Can we improve relevancy? (if desired)

This is to simply open a discussion and harnes some ideas and preferably usable models to improve relevancy - if so desired by the community.

The coop ad netwok is certainly a great tool and has proven it's value.

Personally I do not subscribe to the theory "a link is a link is a link".
Yes presently it does work for G and to a smaller extend on others.
I do happily use it and more sites controlled by me will enter.

Nevertheles I do think that the present link frenzy has started to distort result relevancy.

Therfore I'm asking if it would be of interest to the community to make the coopad more relevant.

Here is where I come from:
ink cartridge link on a travel site or on an insurance site is not as relevant as the same link on a technical or shopping site.

Given a big enough pool a potential aproach could be to broadly "classify" sites and ads (by owner) maybe along the lines of DMOZ entry structure (Art, biz, games, kids,.....)

Those of us who show more than one ad could select to show relevant ads on the additional ads if so desired (all being the default).

Instead of 1 request per hour there could be 2 requests every 2 hours one general one classified or simply alternative requests.

Yes it would double proliferation but allow more relevancy if so desired whilst keeping everyone served.

Don't shoot the messanger this is just a written thought.
M

 

This came up not long ago and I believe the general consensus was that it would be nice but the network wasn't large enough. I have mixed feelings about the relevency issue. Of course it would be nice to have all relevent links but that's not the easiest thing to do sometimes. My theory is to get as many links as possible. Relevent as well as non relevent. As long as you have a good mix I don't think non-relevent links will hurt one way or the other. I look at the ad network links as an enhancement to my theemed relevent links. Relevent or non-relevent all I can say is they've certainly helped with my SERPS in Google. Of course that could change in the future?

 

If you restricted your ads to just showing on themed links in the COOP, you'd be reducing the chance of your link ever getting indexed by Googlebot, wouldn't you?

 

1. I do not concentrate on G but monitor. I see mixed results like competition links reduced from an obcene amount to something realistic. Off-theme links showing but this being patchy....

2. I'm looking at a middle way where both camps (a link is a link / a link needs to be broadly themed) gain whatever they after or how they see the SE world developing.
Basically still featuring all to all links but also have destinguished links peer to peer in a less random fashion.

3. Yes it may well restrict my ad's showing, but again and this is my personal / subjective view ink cartridge site linking to insurance? or viagra to an obituary site?....
M

 

I think that's a logical assumption. As it is now it's kind of hit and miss. Your ad has to be on a page at the right time when Google spiders and caches it. It's probably impossible to do but what would be interesting would be to find a way to corelate ad impressions to links picked up by Google. I bet the ratio would be extremly low. Not complaining mind you. It's a great network and Shawn has done a great job and service setting it up. Just seems it would be even nicer if ads stayed on a page a little longer so they'd have a better chance of being picked up.

 

The problem is that different sites have differing Googlebot frequencies, so one site might be indexed twice per day, while another site might only be indexed once per month.

 

Googlebot hits me pretty frequently, At least I think it does. 231 times so far this month and 606 times last month. I'm sure other sites get hit much more then this. It re-caches the index page almost every day. So the ads appearing on the main page probably get picked up pretty good. I doubt that ads on the internal pages get picked up as often since Google only goes 2 or 3 pages deep on most visits.

 

fortunately or unfortunately on bigger domains I see feeding frenzies allover during the day and there is a world outside G.

G Y M AJ W S all come in and slurp by the multitude. 20+ G Y or M bots hogging bandwidth at the same time is now rather usual, if my DB's like it or not.

Thus there is a potential to inject all 100 ad's (sorry links) during their slurping and if it happens to receive a new batch during this time even more which I think is fair game.

The odds are improving with growth and the content changed flag is no longer a flat lie.....
M
PS Still some relevancy could even improve this when showing multiple ads....

 

I'll have to think about this... I don't think it makes any difference to the statistical distribution of Google picking up your link... If your ad shows up for 1 in 100 ad displays, it's exposure will be the same whether it shows on one site for 7 days or 7 different sites for 1 day each.

That said, I don't know a lot about the selection algorithm, I'm just generalising from what I've picked from the way Shawn talks about it.

 

sorry for regurgitating/duplicating ...

I was not intending to go down the route of infinite categorisation just some broader cats, thus keeping in line with linking to "related".

Also I'd concider if I want my ad on all sites than I have to show ad's from all,
if I only want my ad on lets say travel/recreation I may restrict my sites showing links from travel/recreation.
Thus no need to split anything and it's still be selfregulating.

But ..if it aint broke don't fix it....
M

 

And, what about displaying ads from any category on your sites, but being able to select that your ads are shown only on selected categorized sites, but at a higher frequency rate?

 

obviously yes you are right, although maybe I should not have used sites.

it's actually ad instance. E.G. showing two ads on a page ad1 slot shows all/general ad2 slot shows kids.... thus one can queue two types of ads one shown general on shown kids.....
...OK lets forget it for now maybe sometime later (bigger)...
M

 

The exposure would be the same but the chance of Google actually spidering the page while your link was displayed should, I think, increase? I'm not an expert in probability and statistics. All we really need to do is get Google to give us their bot schedule a day ahead of time so we know where and where to place links.