedded Google Videos. Unlike traditional animated ads, Google has gone overboard making sure they don’t annoy anyone, because they don’t do anything unless you click a really obvious play button.
These ads can be extremely valuable, for two reasons: Because they require the interaction of the user, they are valuable to the advertiser, since it is quite clear that the user noticed and watched the ad, and was at least somewhat interested in its content, unlike every single other type of ad. Also, because the ads appear similar to Google Videos, any website which occasionally embeds Google Videos is going to earn serious money from these ads, since users will be drawn to the ads and their money-making play buttons.
If Google were to give a way to get only video ads, I’d be there in a second.
First, as with all AdWords ad formats, video ads will compete for placement on sites in the Google content network with other text, Flash and image ads — and, as with our other image ad placements, you can choose to bid on a CPC or CPM basis.
Second, these ads will be supported by both site- and keyword-targeted campaigns. You can choose to serve your video ad on a specific site or on pages in our content network that relate to your product or service. As always, you have the ability to geo-target your video ads internationally, nationally, or locally.
Finally, unlike some intrusive advertising, users will have complete control. When a page loads, only a static image will be visible; the video will not start playing until the user initiates it. He or she will be able to advance the video, pause it, adjust the volume or click through to the advertiser’s site, as you can see in the example below:
But, you may say, video is only for big branding oriented advertisers. We beg to differ. This feature makes video ads much more accessible to all advertisers. Now, an owner of a small bed & breakfast in Lake Tahoe can put a video tour of his beautiful chalet right next to an article that talks about skiing the epic slopes of Squaw Valley.
That isn’t a real ad, so feel free to click on it without worrying about click fraud. However, I might interest you in some ads below….
Ha ha. I’m funny.
Anyway, The New York Times talked to Gokul Rajaram, a Google director of product management, who said that the video ads will generally cost anywhere from the single digits of dollars to the low double digits. The bad news for Google: Since the publishers (guys like me) get a pretty decent cut of the ad revenue, that means Google won’t see a lot of that higher price.
Of course, that’s also the good news, both for publishers, and for the strength of the AdSense network, which has seen slow growth. If Google can fill enough ad inventory with Flash ads, it can revive interest and maybe even score some new publishers. If Google ads a video-only ad format, all those video sites out there that repost YouTube and Google Video content might be inclined to sign up.
I hope someone at Google is thinking about that.
Some bonus quotes:
The activation feature could discourage many advertisers from signing up for the new video service, said Jupiter Research analyst David Card. “This isn’t going to be a game changer for Google, but it gives them a much richer palette.”
In addition, it suddenly gives Google a real product for brand marketers, something the company has wanted but hasn’t really had until now. Yahoo! has had much more to offer brands. It will be interesting to see how many of them will try this out.
Advertisements can now offer video tours, normal TV commercials, CALL NOW call to actions (perhaps with an interesting Google Click-To-Call twist), ability to view the video before clicking over to a site, and strong branding via the multimedia spots.
UPDATE: Check out Darren’s take:
One would pressume that video ads would be worth more than normal image ads or text ads and that publishers would have the chance to opt in or out of having such ads show on their sites in a similar way to them being able to select text and/or image ads.
I would also presume that these ads would predominantly be rectangle box ads and that CTR on these ads would be quite good at least initially due to the novelty factor of them.
Google released an AdWords Traffic Estimator, that, without logging into an AdWords account, will give you a decent idea of what you can expect advertising on certain keywords. You put in keywords, select a language and territories, and can enter a max Cost-Per-Click or daily budget, and see your results.
(via Lisa Barone)
While AdWords users can make use of this as a quick reference, it can also be a great tool for guys like me. For instance, I learned that at a “recommended” CPC (that is, whichever CPC will get you a top position 85% of the time) of 57-86 cents, I can get between 18-22,000 clicks for a search for “google”. Meanwhile, the recommended $1.08-$1.61 for Microsoft means I would get 6-8,000 new readers, costing me seven to thirteen grand in ad money.
Problem is, since I don’t make 57-86 cents per visitor to this blog (which would be a CPM of several hundred dollars), how the hell can a blogger make more money by advertising on Google? Short answer: You can’t. As of yet, there is no way to advertise on Google and make money as an ad-supported business. You really need to be selling something, and something that makes quite a bit of money per visitor, to really justify this.
What can Google do? Again, a tough call. Letting bloggers advertise on other blogs for less money won’t work, since you pay per click. To even have a shot at breaking even, every single person who arrived at your site would have to click an ad. You spent a dollar to get them here, that means you have to make back that dollar before you can even turn a profit. Needless to say, it is impossible.
I’ve been experimenting with AdWords to drive traffic here, and I’ve invested a tiny sum on super-low paying ads. I pay the bare minimum CPM of just a quarter per thousand ad impressions. While very few people actually click the ads (in fact, such a small percentage that it is almost pathetic), my ads have been shown over 200,000 times for just fifty bucks. Is that enough exposure to make it all worth it? Maybe, but I can’t be sure and I can’t track it.
So, that is one option: Pay fifty bucks, and marvel as a 200,000 people (maybe) discover your blog exists. That’s a pretty cool way of getting exposure, except how are we going to know if it works?
I’m thinking it is time for Google to start up a AdWords For AdSense program, one that lets AdSense publishers advertise in AdWords on a backwards basis. Basically, Google lets you advertise on low-paying AdWords keywords and AdSense websites for free, but takes a larger cut of the AdSense revenue you make from the people who click through to your website.
In other words, right now, if you bought 500 $1 ads, and those people clicked on $200 worth of ads, you’d be out $300, plus Google would take their (assumed) 20% AdSense cut, leaving you with $160, a loss of $340. In AdWords For AdSense, Google would just give you $100 and charge nothing for the ads, taking a fifty percent cut of the “new business” they sent your way. Sure, you’d be getting a big discount on the AdWords ads, but you wouldn’t have bought them without the discount, and Google wouldn’t have sold them, anyway, since they are low-paying.
There are issues to consider: For instance, Google would have to consider the other ads on your blog (like AdBrite or BlogAds), since you might make money for those. There are solutions as well. Either Google could ban any other ad programs on those pages, or require you use Google Analytics, track the performance of the competing ad programs, and take a cut from there.
To give the new example: Google sends you 500 visitors. You make $50 from AdSense ads, but also sell $20 of Blogads and $15 for AdBrite, and sell two Amazon Associates books for $5. You made $90, so Google only gives you $5 for the AdSense ads, instead of the $50 you’ve earned, since it is taking a straight-up fifty percent cut of your ad revenue that results from this new program.
I think it’s a killer idea, and I’ll tell you this: If I were working at Google, this would be my 20% project. Could someone at Google try this out, helping bloggers and Google extend the long tail of advertising? I hope so.
BuyGoogle’s got the fascinating tidbits from the Google earnings announcement conference call, including this biggie: Over a billion dollars in ad money that was allocated to Google went unspent. Although the period in which this happened is unclear, it is still a significant item Google will want to handle.
What would cause money to be not spent? I’m thinking it’s money allocated to ad spends that wound up losing the Google AdWords auction process. After all, you can plan on spending $4 a click a million times over, and spend exactly $16 for four clicks since you were outbid by a dime per click. Oops.
Of course, the easy (and, of course, not at all easy) solution is that this is an inventory problem, not a system problem. Google needs to find places for this money to go. After all, its not like the per-click prices can go much higher; if it isn’t mesothelioma, it isn’t worth more than what’s currently being paid. Considering how poorly AdSense is growing, I’m not sure how Google can grow that network without slashing profits from that division. Maybe they need to buy one of the smaller players that have a lot of publishers.
This didn’t take too long: As Google Blogoscoped shows, one blogger has already figured out how to combine Google Related Links with advertising, simply by putting the Links box on top of an AdSense ad. This seems to work best with link ad units, but I’m willing to bet it’d work nicely with full-size ads as well. Lets see some experimentation! I want confidential email with click-through rates…
Other stuff I’ve been meaning to get to:
Here’s another good ad placement.
Clear Channel has a deal to use Google AdSense and AdWords.
AdSense CTR’s now go to two decimal places. Useful for those that track this stuff religiously, and those that have to.
Google is seeking a patent for ad-subsidized wifi web access. Not exactly an original idea.
Take a look at Google’s GeoAds beta, for advertising in Google Local with graphics in the ads.
AdWords advertisers can now bid on which position they want.
Verizon SuperPages is selling Google AdWords.
Marketing Sherpa released a report that showed the average cost-per-click rose 24.8% in 2005, from $1.29 to $1.61 on Google AdWords. Yahoo’s CPC jumped 30%, from $1.03 to $1.34. Shopping search engines saw an even better boost, going from $1.87 to $2.56, a 36.9% increase. Looks like a good year for all.
Who thinks this year will be a good one too, but not as good?
(via Search Engine Guide)
AdSense publishers have the option of filtering out ads from websites they don’t want to see. Usually, this can be used to block competitors from advertising on your site, but some people use it to remove ads they believe are irrelevant or low-earning. Nitrous at WebMasterWorld decided to see what would happen if he deleted his 200-site filter list. He had reached the upper limit of the list, and couldn’t add any more, so he decided to see what would happen if he didn’t use any filters at all.
Lets see! Will post later on to let you know what the figure is before I go to the pub! Normally it has always been between 75 and 95 (by 9.30) here in the UK.
Well its only about 15 percent down from normal so far, but the mfas have only just reapeared an hour ago.
By this time of night it should normally be about 60 dollars. At the moment its 46. Thats about a 30 percent drop from the “norm” so far. And the lowest I have seen it at 7.30 pm in over a year.
By now I normally have between 75 and 90. Probably nearer 90 since its a tuesday.
So far we are up to 51 which is a record low over the last year! Traffic is normal to good.
ECPM down by 25 percent.
Earnings are 74 dollars… Instead of a hundred or so expected.
Dont know why. But I may need to continue this for another day. Maybe not! Depends how I feel when sober.
Result was, lower income by approx 21 percent over the previous average from the months before. Not as bad as I expected but bad all the same! Click through unchanged, epc down a bit.
So, what have we learned? While the filter will take out some of your higher priced ads, you need to filter out any ads that appear often on your sites and are completely useless to your visitors. Filtering out Text Link Ads and Rojo would be a good start, as they buy too much ad space that blog visitors will see the ad every day, and get bored with it. Just don’t filter out good advertisers in the belief you’ll get something better.
This is a case where it would be really useful if Google gave stats on individual advertisers and showed how effective those ads were. I’d like to know specifically which advertisers are dragging my site down, and knock them off. Someone has to release a stats package that reads AdSense ads and tracks which ones get clicked on, as well as how many impressions they get. I’d pay for it.
(via SE Roundtable)
Google has announced (via its blog, no less) that it will cost them up to $90 million in compensation to settle a click fraud lawsuit from Lane’s Gifts. Basically, Google says that while it can’t reveal the confidential details of the settlement, the main result will be that it will offer a lot of credits to advertisers who have suffered invalid clicks, credits totalling up to $90 million.
Google currently allows advertisers to apply for reimbursement for clicks they believe are invalid. They can do this for clicks that happen during the 60 days prior to notifying Google. Under the agreement with the plaintiffs, we are going to open up that window for all advertisers, regardless of when the questionable clicks occurred. For all eligible invalid clicks, we will offer credits which can be used to purchase new advertising with Google. We do not know how many will apply and receive credits, but under the agreement, the total amount of credits, plus attorneys fees, will not exceed $90 million.
If I were Lane’s Gifts, I would push for a part of the settlement to force Google to widen that 60-day window, at least a little, permanently.
The probable reason Google is announcing this is to provide warning for investors, since a $90 million hit in its revenue needs as much advance warning as possible. Google doesn’t want any unexpected charges dumping $70 out of their stock again.
For the finance folks out there wondering how we’ll account for this, we can say that the attorneys’ fees (which will be determined by the judge) will be charged as an expense, most likely in the first quarter, once the amount is determined. The credits will be recorded as a reduction to revenue in periods in which they are redeemed.
According to the San Francisco Chronicle, the other defendants in the suit were Ask.com and Yahoo. Because Ask was only being sued over the Google ads it uses, the settlement means they’re off the hook. However, Yahoo is still in, and fighting, and says “it plans to defend itself vigorously against the accusations”.
Also sued, according to Elinor Mills: Go.com (as Buena Vista Internet Group), FindWhat.com (as Miva Media), LookSmart, Lycos, and Time Warner (both AOL and Netscape). Danny Sullivan has interviewed attorneys from both sides about the settlement (naturally, the plaintiffs were more talkative and forthcoming than the defendant).
Google has acknowledged that when it changed the AdSense Referral period to only count people who made $100 in 90 days, it didn’t give enough time. As a result, the latest AdSense update gives a good 180 days, a much more reasonable period, for your referral to earn their $100. I still feel like it should be an open-ended program, but at least Google improved it somewhat.
From the Inside AdSense blog:
You may have noticed today that the time limit for AdSense referrals is now 180 days. Based on the feedback we heard, we agreed that 90 days may not provide enough time for your referred publishers to complete earning $100. Therefore, we decided to double the window. This change is retroactive, so it will also apply to AdSense signups that occurred more than 90 days but less than 180 days ago.
Darren Rowse details the other changes to AdSense. Basically, you can now see how much of your visitors see contextual cost-per-click ads vs. site targeted cost-per-impression ads. There’re other feature change/tweak, go read it at Darren’s blog.
I would be interested in hearing what percentage of people’s ads are CPC vs. CPM, especially since Google hasn’t said yet that you can’t share those numbers. In general, about 10% of my earnings and just over 10% of my impressions are site-targeted ads. About 20/22% of the first ads on my front page are site-targets, 35/60% of the second ads, and 25/75% of the rest of the front page ads (with the first number being impressions and the second being earnings). As for ads on post pages, about 7% are site-targeted, and they earn about the same seven percent. As for the top of the page ad links, .0005 percent are site-targeted, and they earned basically nothing. This is definitely some interesting data, that many publishers are going to want to review more in the future.
Randy : two of the three CPC ads have phone numbers on them.
“The whole point of CPC is to pay for leads when the user clicks on the ads.
Allowing phone numbers in the ad text is simply wrong and Google must disallow this practice,
if they don’t already.”
Doesn’t this undermine the entire pay-per-click system? If ads contain these phone numbers and people start to call instead of clicking on the ad, how can a publisher make money from that ad? He gets paid per click, not per incoming call.
via [Besting Adwords]
I’ve decided to start a new series on this blog, which I’m calling Perspectives. I realize that I end a lot of posts with “we’ll see where this goes”, but we never get to see, since I’m not going to repeat the same stuff every day until a conclusion is reached. Blogs tend to be so obsessed with the “now”, that we never pay attention to what’s happened in the past, and we forget things that might be important. So, every day, unless I’m way too busy (or not around), I’ll post about whatever I was discussing a year ago, and, as time goes by, two years ago, and so on.
On January 1, 2005, I only had one post, part of my contract with Marqui. It seems strange that it was only a year ago that people were trying desperately to figure out how to make money from blogging. While there are some fancy sponsorship agreements, the majority of money these days does seem to come still from advertising.
A year ago, all we had was AdSense, as well as the traditional companies, many of whom didn’t shoot low enough for the blogger market, or weren’t interested in out traffic. Two major competitors have emerged in the last year, Yahoo’s Publisher Network and Chitika. I use all three concurrently, and my experience is:
- Google’s ads are the most relevant, getting twice as many clicks as Yahoo
- Yahoo’s ads are worth more money, paying twice as much as Google
- Combined, Google and Yahoo earn roughly the same. If you can’t get relevant ads out of Google, Yahoo is better, but if you can optimize your content for relevant Google ads, you can earn a freakin bundle
- Chitika, not being a major company like Google or Yahoo, doesn’t yet have a trustworthy enough reputation. But it does add a few bucks in my pocket
- I’d take a sponsor, if offered, but I’m not looking
- Blogads is still a great program, putting a few hundred in my pocket, but they need to work on both smaller (pay-per-day) and larger (full site sponsorship) solutions.
This has been “Perspectives”, I’m Lionel Osbourne.
Well, this is different…
Google has replaced the vertical line that seperates ads and search results with a nice Christmas decoration, and
only on a search for, you guessed it, “Christmas“.
Killjoy me, my first thought was, “This will kill CTR”, but when I saw that it was only on that search, I guess it’ll be so unusual that ad clicks might increase slightly.
Enough! Enjoy it…
(via Biz Stone)
UPDATE: Philipp says that it also shows up for “peppermint“. I have found it on a search for “candy cane“.