|Quality Score Experience|
Wild variations in Quality Score causeing ads not to show
| 7:22 am on Dec 26, 2008 (gmt 0)|
I have a local karaoke business. I made a campaign whose keywords are the names of 3500 artists who have karaoke songs at my place. I inserted the keyword into the ads using Dynamic Keyword Insertion and made automated landing pages to match each keyword. My campaign is only for a small three-city region in the US.
If someone is searching for a particular artist in my local area, my system serves an ad that says "Sing [Artist name]'s songs at MyKaraokePlace" and directs them to a landing page that says "You can sing these songs by [Artist]:[Song 1],[Song2],[Song3]...at MyKaraokePlace".
This seems to follow Google's rules exactly. I have seen many similar ads for CD retailers such as Amazon.
Initially I had poor quality score for some keywords, then I made some new ad text variations and the score rose.
Now I find that instead of the bids going down, the bids are going up and many of the keywords are getting $5 or $10 minimum bids even though there are no other ads (or maybe another one only) on the search engine result pages.
There seems to be no method to this. It doesn't matter if the artist is popular or not, or in the Top 40 or not.
I am getting reasonable CTRs for the keywords/artists that have shown. Some bands with generic names like Kiss of Love have 0.06% CTR but 99% of artists are in the 2-5% range. Even so, a rising number of those keywords have the $5 minimum and a few even have $10 now.
Examples might be George Harrison $5.00, Frank Sinatra $1.00, James Blunt $0.60, Maroon5 $1.00, Meatloaf $1.00, Lynrd Skynard $5.00, Wang Chung 10.00!
There's no logic to it at all. All the ads and landing pages are the same format. There should be no reason why Moby has a Quality Score of 6 and Hoobastank has a Quality Score of 2?
The thing I am concerned about is that many of the keywords have been automatically assigned a low quality score/high minimum bid even though they have yet to be shown. This means that they don't get the chance to be shown, so I can't improve the CTR and lower the bid. I would expect these ads (if they were shown) also to have CTR in the 2-3% range.
I raised the minimum bids to $5 for some of the ads and then I got hit for a $5 click for a single ad click.
The only thing I can think of is that sometime in the past the keywords were used in some other business campaign and did not have good results, so now that historical information is affecting my bid price. But my business is very different from anything else so I don't think it is fair to use historical data to close my keywords out.
Is there anyone high up at Google who actually listens to these types of problems? All I seem to get from support is "you're doing it wrong", even though I think I am following their own best practices.
For the small business owner dealing with Google is just an enormous pain in the ass. It feels like they are trying to make it as difficult as possible for me to give them money. The results pages have no ads on them, so why not just let me put ads on there -- I'm happy to pay $0.50 a click for all of them. Once a substantial part of my keywords moves to $10 then I'll just give up I suppose.
I don't have any similar problems with Yahoo, just put the ads on and they run with no problems. Actually traffic is about the same, but I'd like to have both.
Well I thought I'd share my experience and ask if anyone else been able to deal with a similar situation.
Also, I have three ad variations. Is it possible that Google is using different ad text each day to determine Quality Score?
| 3:24 am on Dec 29, 2008 (gmt 0)|
UPDATE: Well I thought that after putting the minimum bids on the keywords up to $5 that the ads would show more and that the good CTRs would bring the prices down. Well they did show (not actually that much) more but, unsurprisingly, I was charged $5 for some clicks, so I returned all the bids to $1, still more than the $0.40 I think they are worth.
Example: I was charged $5 for the keyword "Sarah McLaughlin". The CTR was 50%.
Because I have reduced the minimum bid to $1, two thirds of my keywords will not show now.
I have done all the correct things...
1. My karaoke business is completely relevant to my ads and to my keywords. Just to clarify, I have a private karaoke room business where people come specifically to sing songs by their favorite artists.
2. I am in a very defined geographical area of just three small cities.
3. The ads use Dynamic Keyword Insertion so they include the search term in them
4. I am achieving high CTR on the ads that are shown (in the case of Sara Mclauglin, the CTR was 50%, but most are 2-5%
5. I have a relevant landing page that lists the karaoke songs we have based upon the user's search and then leads them to our main site.
During the past few days I noticed that the keyword for Eartha Kitt went up to $100, now down to $35 a click. In my local market of three small cities there is only a single advertiser against this keyword (The New York Times running her obituary).
This leads me to believe that Quality Score is simply a way to extract money from the advertiser. Rather than letting the advertiser get the benefit of the extra clicks that would come from more exposure as Eartha Kitt's story becomes popular, Google simply decides to price the keyword as high as it can, hoping some fool will pay the higher bid. This to me seems hopelessly self-defeating. Surely it is better for Google to get 100 clicks at $1 each than zero at $100.
The emphasis on Quality Score means that advertisers have to be aware of the popularity (and other unknown factors) of a keyword at all times, a huge management burden on the advertiser. It is extremely disheartening to follow Google's best practices and to see the minimum bids rise to unaffordable levels. It is infuriating to see high CTR keywords reset to $5 or more for a first page bid when the first page is almost empty.
Another factor is that the technical support I have received has been quite poor. I asked for someone to optimize my site and to check why similar keywords appeared to have different Quality Scores. I was told that I was confusing CTR with Quality Score. After I told them that I most certainly was aware of the difference, I was told that the Quality Score was made up of different factors and was given a list of items I had already done. It would nice to know what the real factors were so I could work to improve them but frankly I don't think the people at Google support really have an answer other than it's our algorithm and if it doesn't work for you then too bad. I never received the optimization. By contrast, Yahoo called me on the phone to help me upload my ads to their system.
I have to conclude simply that Quality Score is hopelessly broken. It does not work in local markets. Having to check the perceived "quality" of keyords every day places an enormous management burden on the advertiser, which simply saps the will to live.
It's tiring enough to start and promote a small business without Google making it as difficult as possible. I would be spending $1000 - $2000 month if they would let me. Maybe not huge, but it's now gone to Yahoo and Facebook. I am certain that many other potential advertisers, who have far less patience than I, have also voted with their wallet.
All I ask is for Google to fix Quality Score and let me advertise my business on their site. Until then I have to say it's simply not worth it.
| 3:30 pm on Dec 29, 2008 (gmt 0)|
In this case, I don't think the Quality Score is broken; I think you just aren't quite clear yet how it works. (Not your fault, Google still doesn't do a good enough job explaining it)
The thing is, if you are bidding on the artists' names as keywords, you're never going to get a good quality score, because those keywords aren't relevant enough. You aren't providing those artists, or actual product by those artists, you're providing karaoke versions of songs by those artists. It's not the same thing, and Google wants it to be very clear to the user what he's going to get when he clicks. So if you want a good quality score, you should probably append "karaoke" to them:
frank sinatra karaoke
lynyrd skynyrd karaoke
etc. If you use broad match, you might pick up some searches for frank sinatra and lynyrd skynrd as well, but make sure you bid LOW on the broad matches.
You could also try 'songs' instead of 'karaoke' but I'd think that would probably have less chance of a decent QS, as you aren't selling MP3s or CDs by these people.
Your keywords have to really reflect whatever it is you are offering. You are NOT offering Frank Sinatra or Sarah McLaughlin, you're not even offering their own recordings, and your keywords / ad / landing page all need to reflect that. That's how you get on the road to a better QS.
Once you start building a good CTR and some good history, you can branch out a little and experiment with some broader keywords.
I had a similar situation to you - I have a site that lists events for most if not all the cities in my state. The first time I tried it, I put all 937 cities in as keywords, and used DKI. The QS tanked almost immediately; my minimum bids were between .40 and 1.00 for everything.
So I then reworded my ad, and added the type of event to each keyword:
ann arbor event
and so on. My QS is 9 or 10 on about 925 of 'em (for some reason it doesn't like the cities with more than two words in them) and I pay between .02 and .04 a click. For the past 3 years.
My keywords weren't good when I just used the city names, because my site and my ad wasn't about the entire city, it's just about this particular type of event IN that city. That made the difference.
| 7:01 pm on Dec 29, 2008 (gmt 0)|
Netmeg, thanks for your reply.
I can understand how your example can cause problems. When people search on Detroit, there are a multitude of items that could be related to that keyword, and competition is strong both ousdie and inside each city. I had a similar situation in a previous business where I wanted to advertise against the "Japan" keyword and Google wanted me to advertise against "Japan News" instead.
However, I am in a very local area, and there is very little competition for the keywords in this area. One or two ads for most of the keywords.
Also, unlike a city, practically speaking, there are a limited number of items you can sell against an artists name (music, ringtones, tickets, t-shirts, news and maybe a few more).
My keywords are directly relevant to what I offer -- the ability to come and sing the artist's song at my establishment. The songs are our product.
I just got off the phone to Google where we were discussing the "Erasure" and "Eurythmics" keywords which has a $5 minimum bid. There's only one competitor ad on the Eurythmics search engine page and none on the Erasure page. What would be the harm to Google or to the user to allow an ad for Erasure or Eurythmics karaoke on those pages?
The Google expert also told me that I should not advertise against artists and just concentrate on Karaoke related keywords only. That's like asking Amazon to only use keywords like "bookstore".
>So if you want a good quality score, you should probably append "karaoke" to them:
I understand that adding "karaoke" to the ad will improve the quality score, however, I won't get any clicks. How many people do you think in a local area are looking for "Britney Spears karaoke"? I'd get 50-100% CTR, but probably only one click a month. I'd much rather have 3% CTR and 30 clicks a month. I also think that my customers would prefer this because they would have found a service that they were interested in (and Google would make more money).
Also, do you think that Amazon adds "CD" or "Book" to the end of all its artist-name keywords? Amazon is not offering Frank Sinatra or Sarah Mclaughlin either, but their ads appear on those keywords.
Anyway I will try it with one ad group and see if it makes a difference.
> etc. If you use broad match, you might pick up some searches for frank sinatra and lynyrd skynrd as well, but make sure you bid LOW on the broad matches.
Can you clarify this? Are you saying that the keyword "Britney spears Karaoke" will actually trigger ads when the user inputs "Britney Spears"?
> Google wants it to be very clear to the user what he's going to get when he clicks.
The user also makes up their mind what to click based on the ad text. After inputing the search keyword the user looks either through the results or at the ads. The user can choose between different advertisers offering Britney Spears CDs, Britney Spears T-shirts, and Britney Spears Karaoke. If users think the ads are not relevant they simply wont click on them, the CTRs will be low and the advertiser will stop advertising.
| 7:24 pm on Dec 29, 2008 (gmt 0)|
|However, I am in a very local area, and there is very little competition for the keywords in this area. One or two ads for most of the keywords. |
Competition has nothing to do with it. It's relevancy. It doesn't matter if it's detroit or paw paw, it's not relevant as is.
Likewise, if you are using keywords that are just the artists' names, that's not relevant enough, because you are not selling those artists, nor product created by those artists.
|My keywords are directly relevant to what I offer -- the ability to come and sing the artist's song at my establishment. The songs are our product. |
No, they're not, I'm afraid. You're selling an experience, not a song, not a product. If you use a keyword like "Britney Spears" to advertise what you're selling, it's not the same thing, and therefore not relevant. Not by AdWords standards, and they wouldn't be by my standards either. I'm sorry.
|The Google expert also told me that I should not advertise against artists and just concentrate on Karaoke related keywords only. That's like asking Amazon to only use keywords like "bookstore". |
Amazon sells products created by their authors; the keywords and ads reflect this. (They also no doubt have a special deal similar to eBay, just due to their size)
|I understand that adding "karaoke" to the ad will improve the quality score, however, I won't get any clicks. How many people do you think in a local area are looking for "Britney Spears karaoke"? I'd get 50-100% CTR, but probably only one click a month. I'd much rather have 3% CTR and 30 clicks a month. I also think that my customers would prefer this because they would have found a service that they were interested in (and Google would make more money). |
If they're not directly looking for what you have, then AdWords is probably not a very useful or efficient advertising solution for you. And you might prefer it, but Google won't prefer it, because in their eyes, your advertising campaign won't offer a good experience. Because if I'm searching for Britney Spears, and JUST Britney Spears, I won't want to see your ad, even if I'm in your area. I'm looking for something else. And your not-so-relevant ad might be displacing an ad for something I really do want.
|Also, do you think that Amazon adds "CD" or "Book" to the end of all its artist-name keywords? Amazon is not offering Frank Sinatra or Sarah Mclaughlin either, but their ads appear on those keywords. |
They're offering product created by those artists, generally the primary product that those artists create. It would be more relevant than what you offer to those keywords, but you're right, it wouldn't be spot on. However, they #1 probably have their own AdWords programs with different rules, and #2 are using broad match - and you might very well pick up on some of those artist name only searches if you use broad match as well.
|Can you clarify this? Are you saying that the keyword "Britney spears Karaoke" will actually trigger ads when the user inputs "Britney Spears"? |
It's possible, specially if you are limiting your targeting to a specific area. It's probably more likely with lesser known artists than the big ones. All you can do is try it, and then run the Search Query report to find out what actual keywords are being typed in. If you're running Google Analytics, you also might want to consider the filter I posted for extracting the true keywords as well - you can find it here: [webmasterworld.com...]
|The user also makes up their mind what to click based on the ad text. After inputing the search keyword the user looks either through the results or at the ads. The user can choose between different advertisers offering Britney Spears CDs, Britney Spears T-shirts, and Britney Spears Karaoke. If users think the ads are not relevant they simply wont click on them, the CTRs will be low and the advertiser will stop advertising. |
That's not how AdWords works. You can argue that it should work that way if you want, but the fact is, it doesn't.
Google wants very targeted ads (as much as an automated system can manage) and figures (rightly, in my opinion) that people interested in karaoke aren't going to be searching for it by using artists name. If they're looking for it, they'll be looking for ways to have a good time in your area, or karaoke bars, or whatever. But do you really think they're going to make a decision to visit your establishment based on a specific artists name?
I had a client who was a local bar with live entertainment. They put in this really fancy lighting that cost them tens of thousands of dollars. They wanted me to prominently advertise this lighting; I argued against it, but they were the client, so ultimately I went ahead and did it. And the campaign tanked. Because people don't come to the bar to see the lighting - they may appreciate it while they're there, but ultimately they come for the live entertainment, or because their friends are there, or because of the drink specials, or the atmosphere - but not the lighting.
Hope this helps. The reason I am taking so much time with this is because these are the same misconceptions about AdWords I see over and over, and maybe it help clear some air.
| 10:12 pm on Dec 29, 2008 (gmt 0)|
I think the big issue as Netmeg pointed out so is that Search Engine Marketers sometimes concentrate so much on relevancy that that we try to be *Too Relevant* on some campaigns and get a strong case of relevancy blindness.
Whilst I can see clearly how CrissCross truly beleives that his keyword choices are very relevant to his local ( very relevant ) audience and by appending Karake to them he will get zero traffic and so doesnt want to do that, Google and users are never going to see it the same way.
Maybe in this case with *Such* a niche market in such a local area, PPC and online marketing are not the way to drive traffic to the buisness and good old offline old fashioned street walking into bars and doing deals with the staff and owners and flyering are?
| 6:05 am on Dec 30, 2008 (gmt 0)|
I agree with you that flyering is important (we are actually doing a lot of flyering around town) but as we are a compeltely new entertainment concept for our area it would be useful to drive customers to our website where they can learn in more detail about how our business works first.
I was receiving around 100 clicks a day at around $0.40 before the bids were raised to unaffordable levels. 100 clicks a day is a good number for a relatively small local business like ours so I was very happy with the performance. To get the same number from TV or radio ads or flyering I would probably have to spend a lot more.
As you say, in trying to concentrate so much on a relevancy (the calculation methodology of which is completely unknown to advertisers and even to Google's support team) they have become blind to the needs of ordinary advertisers.
I ran a very successful advertising-supported magazine for 12 years and dealt with advertisers every day. It would be like me saying to Sri Lankan Airlines who wanted to advertise in the travel section, "No, you can only advertise when we run an article on Sri Lanka". Did we ever get a complaint from a reader about running the ad against - god help us - an article about India? Of course not. When we talked to advertisers about their ad placement it was always with the aim of giving them more readers, not less!
Forcing advertisers to narrow their keyword choices to the point where they don't get any clicks from those keywords is insane. I am quite persistent so I have tried to make this work, but I am sure there are thousands of advertisers who just give up and never complain.
Not that you can complain: When talking to Adwords support today I asked if I could talk to a superior to explain my case further? NO. Was there anyone available who had any influence on the policy?: NO. Was there any way to complain? NO. I was told that no-one had any influence over the algorithm and here was no way to change it -- it could have been a scene from the movie Brazil.
All I want to do is to capture the 3-5% or so of local users who could be interested in singing an artist's songs. I can't do that by asking them explicitly ("Britney Spears karaoke") because they don't know they are interested yet, and if they were interested they would most likely already know about my business.
However I do know, from common sense and from the CTR on ads that are running, that many people who input "Britney Spears" and see an ad for Britney Spears karaoke will be interested and click on my ad and find out more about my business. What is so wrong with this that Google has to squash it?
Despite Netmeg's belief that my business is not as relevant as someone selling a CD or ringtone (is buying an experience really that different from buying a song or a T-shirt?) I believe that about as many people would click a Britney Spears ticket or T-shirt link, as would click a Britney karaoke link, especially considering I have an unusual local business and have no competition in this local area.
Nutmeg, like Google, seems to want to define how relevant my business is to the user. His post is full of his subjective decisions about how relevant my business is. With all due respect, neither Netmeg nor Google knows anything about my business and its relationship to my local market -- that has still to be revealed. The argument that Amazon is somehow more relevant to a user, or that their ads provide a better "experience" to the user than my business is simply not proven. The local user may well not be interested in buying a CD at all and would rather come to my place to sing.
If the relevancy was based on CTR then we could clearly see how many people choose between downloading Britney's songs, buying her T-shirts or singing her songs. But Google has decided to use a different, highly volatile, and unknowable relevancy measure that effectively locks me out of advertising against these other products.
Netmeg also seems to think that it is fine that Ebay and Amazon have special deals due to their size. That's not really a level playing field and rather than accepting that it would be better if he pushed for the same rules to apply for all advertisers.
In short, rather than telling me my business is not "relevant' Google should be trying to help me make a better connection to new customers rather than hinder my connection to them.
No doubt the coming severe downturn in advertising will make Google realise that my $1000-2000 a month (and that of other advertisers in a similar situation) is important -- it's difficult to be choosy in a recession. Until then most of my budget is on Facebook and Yahoo (and flyers).
| 3:20 pm on Dec 30, 2008 (gmt 0)|
(netmeg is a she)
Ok; good luck to you.
| 6:30 pm on Dec 30, 2008 (gmt 0)|
I think Google is well aware that you can never make 100% of the searchers happy 100% of the time, so they play the odds and serve ads with the best chance of being on target. And for any search query Google can easily determine the breakout of what past searches have clicked on and were likely looking for.
So crisscross, you are correct that Google may well be leaving out a great option your business offers, but they also know that it represents a very small percentage of searchers. A very large (and I am just guessing VERY large) percentage of searchers were looking for music, or t-shirts, or something else.
So if your goal is to Catch 3-5% of local searchers using general terms, Google would be serving your irrelevant ad 95-97% of the rest of time.
As Netmeg said, good luck, in my experience Google will never let you display an ad that from their vantage has that high a liklihood of being off-target. Better to let the searcher realize they didn't find what they wanted on the first search, and the 3-5% who were looking for you will refocus and search again using terms more specific to you.
| 1:11 am on Jan 1, 2009 (gmt 0)|
|Hope this helps. The reason I am taking so much time with this is because these are the same misconceptions about AdWords I see over and over, and maybe it help clear some air. |
I want to express my thanks. Your posts here have been some of the best on this subject I've found. We've played with adwords a fair amount, but with limited roi, probably due to lack of skill.
Now, we want to get serious, and put in more money, and learn more, but it's hard to find clear explanations of the kinds of things we need to understand to make this work.
We're reworking all our landing pages to function as better sales platforms, and we'll be redoing all of our campaigns, certainly incorporating many of the things you've talked about.
| 2:25 am on Jan 2, 2009 (gmt 0)|
As suggested above by lady Netmeg, I took all the artist name keywords that had gone up to $5 minimum bid and then I made a new ad group and added all of the keywords with "karaoke" appended to them.
It's been running for a day or so now. Many of the "karaoke" keywords now have a quality score of 7, many do not. There are still 65 keywords that have quality scores of 5 or less. One has to wonder about the mysterious reasoning why "Usher Karaoke" has a Quality Score of 4 only, while "4 Non Blondes karaoke" has a Quality Score of 7.
However, the real issue is that, as expected, the artists with "karaoke" appended have *zero* clicks at all. This is because, as I explained above, no-one searches for karaoke directly unless they are already interested in it, and if they are already interested in it we don't need to advertise to them (and a broad match "karaoke" keyword will catch them anyway).
By contrast the artist name-only keywords that have a minimum bid less that $1 have produced 50 clicks in the same time. If I had access to the artist name-only keywords that are over $5 I would probably have another 100 clicks/day and Google would be getting another $1500/month. However, Google has decided, through the Quality Score that it knows better than I do how to advertise my business.
The objective of most of the world's advertising media is to place ads in positions where they will create the best connection between the customer and the advertiser. By forcing customers like me to place ads where there are no customers, Google does itself, its advertisers and its users a disservice.
| 4:23 am on Jan 2, 2009 (gmt 0)|
|The objective of most of the world's advertising media is to place ads in positions where they will create the best connection between the customer and the advertiser. |
Connections? I'm curious. How are you evaluating the return on your investment here? You aren't, are you?
I'm also curious as to where your enterprise will be in 18 months. It's possible you have a flawed plan going here.
| 8:13 am on Jan 2, 2009 (gmt 0)|
I am curious why you think my plan is flawed? I am evaluating the return on investment by the number of people who view our website, join our Facebook group and ultimately call or visit our business. The advantage of internet adverting is that I can track everyone who visits the web page, views the flash and so on. In this small city 100 - 200 visits to my site a day is a lot.
Regarding the comment about "connections" I mentioned above I ran a highly-successful advertising supported magazine and website for 12 years. My magazine dealt with hundreds of display advertisers each week plus about 1000 classified ads each week too.
Media makes money by connecting the advertiser with the customer. Usually media want to maximize the number of customers responding to an ad, not minimize it! If they make a good connection between the customer and advertiser then both are happy and the advertiser continues to advertise. Usually this is done by placing ads near to areas where the advertiser and media agree that they will get the most response. Some advertising is content dependent, such as airlines advertising in the travel section, but often it is not, for example Coke advertising at the Superbowl. This is all elementary stuff.
Perhaps Google's system is to stop large advertisers like Coke advertising in unrelated sections, but in my situation Google is stopping adverting on relevant keywords that work well for my business. Adding "karaoke" to the keyword simply does not work because nobody searches for that.
My business is a new concept and people simply aren't searching for it in that way on Google and I doubt they ever will. How can people find you if they don't know the search term to use? Or are unwilling to search at a deeper level? They are, however, interested when they see it advertised against the artist's name. In my previous life we had lots of advertisers in our business who were not perfect fits with our content sections, but we never were daft enough to turn them away.
| 5:16 pm on Jan 2, 2009 (gmt 0)|
What you want to do, what you view as *your* business plan, is just plain not Google's business plan.
Of course Google doesn't know your business better than you do, but I'm pretty sure they know the search business better than you (or I) do, and they know more about click patterns and conversion rates and all that information that they've collected over the years from AdWords, AdSense and Analytics, their own logs, Google Toolbar, etc.
If people aren't searching for your business yet in Google, then advertising in a search engine may not be the best fit for you. It's definitely not the best fit for Google, and they know it, and that's why you see the QS issues that you do.
Over the long haul, they would rather take $500/month from an advertiser whose ads and keywords are spot on relevant to their business, than $1000 to $2000/month from someone whose keywords are more general, and who is just hoping to catch the eye of a searcher who doesn't even know he wants the product yet. The $500 advertiser's ad campaigns are much more likely to sustain and grow, making it a better long term strategy.
Of course, they're not 100% there yet, but they are working on it (and believe me, it was much worse a couple years ago when you'd click on an ad and have no real idea what you might find on the other side)
AdWords text ads are *not* the same as display ad banners. They are contextual, so they have to be more relevant than a display ad (which is usually cost per thousand impressions) would be.
Google is also trying like crazy to build and improve trust in their ad product on the part of the users (searchers, clickers) because declining ad relevance serves nobody's purpose.
I don't hold out a lot of hope that any of this will help you understand how AdWords works, because you seem to have already made up your mind how YOU think it should work. Google's not going to change their business model for you. We get that you don't like it, and I'm sorry it won't work for you. But it won't. The artist names that you want to use, by themselves, are *not* relevant keywords for a karaoke establishment.
FWIW, you would likely have the same experience advertising with Yahoo & MSN, although of the three, you'd probably have the best shot of making it work with MSN. For now.
| 8:47 pm on Jan 2, 2009 (gmt 0)|
You seem to have missed my point entirely: about half of the artist names have high quality scores, but many do not. You can claim that my keywords are not relevant, but Google itself says that half of them are. It's the other half, which appear to be no different that the ones that are allowed, that I would like to be able to use. As I said above why should "Wang Chung" have a quality score of 2 and a minimum bid of $10 while "Bette Midler" has a quality score of 7 and a bid of 0.40? I would be surprised if you could tell me why, when the technical support people at Google couldn't even tell me.
As my business is an entirely new concept Google could not have any historical data about the relevance of the keywords to my new business.
Your point about the $500 advertiser growing to love Google is bogus because I am such an advertiser and half of my relevant ads are being blocked. Worse still, they are being blocked on pages that have hardly any advertising on them already.
And really, no-one is going to complain that they were served a Britney Spears karaoke ad when they search for Britney Spears. I never buy ringtones, but I don't get all offended when I see those ads on Google next to my favorite artists. I don't think OMG Google has lost it now and demand that ringtone ads appear only when I specifically search for them. It's ludicrous.
I understand quite clearly how Google wants the system to work, but it's patchy application makes it difficult and frustrating to get reasonable results. I am pretty certain that almost all advertisers have had similar issues with Quality Score, and that a large majority of them would rather bid on less targeted keywords that have lower CTR but higher click volume.
The time and effort to manage this Quality Score issue is a major turn-off to any potential advertiser. I actually appreciate your efforts to try and explain it (and followed your advice), but do you really think most business people have the time and patience to really do it the way you and Google want, especially when doing it that way goes against users search behavior, and in the end gets them less results?
The simple fact is that my case shows that the Quality Score make it harder than is necessary for any small business person to effectively advertise their business. I also think that Google's investors will be interested to know that according to your example that in this economy the company would turn down $1500 on a principle that makes no sense.
| 10:19 pm on Jan 2, 2009 (gmt 0)|
Chris, what percentage of people who click on your ads have gone to your establishment?
I actually agree with some of your more general comments. Yes, I think there's inconsistent application, but it's how it works, and it will always be that way, as is the case with many things in life.
Yes, QS adds a layer of complexity that discourages advertisers, I think, although one needn't get bogged down by it.
On the specifics of your situation, I disagree, but I understand you can't share all your info here.
You are claiming some things that are a little unlikely, and trying to market in a way that I find unlikely to work, but heck, if it works, great.
Personally, I happen to think that terms like Chicago karaoke would be relevant to an establishment in that city that has karaoke, BUT
I do NOT think that Britney Spears Karaoke Chicago is terribly relevant to anything except if a person inputs those exact words.
I think you may have made a strategic mistake with your bids, the structure of them, etc, also. Too many shotgun approach tends to lower QS, I think.
| 7:43 am on Jan 3, 2009 (gmt 0)|
Coach, we already use "city name karaoke" -- that, and city-related keywords, get us about 10 clicks/day. We have just started tracking our customers so I don't have too much data yet, but anecdotaly, we have had people coming to our business as a result of our web ads and with only flyering, web ads and a few PR pieces around town business is steadily growing.
To give some perspective, as a result of Facebook's PPC program we have 125 Facebook fans, from 800 clicks (around $400). I am certain that all of these fans will visit our place at least once and tell friends etc. If Google's ads can convert at a similar rate then we will very happy.
I have to decide in the next few weeks if it is better to spend $1000/month on Google, or the same on a radio or cable TV campaign. Each one has its advantages. The problem with Radio and TV is that it is difficult to track who has actually seen your ad, and I don't see much point in using radio or TV to drive customers to a website when you can place ads around the web anyway.
The big advantage with pay-per-click is that I know people that click on the ad are specifically interested in Karaoke, because the ad specifically mentions karaoke. Also because the artist name is the keyword I can feed them information about the artist's songs that we have on our system. I am getting about 80% click through from landing to the actual main site, which further proves that people clicking on the artist names find our content relevant.
I don't agree with the sentiment "but it's how it works, and it will always be that way". It is not good for me, Google, nor I suspect many other advertisers, for things to stay the same. It is not good to have policies that decrease earnings. Like all businesses, Google must adapt or be surpassed by those who do accept that things can work better if some rules are relaxed.
It is also a terrible indictment of Google's inflexibility that you think that nothing will change the situation. How can a company that once offered so much hope become so hopeless? How can something so entrepreneurial become so beauraucratic? If, as customers, we cannot make our voices heard, and Google cannot reach out to us, then we are all just slaves to the machine.
| 8:42 am on Jan 3, 2009 (gmt 0)|
"we are all just slaves to the machine. "
Saw "Matrix" rerun the other day.. ;-)
Netmeg gives correct advise. If it doesn't work.. the way so you want, you have really no other option.
We are really in a sense .. slave to the Big G machine. You are bang on 'target'.
[edited by: Green_Grass at 8:42 am (utc) on Jan. 3, 2009]
| 6:37 pm on Jan 3, 2009 (gmt 0)|
1) Ok. So you are NOT tracking ROI as you suggested earlier. I didn't think so.
2) Run media ads with a discount code to track their effectiveness.
3) It's been explained to you why google is doing this (or trying to do this). It IS indeed in their interests. It may actually be in your interests too. I'd suggest you don't know yet.
4) You aren't a slave to the machine, and it doesn't reflect very deep business thinking to make these comments. It's YOUR business model, and YOUR marketing model so it's up to you to make it work. Adwords has no responsibility to you other than what it says it has. If you don't like it, you need to find an alternative model that works for you.
5) Facebook isn't google. You can't draw conclusions from one to the other.
I'm trying to be helpful here, but I really feel like your conclusions about your business may be faulty. If they are, it suggests a higher probability of failure. I'd suggest you have someone who is brutally honest, AND analytic to look over your conclusions and the data they are based on. You could be absolutely right on everything, but if you are wrong on just a few, well...
Again, good luck.
(local radio and tv promotion seems best for you. Free givaways via radio, coupon discounts, etc.
| 9:59 pm on Jan 23, 2009 (gmt 0)|
Wow a good bit of defense for Google in this thread.
Rather than pile on Crisscross that he is wrong and Google is right I think he makes complete sense in every single argument he has made.
Unfortunately the way Google has decided to evaluate "relevancy" is not a good match for how he wants to run his campaign.
To me, based on what he has said about his campaigns, and if for a moment we disregard how Google's QS worked not many people would argue that it doesn't sound relevant and a good match for his target audience.
It's only when we apply the skewed reality of "relevance" based on Google's rules do things start to suddenly sound irrelevant.
And that unfortunately, is how we must play the game.
| 8:48 am on Jan 26, 2009 (gmt 0)|
Criss Cross, reading this thread my overwhelming impressions is that you have not realised that print advertising, which you have a lot of experience of, is in no way related to on-line advertising. Unless you let go of the idea that you "know how this should work because of your previous experience" you are doomed to struggle with Google,a dn other forms of on-line advertising. I humbly suggest that you stick your experience in a box, and try to learn from someone who is very experienced in this form of advertising, and you could do a lot worse then actually read, and believe what Netmeg is telling you, very few have her level of experience. (said as someone working as the online expert, in the midst of a large number of offline advertising experts)
| 10:43 pm on Jan 26, 2009 (gmt 0)|
Am I the only one that thinks he makes sense?
I've been managing Adwords campaigns full time for 5 years now so I'm not new to this game.
First and foremost my position and perspective is that of what might be perceived as actual relevance to the searcher...not Google dictated algorithm relevance...
(yes, I know the algorithm relevance is how we must play the game but that does not equate to actual relevance...and yes the argument could be made that Google's relevance is actual relevance in order to play the game...yadda....yadda....yadda....)
So that said, the gist of what I am hearing is that if I am a Britney Spears fan, it is not relevant and is foolish to think that if I search on her name I would not be the least bit interested in doing a little Britney karaoke?
Or that I might want a Britney Spears ringtone, or t-shirt, or lunch box, or super awesome Britney action figure?
None of that is relevant to the phrase "Britney Spears"?
My point is that yes we are slaves to Google's algo, and yes we are forced to play by their rules...but to say CrissCross is silly for thinking his campaign is actually relevant (as in real human beings and not a computer would find the stuff relevant) is just as foolish as Google's interpretation of relevance in many circumstances.
| 5:31 am on Jan 28, 2009 (gmt 0)|
crisscross, I agree with you, the very idea and suggestions that your business plan is flawed is a Google Myth.
Their system is definitely designed to extort as much money as possible. If you contact them you will only get an automated e-mail back.
Besides, who are they to tell us how we must run our business and how we should advertise? Yes, I know they're the big G, but one day they'll be slapped with an anti-trust lawsuit worse than Microsoft had feared.
His keywords are clearly relevant, and I've seen many cases like crisscross's.
And Google still shows a lot of irrelevant ads when I search for any number of things. Big name advertisers can come up for all kinds of things even though it's completely unrelated.
| 9:31 pm on Feb 3, 2009 (gmt 0)|
Very interesting thread. I feel I have to butt in.
Netmeg is exactly right in what she says. We should all learn from her. And I know what I'm talking about having managed dozens of campaigns myself, large and small.
I've seen campaigns built just like crisscross is doing and NONE has ever been successful.
Another thing that was overlooked is that the landing page affects the QS. So even if you bid on "Britney Spears karaoke", while karaoke is the underlying theme, maybe Britney Spears is not. So crisscross should have a landing page for each of his keywords to make it more relevant.
> if I am a Britney Spears fan, it is not relevant and is foolish
> to think that if I search on her name I would not be the least bit
> interested in doing a little Britney karaoke?
> None of that is relevant to the phrase "Britney Spears"?
That's right. "Britney Spears" by itself is not *highly* relevant. I may want to read news about her or read her bio. If I typed on the other hand "Britney Spears CD", that is more relevant and that is the keyword I should be bidding on if that's what I'm selling.
> His keywords are clearly relevant, and I've seen many cases like crisscross's.
I've seen many cases myself like his. And they were NOT relevant either.
I agree with Coach about using radio in this case. Use Adwords for "karaoke in city" type keywords. Less impressions but highly relevant, should be high CTR, low cost. He could try text and banners ads in selected sites that people in his city visit regularly.
| 3:27 am on Feb 5, 2009 (gmt 0)|
|That's right. "Britney Spears" by itself is not *highly* relevant. I may want to read news about her or read her bio. If I typed on the other hand "Britney Spears CD", that is more relevant and that is the keyword I should be bidding on if that's what I'm selling. |
"highly" is the key term here, if "highly" relevant were required for every single search term then it would be next to impossible for any advertiser big or small in any industry to bid on any broad/generic keyword...which we all know is not the reality of things.
And so if someone were top search on "Britney Spears" and saw an ad that had Britney Spears CD's on sale, that would not be relevant or of interest to the searcher?
Google has their rules, nothing we can do about it, that is how we must think and unfortunately "see" relevance as search marketers but it doesn't mean that things are actually relevant.
Google is just the best at getting things wrong less often than anyone else.
| 9:44 am on Feb 6, 2009 (gmt 0)|
I'm sorry JKwilson, but in this case the highly relevant broad match term is Karaoke. Karaoke in itself is not relevant to Britney dspeaqrs of any other singer.
It's all about probability and intent. If you search for "Britney spears" there are relatively few likley intentions from that act,
1) you like her music
2) you want to read about her latest issues/boyf/other gossip mag info
3) you like how she look sand want pictures of her
so if you are selling CD's of her latest album you are highly relevant to people with intention 1. If you are selling calandars you are highly relevant to intention 3, if you are a gossip site you are highly relevant to intention 2.
However Karaoke is not relevant to any of the above, bidding on this keyword is only relevant for the advertiser, in the form of branding, it is not relevant to the searcher whose intent is almost never going to be to find a Britney spears karaoke bar.
So from G's perspective who do they place on the page, an ad which is likely to be clicked on because it fulfils a searchers requirement, or one that is highly unlikely to be clicked because it is not relevant.
Google is not a charity, and i find it amazing that in the adsense forum that are hoards of people complaining because Google won't let them ban irrelevant, non money making ads from their sites, yet here you are actually advocating that Google lets a less profitable ad remain, because of the highly tenuous link that karaoke involves singing, and Britney spears sings songs.
| 11:55 am on Feb 6, 2009 (gmt 0)|
what exactly did you mean earlier, when you were talking about Amazon saying that they "#1 probably have their own AdWords programs with different rules"? Them, ebay, other big companies or pharmaceutical firms, you say, actually Have a diff. policy? - What does that mean for lower profile AdWords-customers?
| 6:02 pm on Feb 6, 2009 (gmt 0)|
Nothing different than it has been. I don't know that the large companies have different terms, but it would stand to reason. They do in AdSense. I don't worry about it.
| 6:07 pm on Feb 22, 2009 (gmt 0)|
Nutmeg's guess is correct. There are different rules for the large advertisers. Google gives them preference.
Amazon is a special case. Google tries to suppress people who use Amazon store (the affiliates). If you use Yahoo Store as the shopping cart/transaction engine for your ecommerce site, that's okay. But if you use Amazon for the same thing, your site will move down to page 5. Why? Amazon and Yahoo stores are competitors, but Yahoo owns 20% of Google. So, Google helps Yahoo and suppresses Amazon.
| 11:38 am on Feb 25, 2009 (gmt 0)|
I think Netmeg did a very good job explaining things here.
I know the OP is frustrated, but realistically, what percentage of people who search for Britney Spears are going to be interested in your product?
I am guessing that the actually percentage of people making that search are not interested in a karaoke song version. Sure you might hit some of them with a curiosity factor when exposed to it, leading to lower conversions anyways.
If only 1% of the audience might even care about your ad, it is hard to claim it is relevant. While you don't see the traffic with the karaoke word attached, those are highly relevant, so your ctr will be significantly higher as will your conversions. The key is to find what other words people use when searching for karaoke.
I simply think you are over estimating your market here and think you should be getting a lot more traffic than you are, but in reality it is a very tiny market. You are trying to force action and ultimately conversions by casting a wider net, and while this can be viable when you can get extremely cheap clicks, it is hard to do on adwords due to relevancy.
I just don't think the average person who is actively searching for a karoake version of a popular artist will just type the artist's name in. It is not an intuitive way to use a search engine like that. So those searches are more impulse views, and like I said your conversion rate will be many times lower even if you could get the traffic.
I would focus my efforts primarily on trying to make sure your ad is showing every time someone in Detroit wants information about karaoke, however that may be. If that doesn't amount to much, well it is what it is.