You have a keyword result, say, "pet rocks", and 20 people clamoring to bid for that keyword.
Now, do you put the people at the top who
a) bid the highest?
b) or have the most relevant landing page to the user and bid the highest?
I think b is an easy answer.
Now, to extend this experiment, how do you pick the most relevant landing page?
do you
a) hand review every single landing page?
b) automate the review of landing pages?
Again, I think b is pretty clear.
Now, finally, do you
a) tell people exactly how to change their web pages to satisfy your algorithms (basically, making it trivial to game) or
b) keep it to yourself.
I'll leave that one up to the reader.
Personally, I really hate what Google has done. But, seriously, what choice do they have?
[edited by: mona at 10:26 pm (utc) on July 25, 2006]
[edit reason]
[1][edit reason] tiding up [/edit] [/edit][/1]
more sign up so no loss to them. In fact the short term boost to earnings as they reached into tons of pockets will probably show an increase in revenue.
That could be part of what happens. But the other source of revenue increase will come from higher CTR's from existing advertisers running better business models, and already bidding higher. That will work where you have a wide spread of bids.
In other words, you remove the poor low paying ads which are also associated with sites with no value added, and the result is you increase the visibility of the higher paying ads going to higher quality sites. Higher clicks at higher rates.
The result - less junk for the user, and the same or more revenue for google.
[edited by: mona at 10:26 pm (utc) on July 25, 2006]
[edit reason]
[1][edit reason] tidying up [/edit] [/edit][/1]
The reason for the secrecy is not so that they don't get "gamed". It's because if we knew how their algorithm worked, it would open them up for lawsuits. Because it is obvious that their algorithm as it is working now - is a major hack job. In order to accomplish what they are claiming to accomplish, they need a precise operation. However they have given the surgeon a machete with this algorithm.
If you've made such a great algorithm that it can determine quality, how can that be gamed? Quality is quality.
You are mistaking the word for a real thing. Reification, it's called. I don't expect that will make any sense to you, so I wish you luck. Your venom shows, as it does with others and obviously colors what you can see and not see.
But I wonder if you could keep an open mind, since you can't read the minds of google.
Might help you if you could do that.
I doubt venom will be all that useful to you.
You know, the sad thing is, I think probably 90% of knew this day would come but just kept on kidding ourselves. I know I certainly did.
All we can really do is evolve or go back and work for the man, I guess.
Now, to extend this experiment, how do you pick the most relevant landing page?do you
a) hand review every single landing page?
b) automate the review of landing pages?
This is the key point... rhe search engine who will do this better will win the race. But not easy.... PageRank is part of the algo surely.... but a mixture of A and B is best, because it can not be reversed-engineered and helps with your next question
a) bid the highest?
b) or have the most relevant landing page to the user and bid the highest?I think b is an easy answer.
Why is it the easy answer?
It's been said before, and I'll say it again - this is not search, this is paid for inclusion.
No one really expects these to be supreme quality links when they click on them. At the most, they expect them to have something to do with it or sell it. Nothing more.
Searchers are not the morons they once were. They all know what sponsored links are now, and knowing that I'm sure they don't give a damn for the page quality so long as it'll sell them what they want.
Quality should be in the search results (which, sadly, it isn't at all). All the quality a sponsored link needs is the money to bid on the word and something vaguely to do with the search term. If its not what the searcher was looking for, they'll go back and look again.
On the flip side... my site does do all the right things. I'd be fine with the quality score, but all the competition with rubbish quality landing pages are now just bidding stupid money and pricing me (the innocent, good quality yet low spender) out of the game.
So much for quality results.
In the beginning, there were few advertisers, relevance of paid ads did not really matter. Today just money / CTR is no longer enough: there are so many ads that Google can classify them in order of relevance, the same as for the organic results.
Based on the range of experiences out there -- from ecommerce sites with moderate min bid increases, to content sites getting hammered with high min bids, to MFAs staying untouched -- I think the hybrid quality/profit algo is the only explanation that sticks.
If short term profit optimization is part of the algo, then G really has no choice but to keep the algo cloaked in secrecy and purely under the guise of a 'quality score'. You can't exactly tell advertisers that they are too successful at getting cheap clicks, and therefore must pay more.
Whatever the case, raising this sort of uncertainty on a product that accounts for all of Google's income is risky. Most advertisers I know don't like to leave giant unknowns (i.e. will I wake up one day to find out Google has put a huge dent in my business?) in their plans. We'll see how it plays out.
5 days later there are 6 new relevent pet rock sellers
But what on earth is a "relevant pet rock seller"?
Perhaps you meant pet rock sellers with relevant landing pages. I should save "relevant", because there's little evidence that Google's algorithm is very good at measuring relevant. So, landing pages that satisfy Google's algorithm.
Is this what users are looking for? Landing pages that satisfy Google's algorithm?
Silly me! I thought they were looking for reliable sources with good prices, good pre-sales information, excellent customer service, etc.
How do you determine that from a landing page?
Unfortunately, Google has removed market forces from the equation.
Take off the artificial constraint of "landing page quality", and market forces should work. Those with low-quality ads, products, landing pages, support, poor prices, etc. will not achieve a level of sales that will support continuing to advertise.
Perhaps you meant pet rock sellers with relevant landing pages. I should save "relevant", because there's little evidence that Google's algorithm is very good at measuring relevant
They don't have to. I've explained this a bunch of times, and it's getting really old. It's all done empirically and statistically, and uses correlates of "quality" that distinguish between sites they like and sites they don't.
It's a type of profiling and use of statistics for prediction.
Landing page is likely a part of the equation. Could be a huge part or a teeny part.
If you want to understand this, you're looking in the wrong place.
Might be an age thing as a "pet rock" used to be a big selling item. Maybe we should stick to fuzzy widgets.
Rbacal, No need to pretend it is all done in some mysterious way. If it is done by computer then the computer could simply spit out why such and such a site did not pass muster. They could easily provide some direction for obtaining a quality landing page. Ok. provide better direction then.
f it is done by computer then the computer could simply spit out why such and such a site did not pass muster. They could easily provide some direction for obtaining a quality landing page. Ok. provide better direction then.
It can spit out information, but that information would be useless to you. Sorry. It doesn't work that way with complex algos.
It could spit out a numeric quality score, which wouldn't tell you which of the interacting variables caused the problems. Since the variables interact (that's the part you're missing), you can't point to one or two factors, because they all interact to produce a score.
The only situation where it could do what you ask is if the algorithm is linear, simple, and ADDITIVE.
Even then you'd have to program an algorithm to analyse the algorithm and translate it into english, which would be a huge and, frankly, wasteful, undertaking. I wouldn't do it.
I'm not going to spend, oh a million dollars worth of programmer time so I can send an informative note explaining to the proprieters of best200sites.tld why their sites suck. And, in the process drop hints to said proprieter as to how they can game the system some more.
Just as an example.
5(w) x 5(w) x 7(w) +(2(w)) x 8(w)...
to yield a QS score that causes a problem. With twenty variables or more, and twenty weights or more, what do you tell the person?
I would if the people i had lost were truly useful sites that were added value to my current advetisers. The 1 mil price tag is very small compared to saving .1% of advertisers. From the looks of forums I go to. I think they lost more then that in advertisers.
Why does it matter how they determine if they like you are not? This is a dangerous path to follow, allowing single entity to determine what is good and not good. This has always been and will always be a bad idea.
Sometimes I think people don't understand what is actually taking place when they are buying ads on google search.
Google markets a product to the general public, a search engine. In order to make money they have unmarked and relatively unobtrusive advertisements on their site."
It is entirely in Google's purvey as to what they like or don't like. Google is not impacting your website or what you do when advertising on other sites. They are simply telling you that if you want to advertise on THEIR page you have to meet certain of their standards. What those standards are or whether they are made known to you is completely irrelevant. Google is in a position where they can do this because they have access and the ability to deliver a LARGE number of visitors to advertisers. You can like or dislike what they are doing, but you can't claim they don't have the right to do it. They have every right to do what they think is in the best interest of their business.
I agree with you, it is their site and they can decide who can advertise or not. Here’s what they cannot do according to the law (Robinson-Patman Act 15 U.S.C. Section 13)
If you and I both sell "paper" via a website and Google allows you to purchase advertising from them at .10 and then tells me that they don't like my site so I will have to pay $10.00 for the same advertising you get for .10 in an effort to leave me unable to compete with you in an open market, that is illegal.
Now if Google was saying that they don't like websites that sell "paper" and all website that sell "paper" will have to pay more for advertising that would be a different story. That’s not what’s going on here; Google is creating price discrimination in an effort to control the market place and restrict trade which is illegal.
limoshawn. You bring up a good point but who has the deep pockets to bring it about.
I'm not sure how good the point is. Limoshawn's interpretation and application of the law is largely irrelevant, as as the case for all of us. The ONLY thing that matters is how the COURTS apply the law.
The courts interpret the law, not you, not me, and certainly not limoshawn. Thank goodness.
I know there has been talk of a class action suit, and that would certainly be done on a contingency basis, if lawyers believe it's a legitimate case (ie. winable). Or the government could, itself, charge google with restraint of trade, IF Limoshawn's interpretation is accurate.
On another forum, someone has spoken to a number of lawyers (or claims to) and none of the lawyers have expressed any interest in being involved in such a class action suit.
You’re correct; for the most part our interpretation of the law is irrelevant. I have based my opinion on how the courts have interpreted the law in previous cases pertaining to price discrimination. The only variable is intent and never before has the intent of the price discrimination been so obvious. Google will tell you what the intent is, an excerpt from an email received from the “AdWords team” clearly states what the intent is: “We understand that high minimum bids are not cost-effective for advertisers, and we don't expect advertisers to continue to raise bids to that level. We're hoping that by setting high bids, advertisers stop advertising on those keywords”
15 U.S.C. Section 13 (a) It shall be unlawful for any person engaged in commerce, in the course of such commerce, either directly or indirectly, to discriminate in price between different purchasers of commodities of like grade and quality, where either or any of the purchases involved in such discrimination are in commerce, and where the effect or intent of such discrimination may be substantially to lessen competition or tend to create a monopoly in any line of commerce, or to injure, destroy, or prevent competition with any person.
You’re correct; for the most part our interpretation of the law is irrelevant. I have based my opinion on how the courts have interpreted the law in previous cases pertaining to price discrimination. The only variable is intent and never before has the intent of the price discrimination been so obvious.
1) Then how about you cite the case law that is relevant. Pronouncements (empty at that) aren't really that informative, and neither is quoting a statute. Even case law is tricky.
2) The intent has "been so obvious" to YOU. Not to me. I don't find jumping to conclusions an effective business strategy, and neither are attempts at mindreading, although that might work for you.
Related -- if people made better distinctions between what is relatively established fact, and what is relatively baseless opinion and speculation, they might be a little more successful with their businesses.
As they say: It's not what you don't know that kills you. It's what you don't know, but you think you know.
“We understand that high minimum bids are not cost-effective for advertisers, and we don't expect advertisers to continue to raise bids to that level. We're hoping that by setting high bids, advertisers stop advertising on those keywords”
Sea-Roy Corp. v. Parts R Parts, Inc., U.S. Dist. Lexis 21809 (M.D. N.C., Uurh. Div.)
Omega Envtl., Inc. v. Gilbarco, Inc., 127 F.3d 1157, U.S. App. Lexis 29863 (9th Cir.)
Hansel 'N Gretel Brand, Inc. v. Savitsky, U.S. Dist. Lexis 13324 (S.D.N.Y.),
Chroma Lighting v. GTE Prods. Corp., 111 F.3d 653, 1997 U.S. App. Lexis 6642, 1997-1 Trade Cas. (CCH) Para. 71,769 (9th Cir. )
George Haug Co. v. Rolls Royce Motorcars, Inc., U.S. Dist. Lexis 13650 (S.D.N.Y.)
Parker Ice Cream Co. v. Conopco, Inc., U.S. Dist. Lexis 8570 (E.D.N.C., W. Div.)
Your venom shows, as it does with others and obviously colors what you can see and not see.
Are you calling what I said above venomous? Wow - you need a thicker skin than that to stay alive in this business. Disagreeing with google's algo does not make me venomous. And from what I'm reading from these forums, the majority here would say that it makes me smart.
I haven't made a killing in this business over the last decade by letting every foolish corporate decision bulldoze me over. And my ranting (you call venom) pays off for me quite heftily. I've managed to get my account to almost pre algo prices by "spewing my venom" at my adwords rep this past week.
