Forum Moderators: martinibuster
I can't prove much regarding this, but it seems like just basic common sense some may miss. Its simply good for the advertiser, as any will be willing to pay more for visitors who are more qualified than others. Any other ideas or thoughts?
Dare I say, both you and Anne may benefit from setting up an AdWords account of your own. I'm not saying so you can be a MFA hated person, but I do think that if you are having problems with Adsense perhaps one of the first/best things to do is get a better grasp of AdWords and understanding things from an advertisers perspective. Setup a campaign (its only $5) and experiment with AdWords directly yourself. Understanding things from the perspective of the advertiser is one of the most important factors when it comes to improving your individual situation. Afterall, they are your customers in cooperation with Google.
I sure hope this goes a long way towards helping...it has gotten really bad since Dec. but the visitors are about the same only with a slight increase. Clicks have gone down to half or worse and payout per click is in the....basement.
Keeping my fingers crossed!
Ann
Dare I say, both you and Anne may benefit from setting up an AdWords account of your own
kaz, I'm a long-time user of Adwords and a lot further down this road that you appear to give me credit for. I already 'grasp the role' of negative keywords and I also understand the value of Adwords to Adsense publishers - eg: if you look in the library you'll find a post of mine on smartpricing that details, step-by-step, how to use Adwords to fine-target those pages that convert poorly for Adsense publishers.
What I'm asking you for, if you are able to provide it, is not a statement that I'll develop a feel for it by experience, that's not actually an answer, but explication of how I use my own logs to infer the negative keywords used by my advertisers. That is, what do I do with my list of 3,000 words (and 7,000 phrases) to identify negative keywords used by advertisers?
It seems to me that this is not straightforward, and perhaps not even possible, which can be illustrated using the word "free". Whilst this word appears in my logs (ie: there are some surfers who find my site using that word) it only appears in a very tiny proportion of search phrases, and even those appear to be appropriate to my site. Eg: one phrase might be "free information on blue widgets", which is exactly what I provide at my site. By looking at my logs I would not realise that 'free' would be a negative keyword - rather, it is a very appropriate keyword for my site.
It really is a process of trial and error.
by referencing their logfiles and tracking which words convert and which are dead ends.
Are you tracking any conversions?
By looking at my logs I would not realise that 'free' would be a negative keyword - rather, it is a very appropriate keyword for my site.
If you can't infer that free is a term advertisers would avoid, i'm not sure how to overcome this. It just seems self evident to me. Other words also are to me. Does someone looking for blue widget pictures sound like somethin you pay for often with all of your experiences as an adwords advertiser?
Are you tracking any conversions?
Only an advertiser can track conversions, not a publisher (although my library thread explains how to infer whether ads on a page convert well or not, that is an approximation and I know of no technique a publisher can use to track conversions). Conversion-tracking is available to advertisers only - isn't it?
If you can't infer that free is a term advertisers would avoid, i'm not sure how to overcome this.
Again, please don't underestimate (or patronise) me, the situation is a lot more complicated than this. My small brain is perfectly capable of seeing intuitively that "free" may be a word that big payers in the industry might want to avoid. But my small brain also recognises that what seems intuitive can sometimes prove to be incorrect - it has to test things out.
In fact, I've just started to compile some statistics on this and found that "free" is a word that is used by advertisers in one of the industries covered by my site. For example, I've searched the website of a major advertiser in the blue widget industry for the word "free" and the resultant search yielded nearly 20,000 pages. And I've done a search on blue widgets in Google and selected the first entry in SERPS; on the landing page are three adverts, two of which contain the word "free". So, whilst the word 'free' may seem an obvious choice for a negative keyword, there is more to it than that.
Perhaps we are talking at cross-purposes. When you use the word "infer" you mean to guess. However, when I use the word "infer" I mean to "reason from evidence/circumstance". What you seem to be suggesting is that I look at my 3,000 words and intuitively guess which ones might be negative keywords for my advertisers. What I was looking for was a methodology based on evidence, not guesswork.
When you use the word "infer" you mean to guess. However, when I use the word "infer" I mean to "reason from evidence/circumstance".
lol... i am saying to reference conversion tracking for evidence. Hardly guess work to me.
Conversion tracking is available for any site. There are tons of programs out there that can assist with tracking conversions on your site. Does your site convert anything besides adsense ads?
lol... i am saying to reference conversion tracking for evidence. Hardly guess work to me.
Kaz, in the context of this and the Adwords forums "conversion tracking" means tracking how well clicks on Google ads convert (for Google's advertisers). That information is not available for a publisher.
If, however, by "conversion tracking" you are referring to tracking whether my visitors are clicking on Google ads, then I have that information aplenty. What I don't see is how that information can be related to my logs to work out negative keywords for my advertisers. If you think it can be done, then please explain.
If you are suggesting that I track (non-Adsense) conversions on my site to extrapolote what is happening on advertisers' sites with Google, then because of the diversity of my site, such a comparison with niche advertisers is unrealistic.
To cut a long story short, could you please just explain how you determine negative keywords from your logs.
It really is a process of trial and error.
Your explanations themselves seem to prove this.
I too have never seen any software that will tell a publisher this.
Ann
i don't think so, but a great suggestion for google to improve smartpricing perhaps. If publishers could actually have some guidance on how their advertising converted, wouldn't all benefit? Interesting idea.
Sure, third time
I have asked you not to patronise, yet you seem to persist in patronising without actually answering my question about how a publisher identifies an advertiser's negative keywords. You probably know how an advertiser chooses the negative keywords that are best for them as an advertiser, but you are talking to a forum of publishers and you haven't answered the question how I as a publisher can identify the negative keywords of my advertisers. I don't think we are going to get any further, so it's probably best to leave the discussion there.
Incidentally, if you want 'guidance' on how your adverts convert, then see my thread in the library [webmasterworld.com].
Incidentally, i think you are a bit patronising. notice, I answered your question time and again the best i could. Also note, you ignored my questions to you and chose to ignore my communication. I'm sure each can read into as they will.
It probably isn't a smartpricing issue, but simply negative keywords reducing the ad stock and competition on those pages.
You are right but if the negative key word (s) is extensive causing a lot of pages to get poorer paying ads I think this could eventually affect smartpricing.
I'm not in a position where I can do a lot of trial and error with AdWords. All I really want to know is what the common words are that advertisers block.
Free makes sense as does anything about clip art. But are there any other big ones I should be watching for?
Thanks much.
Death
Dying
Guns
Bullets
Pistol
Rifle (in one case air rifle:))
Drugs
and now Free....
Anyone else want to add to this list?
Maybe we can make our on right here.
Ann
Ann
It's good to have a list of stop words too but what I'm curious about is the words that a lot of adwords folks are blocking.
directory
links
info
information
or phrases such as:
widget directory
widget info
free info
Just got tired of waiting for the ads to change and deep sixed about 25 of them. Maybe it will help, maybe it won't but I will have tried.
Ann
On this changing out the free from my site and not seeing ANY change in the ads, I Filtered quite a few, BUT Here is what I happened to notice while I was working at it:
There are 2 persistent ads I would love to get rid of and have had them blocked for weeks. It seems Google overrides the filter and never stops them. I tried again to filter them when I got the duplicate warning...What gives?
Anyone have any thoughts on this? I thought the filter would be honored and one could count on the ads being gone.
BTW, I am seeing new ads showing up from my earlier efforts and they look pretty decent.
Ann
There are 2 persistent ads I would love to get rid of and have had them blocked for weeks. It seems Google overrides the filter and never stops them. I tried again to filter them when I got the duplicate warning...What gives?
Is the display URL different to the destination URL? (To check, use the Adsense preview tool, check the box by the ad you want to block and click on "show selected URLs".)
On the topic of negative keywords, I've worked out a technique to identify what words advertisers may be using to block, by comparing the words that appear on high and low EPC pages. It takes about 15 - 20 minutes to compare them, and depends on you having either URL channels or customer channels on pages (to identify the relative EPCs).
I've found, for example, the name of a licensed product in the low-EPC pages but not in the high EPC ones. This makes sense - some advertisers would likely block it because, if they aren't licensed in that product, they can't provide associated products/services.
I'll write up the technique (it's a little involved) and post it in a separate thread.
I am going to check for link, directory, etc words on my other pages though.
