Broad match advertisers, how deep are you tracking?

[Shak]

Can all the advertisers who use Broad Match raise their hands please :)

aha, quite a few.

OK, how many of you are TRACKING which exact "More Specific Keywords" and "Expanded Broad Matches" are working for you?

example:

You are bidding broad match on books

do you know what terms are delivering you traffic?

such as free audio books? and even more importantly did you know you are getting traffic from words such as literature

lets get some answers and then I will share why I have raised this topic :)

Shak

[cline]

Yes, I do, at least in a general sort of way.

I always read the log files to look for undesirable words included in the search keyphrases. For example, if I'm selling widgets and someone clicks on "widget clip art" I'll add "-clip art" to the targeting. No matter how thoroughly one researches keyphrases looking for terms to block, there are always more to be blocked.

I also look for large volume sub-terms and consider whether I should manage targeting and bidding on them separately. For example, if in targeting "widgets" 25% of the volume is on "red widgets" I might add "red widgets" to the targeting to manage separately.

[skibum]

Extended broad match is virtually never used here but with all campaigns we catch the detail on actual keyword searched for all campaigns.

Extended broad match is good for short term research in this way and long term ROI when ya pull back and put the money terms in phrase or exact.

[anallawalla]

and put the money terms in phrase or exact.

I am ploughing through a massive amount of historic data from a Custom report for every keyword we used since the year dot. Small client company, CEO loves keywords and tells us what to add, sometimes checks stats and tells us what to change. No drama there. But several in-house employees have come and gone, as has one outsourced firm. Each has played with keyword variations. I was called in to analyse and revamp the whole campaign.

Some surprising revelations include terms that are absolute money wasters as plurals when singulars are money makers; some work better as broad match; some as phrase, some as exact. Some were run during periods of content targeting; some not. I cannot find a way to display date ranges for these results, so I can't generalise that phrase and exact give more conversions. They give fewer impressions and clicks, so absolute numbers of conversions are also lower.

Shak, are you referring to the effect of stemming?

[eWhisper]

I use broadmatch and track the associated search keywords. The keywords used to find some of our ads blows my mind on occasion.

It's the best way to find more negative keywords to further optimize broadmatch terms and get rid of some of the more liberal matching options.