I'm not including keywords that get zero impressions... but do you think 1 keyword to each unique adgroup is the way to get the best Quality Score and make it easier to optimise? I have some Ad Groups and I've been doing some split testing, but even if Ad A outperforms Ad B .. it doesn't always mean that A is the best for every single keyword in that Ad Group - it could be the best for 9 out of 10 keywords - and Ad B is the best for the other keyword. So I can see why people would want to have just 1 keyword in each Ad Group. What's your thoughts on this? Would you have both widgets & [widgets] in the same Widgets Ad Group or have 2 Ad Groups for each keyword (3 Ad Groups if you go for phrase match as well). Discuss ...
i dont see the point of 2 adgroups for the exact same word just because they are in brackets. i can understand making different ads for each keyword so that each keyword has its own words in it, but not for the same words in brackets...
that depends on your keywords. if you have long tail keywords that are related to each other, or one word of 2 words keyword is a match for few, then you can group them togather.
It makes sense with small campaigns (less than 100 keywords) since you can easily track the performance on keyword level. Of cause, all match types in one ad group. Another thing is software you're using to save time. Some people do it manually, some with Adwords Editor, other with third party software.
Yes I do agree that all 3 match types should go in the same Ad Group, as the Ads would relate to all 3 of them. I'm thinking Quality Score as well. It seems like it would be much easier tracking this, when a keyword's QS goes up or down, as you can check each Ad Group separately - if it goes down, you know to change the Ad. You couldn't do this if you had lots of keywords in the same Ad Group, as the original Ad might be good for some keywords but bad for the rest. This way, each Ad is for each keyword.
I've found that the ad you would use for an exact match keyword is not the same ad you would use for that same keyword broad matched....assuming of course you wanted maximize the effectiveness of the ad. Just something to keep in mind. The one keyword per ad group way of doing things is not applicable to every situation. Over time and with experience you'll figure out when it's the way to go and when its not.
One keyword per adgroup can be a good way to go if you're mainly running on the content network. With Google's own keyword tracking (as in when you use {keyword} added to the subid on your CPA/Sale offer) they randomly pick a keyword from your adgroup to assign as the keyword for tracking purposes when it comes to the content network. With only one keyword per adgroup you will know exactly which keywords are bringing you conversions on content. Jill
It all depends on whether you intend to write a different advert for each Adgroup. My main argument against it is that it takes far longer to get significant results on advert tests. I would do it for keywords with large traffic volumes, but there's no point for long-tail keywords - you'll never tell which advert is the most effective, which was the whole point in the first place. The only time that I'd do it is for keywords that you want to show different adverts for, which have large search volumes. It only affects the QS indirectly, through the advert CTR and relevance...
Hi CM, thanks for your response. With regards to Ad Testing, yes, an Ad Group that gets lots of impressions will be better for finding out which ad is more effective, but what if keywords 1,2 & 3 ... do very well with Ad A, and keywords 4 & 5 do better with Ad B? That's the thing with split testing, surely, removing Ad B and keeping Ad A, because *overall* it performed better, could be effecting some keywords negatively? What would you do in this scenario? Would you put the keywords that did well with Ad B into another targetted Ad Group and start it off with that Ad, whilst making a new one to compete against it? Or would you keep them all in the same ad group, and just pause/delete the advert that 'lost', regardless?
If you believe that different adverts will, or even might, work better with different keywords, then you should put those keywords into different Adgroups...
I always toss my top 5-8 keywords in their own ad groups so I can target the ad text as much as possible. Past that, I do standard groupings of like-keywords into ad groups. It's 100% unnecessary to to an ad group per keyword on low volume keywords though, it'll waste your time & you'll get no benefit from it.
"It's 100% unnecessary to to an ad group per keyword on low volume keywords though, it'll waste your time & you'll get no benefit from it" I do tend to agree - what I've done is for keywords that get very few impressions but are related, I've put into a single Ad Group as they require very little maintenance. Some aren't even getting any impressions so I'm deleting them. As time goes on, I'm splitting high impression keywords into their own Ad Group, depending on how they get on with each Ad.
That's absolutely not true. You can better target ads to a keyword and maximize traffic potential on low volume, high converting keywords. For example let's say you have a keyword that gets 100 impressions, 9 clicks and 6 conversions per month that's part of an ad group with 10 keywords. You could theoretically split it into it's own ad group, increase it's CTR via ad testing (which would improve the QS and lower CPC) and capture a few additional conversions per month if your conversion rates hold. The whole point of one keyword per ad group is to maximize your effectiveness by writing ads specific to a keyword vs writing an ad that works for a group of related keywords. If you're not going to write specific ads for a specific keyword just group everything together and use DKI.
Breaking out your keywords into individual adgroups and even separating match types is known as one to one and is a very common and effective practice. example: widget cheap widet these could prove to perform completely different and best with their own ad tailored to each variance. Ad 1 - widget Headline: Widget $99.99 ad 2 - cheap widget Headline: Cheap Widget Sale Keyword "widet" might be a high volume keyword and be most effective with the price listed to prequalify the visitor and keep ROI high and expenses low. If keyword "cheap widet" was used in the 1st ad, the headline would not be in line with what the searcher had in mind, since price is displayed up front and no longer can be perceived as cheap. Keyword "cheap widet" will most likely have lower search volume therefore capturing more searchers with a more vague ad would be crucial for volume. Now separating match types is also very important. If you do not separate the keywords I would still HIGHLY recommend separating match types, since they are going to perform extremely different. Example widget - broad widget - exact ad1 - widget broad Headline: {KeyWord: Widget For $99.99} Ad2 - widget exact Headline: Widget $99.99 Free S&H For "widget broad" we never know what the user is going to search for so using keyword insert is very important when broad matching. I always want to try and match what the user is searching for so I can get my bold headline but since broad has many variations it can match to, I prefer auto insertion in the headline then pre qualifying in the D1 and D2 fields For "widget Exact" search volume is going to be much less, therefor you want to make a highly targeted and outstanding ad that will get as many clicks as possible. Ad 2 headline will get a bold plus display additional value. Lastely, there are many QS performance bonuses you can gain when breaking keywords out 1 to 1, which I will not go into for post length reasons So, if you have a small portfolio I say, 1 to 1 everything so everything has the most highest relevancy possible and its a great way for noobs to eliminate mistakes with improper keyword groupings.
I'd agree with you to a point. The only observation that I'd make is that if you have little experience in writing PPC adverts, then setting low-volume keywords up in their own Adgroups makes advert testing virtually impossible. So you could be stuck with a poorly-written advert forever. If you know what sort of adverts work, then the benefit of a more-targeted advert probably outweighs the disadvantage of not being able to test. If you don't, I'm not sure that that's still the case...
I have Ad Groups related to a particular subject which I've only just produced. I haven't done enough testing yet to determined which ads are proven to be better than others, on this subject, so putting a single keyword into it's own Ad Group based on each Match Type .. well .. as pointed out, I could be waiting weeks or months before I get enough data to see which is the better of the 2. If they both happen to be poor - the keywords won't get any clicks. I've set a few Ad Groups up that are getting a 'trickle' of clicks, but have all 3Match Types as part of the Ad Group. The 2 ads are related to all 3 match types (I think anyway), but if the broad match gets lots of impressions but not clicks (which would have enabled me to add negatives), then I'll consider moving it to it's own Ad Group and trying more inviting ads. Either that or just deleting the Broad Match.