How diverse are your keywords

[AussieWebmaster]

We use a very aggressive approach to AdWords and the selection and development of keywords.
Apart from cleaning out the terms that have not provided CPA - spent clicks but did not convert, we are constantly testing new terms, getting creatives changed to reflect the changes in the market, etc. etc.

But what we have been recently trying are keywords that test various demographics - takes a little work with landing pages to get by editorial (unless you go for the drive by - submit take the autotraffic and deal with the denial that editorial will soon send).
I am interested in knowing how much time people are spending on honing the keywords and finding new terms.
Obviously most people here are tracking and AB testing creatives etc. but heavy testing of fringe terms - hey even misspellings - are worth the reward of finding that hot term that converts like crazy.

[Dave Hawley]

We have about 20 campaigns and they took me about 8 months to get to the point of best return. For us, we have found exhausting every single keyword/term and then mixing up the order of the terms works very well. So much in fact that 95% of the terms we use only cost the min. It took nearly the whole 8 months to get to the point where we were not getting terms disabled as they dropped below the CTR though.

For us, it's has been very much a trial-and-error process.

[AussieWebmaster]

It definitely takes a committment. But the returns are well worth it.

[Opie1Canopie]

Wow AW - I would say compared to what you are doing, we rarely refine search terms. Maybe once in a quarter or 6 months we'll take a look at each of our campaigns and do some type of keyword restructuring. Unfortunately this is because of resource constraints - once a keyword list is successful, we don't have time to mess with it too much. More frequent reviews/updates and better list targeting by demographics is something I'm trying to work towards however.

Opie1

[mcanerin]

<bad fake eastern european accent>

Regarding the title of the thread, I vant to point out that the less testing you do, diverse your keywords (and ROI) vill be...

<bad fake eastern european accent>

Ian :p

[AussieWebmaster]

If you look at your terms and see what has converted for an acceptable CPA you can also determine the ones that have spent money for no return... these we get rid of weekly if the spends are high... for the lower spenders we give them 3 months and then delete the ones that were too expensive.

While this is going on we will separate what we categorize as Valuable terms... words that are great CPA and usually have decent impressions available.

These are given individual campaigns/ad group and we then start testing creatives (varying the tracking to reflect the particular creative) and allowing the Google optimizer to push the better ones at first.
Since some creatives may have a higher CTR but a lower CPA we analyze those numbers to further tighten the proceedings.

At Overture we test the position results - varying positions for the same creative and then doing the same for another creative - allowing a little more time to get a solid set to draw analysis from.


Hope these few tips give some marketers a few ideas.

We can also do duplicates of the creatives and vary the landing pages again with the proper tracking to reflect which has the best conversion.

[AussieWebmaster]

<bad fake eastern european accent>

Regarding the title of the thread, I vant to point out that the less testing you do, diverse your keywords (and ROI) vill be...

<bad fake eastern european accent>

Ian :p
And I work in the European and Asian markets... will have to choose my words better in the future!!!!!

[PPCPro]

The key to a truly optimized adwords campaign lies in the ability to generate,classify,select,filter and group the keywords into optimal sets, and constantly measure the performance against established objectives and benchmarks.

Though we have a well established system for performing the various sub-stages in the keyword phase that help us to reduce trial and error, still it takes 70% of our project time. However, results have repeatedly shown that this is time well spent.

I am intrigued by the possible combinations that can affect an adwords campaign. Clearly, there seems to be no way out except for devoting substantial person-hours for tinkering with the possibilities.

Cheers