Relative value of phrase match & broad match?

[Tonearm]

On average, what do you think a phrase match and broad match are worth relative to an exact match? I'm thinking 90% for phrase and 25% for broad.

[D_Blackwell]

There are others that can peg the numbers better than I can, but I wouldn't argue your thoughts. Broad match can paint a very wide swath. A lot of people with zero interest are going to see the ad - but they still get a chance to burn a click.

Don't underestimate the value of negative match.

[ihavekeywords]

Broad match takes a lot of maintenance to keep going, but I look at it many times as a good way to be finding keywords I should be be bidding for individually.

My feeling is that broad is worth more than 25%, but that's with pretty extensive negative keyword lists.

[Tonearm]

That's the thing. I have thousands of keywords and I can't think of a way to make negative keywords scalable to that level without too many false triggers from them. I don't use negative keywords at all because of this.

[buckworks]

If you're using phrase or broad match, it's vital to have at least some basic negative keywords in place.

Example: if you were selling pearl jewelry, it would be foolish to bid for broad match on "pearl" without some negatives, because many searches that include the word "pearl" are about something completely different than what you're selling.

Pearl Harbor, Pearl Bailey, Pearl Jam, Pearl drums, Pearl S. Buck, and so on... there's no point showing your ad to the wrong eyeballs just because there's some overlap in vocabulary.

A few simple negatives could save huge numbers of wasted impressions:

-harbor
-jam
-drums
-bailey
-buck

Invest some thought to identify searches where your ads ought not to appear, and enter some select negatives to block them.

Pearl of wisdom: Reducing off-target impressions is one of the easiest ways to improve your clickthrough percentages.

[Tonearm]

Thanks buckworks. I gravitate more toward a strategy like this:

adgroup1:
[pearl]

adgroup2:
"pearl"

adgroup3:
pearl

adgroup3 will have dismal conversions and the max cpc will drop accordingly. adgroup1 is straightforward and adgroup2 will be somewhere in-between. The advantage of this is absolutely no time spent on negative keywords.

As for all the juicy multi-word keywords you miss out on in adgroup3 because your max cpc is so low, they can be added like this:

adgroup4:
[pearl jewelry]

adgroup5:
"pearl jewelry"

adgroup6:
pearl jewelry

[buckworks]

The advantage of this is absolutely no time spent on negative keywords.

How much time are you thinking here?

You wouldn't have to change a thing about how you've structured your ad groups.

If you spent even ten minutes with a good keyword research tool you'd have no trouble spotting some candidates for blocking, and it wouldn't take long to add them as negative keywords at the campaign level. I guarantee that those few minutes would have some of the best ROI of anything you did all year.

You could do a lot more than that, of course ... some people have hundreds or even thousands of negative keywords ... but identifying even the top half-dozen negative keywords would improve the profitability of a campaign if broad matching is involved.

[Tonearm]

The whole negative keywords thing has always felt like a case of "too many moving parts" to me. What if you add -harbor and then a hot new line of jewelry called the Harbor Collection comes out? Everyone loves to talk about how much work Adwords is, and it seems like negative keyword management is always cited as the main reason.

Can you tell me why the negative keyword approach is better than the one I've described? If you isolate broad matches in their own ad groups and add new keywords in new ad groups based on the good stuff broad match finds, you've avoided paying for the bad matches without the risk of negative keyword false positives and without the negative keyword management time.

Just like a broad match keyword can bring in a bunch of crap, a negative match keyword can keep relevant stuff out, especially if the product line is evolving.

[buckworks]

Can you tell me why the negative keyword approach is better than the one I've described?

Using negative keywords is not something different from what you've described, it's a technique to fine-tune what you've described, to further tighten your targeting.

you've avoided paying for the bad matches

No, if you have bad matches, they cost you, period. If you have ads appearing in off-target searches simply because of vocabulary overlap, you lose in two ways:

(1) Your CTR will be lower, and thus your Quality Score. That results in less visibility for the same ad spend.
(2) Clicks that you do get from off-target impressions will convert poorly.

a negative match keyword can keep relevant stuff out

Only if you've chosen your negatives badly.