But does anyone care to speculate as to whether there's a general rule that correlates CPC for positions 1-5 to # of bidders for that keyword?
Thanks,
Shorebreak
Don't ask me where they come from.
Adwords is unlike other CPC programs in which you buy your position. Adwords rates ads not only on the CPC but also CTR so if an ad is getting a 90% CTR and paying .05 it will out rank an ad that gets a CTR of 0.5% paying .10
Google didn't build its success on selling out to the higher bidder, its all about relevancy and more relevant the ad the better for the user and most importantly money doesn’t equal relevancy.
I'm not sure which was the main motivator for this model - relevancy or revenue maximisation :-))
I am sure, it was revenue maximization, but you already knew that!
Google is now a 95%+ PPC search engine, amazing it keeps the current popularity!
Adwords looks at the performance of your ad to gauge where they rank you along with your CPC. This is why so many smart Adwords users are taking the time to program local targeted Adwords.
Good example of this is: I’m based in Las Vegas and I do a search for "stripper" now a quick review of the Adwords that are shown and I see:
<edit>
Now even if this advertiser would pay give a maximum of $50 CPC Adwords would be more likely to show a advertiser paying a maximum CPC of $1 that had say this ad:
<edit>
The reason behind this is simple, Adwords knows that the ebay ad gets very low CTR but the LVS ad get a better CTR so in the end Adwords does make more money by showing the $1 rather then the $50 as out of 100 people 80 people will click the LVS ad where are only 1 will click the ebay ad.
[edited by: eWhisper at 12:21 pm (utc) on Nov. 21, 2004]
[edit reason] Please Don't Copy Other People's Ads [/edit]
I would argue that Overture's system is better for the *advertiser* experience than Google's max CPC * CTR system.
click-throughs mean nothing, conversions are everything, and a free-market ppc system where position is based on CPC ensures that smart advertisers who know their ROI metrics are positioned exactly where they want to be to maximize against those metrics. Google's system makes the assumption that clicks = success; that's true for *their* success, but the true measure of both searcher and advertiser success is a transaction, and so, IMO, THAT'S what they should be optimizing to, and will ultimately net them a stronger, more consistent revenue model.
I would argue that Overture's system is better for the *advertiser* experience than Google's max CPC * CTR system.click-throughs mean nothing, conversions are everything, and a free-market ppc system where position is based on CPC ensures that smart advertisers who know their ROI metrics are positioned exactly where they want to be to maximize against those metrics. Google's system makes the assumption that clicks = success; that's true for *their* success, but the true measure of both searcher and advertiser success is a transaction, and so, IMO, THAT'S what they should be optimizing to, and will ultimately net them a stronger, more consistent revenue model.
Point well taken, shorebreak. I would argue, though, that CTR as a measure of relevance of ads to our users is all important. Before a transaction takes place, they have to first click on an ad - and they won't click unless the ad is relevant to the search they've done. The relevance of the ads that users see (as measured by them, and reflected in CTR), is all important, so that they'll trust and value (and click on) AdWords ads over time.
And remember, it is not the number of clicks that matters at all. It is the ratio of impressions to clicks. In other words, what percentage of users, having done a search, find the ad relevant enough to click on.
To put it into a slightly different perspective, in which it is not all about money:
Excellent success: 10 impressions with 5 clicks at $0.05 = 50% CTR. Total revenue $0.25.
Not successful: 1000 impressions with 4 clicks at $1.00 = 0.4% CTR. Total revenue $4.00.
AWA
Searchers' goal is not to successfully click on an ad. Their goal is search--->find--->obtain, with an emphasis on 'obtain'.
An ideal search engine minimizes total # of clicks for the user while maximizing total # of transactions for the advertisers. IMO, Google's max CPC x CTR model accomplishes the latter but not the former.
I hear your opinion, though, on clicks being a *measure* of user satisfaction, but ultimately think that searchers want less searching and more finding.
From a long-term revenue maximization standpoint, google could try a system as follows:
For all ads above the network average CTR, the CTR portion of the equation CTR * max bid = rank gets weighted more and more heavily, and likewise, for all ads under the network average, the max bid portion of the equation plays more of a role. E.G. something like:
Ad CTR..Bid WT..CTR WT
8%......30%.....70%
4%......40%.....60%
2%......50%.....50%
1%......60%.....40%
0.5%....70%.....30%
This would have the effect of drawing more and more eyes to the ads over time.
Food for thought.
I hope that this isn't too far off topic. When I modify the bid price of a keyword I like to see if that price has made a difference to my position so I do a search and see if I've moved up, then I do another search and another search etc.
Is this stupid? But I can't wait 24hours to see my Av. Pos.
Shorebreak - if efrontier will potentially modify my bid every 10 minutes does it mean you have access to data that I do not - I mean my average position over the last 10 minutes - because that's the information I need to make an informed decision about what to do next. And the same goes for gotoast/atlas onepoint I guess?
When I modify the bid price of a keyword I like to see if that price has made a difference to my position so I do a search and see if I've moved up, then I do another search and another search etc.Is this stupid? But I can't wait 24hours to see my Av. Pos.
Personally, I don't fret about loading a page with my ad a couple times a day to sometimes check on things, but any more than that and I feel guilty about messing with the real CTR of the system... If it's a high-traffic keyword you can ethically get away with more views, if you must.
I've read threads on here (some just recently) where people admitted to loading their ads across IP addresses dozens of times a day to get metrics info. That's a big no-no, especially because a schemer could do such a thing to purposefully try to bump lower ads below critical CTR and off the map.
If 10-20 advertisers are all checking their ads dozens of times a day for a KW, that can have a real effect, especially on lower traffic terms. In general it's a bad practice, so i try to avoid doing it.
I agree with AWA that CTR is very important cause if the CTR is low it means that the ad is not relevant. And if and ad is irrelevant in SE then the meaning of the SE is busted. But I hate when adwords disables some of my keywords which convert.ARGHHH.gonna kill ya stupid machine(that brings me revenue).:)
Asen Yanov
To put it into a slightly different perspective, in which it is not all about money:
Excellent success: 10 impressions with 5 clicks at $0.05 = 50% CTR. Total revenue $0.25.
Not successful: 1000 impressions with 4 clicks at $1.00 = 0.4% CTR. Total revenue $4.00.
The only metric at hand is "did the advertiser profit?" Understanding this simple thing is how you benefit your customers. Which in the long run, is good for business. This statement from AWA elegantly demonstrates that Google still doesn't get it, and here's why:
In the former case, an advertisement with a 50% response rate is unheard of in the field of marketing. The item would have to be so obscure that there is simply no competition for the term from other advertisers or from natural results. Given the minimal spend on the term, it's a safe bet the item is either of low value to the advertiser, or they're the only game in town anyway.
In the second example, we have a very reasonable CTR, 0.4%. The advertiser is spending a bundle of money on each click, and so is obviously much happier to get a click through. Result? Google makes more money. Advertiser makes more money. Consumer spends more money. So in who's world is this a less successful scenario? Let me explain with an over-simplified example.
Ad number one, with the fantastic 50% CTR and 0.05 CPC, is for "light mauve left handed widget". Not many of them to be had, not many natural results, and not many advertisers. High CTR expected.
Ad number two, with dismal 0.04% CTR and $1.00 CPC, is for "new car". Trillions of dollars spent every year, millions of natural results, thousands of advertisers, clamoring for a single click thru.
But in Google's eyes, the former ad is "successful", the latter is not. I'm not sure how you can reach through such short-sightedness. Or why you'd compare apples to oranges. My theory is that Google just makes stuff up as they go along, and really doesn't know much about commerce, customers, service, and so on.
To say that high click-through rate = relevant is great for the G business model. An adult site with a nice family friendly cloaked title and decription in the natural results would draw a lot of clicks and be highly relevant on a search for Disney Toys by that definition.
We all know that if Google measured the success of their editorial engine like this they'd end up on the scrap heap with AltaVista. Obviously ads have to pass editorial review so mis-matches like this between ads and landing pages don't show up (at least not for long) but from the users point of view what matters is the end result, finding the right info, making a purchase, etc..
If you have all your ducks in a row and do a good job of matching targeted ads to targeted keywords and landing pages that provide exactly what the person is looking for then the user is happy, as is Google and the consumer.
I think the G model is better than OV because if you are good, you'll probably generate more sales at a lower cost. It does suck though if there is a keyword you want the top spot for and the number 1 bidder is pulling a 30% CTR as it can cost a FORTUNE to claim that top spot.
The only metric at hand is "did the advertiser profit?" Understanding this simple thing is how you benefit your customers. Which in the long run, is good for business. This statement from AWA elegantly demonstrates that Google still doesn't get it...
But in Google's eyes, the former ad is "successful", the latter is not. I'm not sure how you can reach through such short-sightedness. Or why you'd compare apples to oranges. My theory is that Google just makes stuff up as they go along, and really doesn't know much about commerce, customers, service, and so on.
Welcome to WebmasterWorld, josebrwn.
And with that sincere welcome, I also have to add that it bothers me just a bit to have been quoted out of context, and to then have a broad (and unintended) meaning applied to my quote.
The quote you've pulled was made within the context of illustrating that (IMO) CTR is a meaningful measure of relevance of ads to our users, that relevance as judged by our users is important to Google, that the raw number of clicks received is not an issue, and that 'success' is not measured by how much money AdWords makes on a keyword.
The only metric at hand is "did the advertiser profit?"
I hope that you'll not be surprised to find that I wholeheartedly agree with this, as one measure of success for advertisers. As a company, though, Google must also consider the measure of success for those who search on our site.
AWA
I should clarify, if it is necessary, that I understood you intended to show the altruistic nature of your ranking mechanism. I found the point trivial and meant to show that your mechanism is short sighted and compares apples to oranges, no matter how well intentioned, and I believe you demonstrated this perfectly.
A search term that garners 50% CTR on 10 page views is considered by Google to be "very successful". But I say that such an ad gets this favorable estimation by Google because it is either:
A) obscure content, or
B) really far back in the rankings.
In the case of A) we have different semantic spaces, i.e. your "excellent" and "poor" keywords have nothing in common, while for B) the ads are far apart in rank. If they're the same keyword and "poor" garners 0.4 CTR on 1000 views while "good" gets 50% CTR on 10 views, how many pages apart would they have to be? In both cases, they simply shouldn't be compared.
Now, if you can manage to eliminate cases A) and B) in your intended message, which may yet be escaping me, I believe you are left with this:
C) the ad that Google values and will rank most highly has the most compelling creative.
Well. We all should be careful to write nice creatives. But let's not mistake altruism or arcane mathematics for something this simple, that the more interesing and appealing creative tends to get a few more clicks, when CPC is constant.
Meanwhile this ignores the very good point made above that it is the more interesting and appealing destination that really matters to the user and the advertiser, and I think it's a stretch to imagine that there is some correlation between a compelling creative and a compelling destination. The compelling destination determines revenue, and revenue determines bid. Thus the more compelling destination will on aggregate be bid higher, and everyone's happy, as I tried to say earlier.
In other words, perhaps consumer satisfaction is not measured by CTR, but by conversions, which hopefully will forevermore remain something the provider can only infer from CPC. What I am getting around to is that I would add to "shorebreak"'s comment that the Overture model is not simply better for the advertiser, it is also better for the consumer. Who we all care most deeply about, right?
