Your Title Here <data:blog.pageTitle/>

Wednesday, August 27, 2008

How Google Suggest Changes Paid Search

On August 25th 2008 Google announced the death of long tail in paid search. Of course they didn't come right out an say it but I doubt anyone reading here is surprised about that. Google called the post, "At a loss for words?"...which when I first the news was an accurate description of my state of mind. The death of the long tail in paid search was not something that happened overnight and the full launch of Google Suggest is not the sole reason, it's just the final nail in the coffin. Lots of little things have got us to where we are today but in my opinion two major events over the past 12 - 24 months are primarily responsible for killing of the tail in paid search.

Reason 1: Google cranked the dial on EBM (expanded broad match) to a whole new level in 2007.

As recently as 2005 - 2006 you could set up a new campaign with 10's of thousands of long tail exact matched keywords and reap the benefits of your precise targeting via an extremely low cost per click. Odds are most of your keywords would have little to no competition and as long as you could maintain a decent quality score your ad would stay live and your bids would stay low. Each term would not produce many clicks but cumulatively the clicks/traffic for your 100,000 keyword account would add up to numbers that most would consider "significant".

For example maybe one of your exact matched keywords was something like "1997 honda accord replacement left mirror". Odds are you wouldn't have much competition on that specific phrase so even though it may only be handful of queries a year you could still capture that traffic at a very low cost. All those big name advertisers bidding $20 a click on broad matched terms like "cars" and "auto parts" wouldn't be an issue, they were not chasing the tail. Google realized they were leaving a ton of $$ on the table and loosened up EBM. Once that happened all those advertisers spending big money on broad matched terms started showing up in the space and driving up costs for everyone else. Whereas before you may have been 1 of a few ads showing on the page, Google decided via EBM that people bidding on "cars, "auto parts", "mirrors", "autos", "honda", etc should start showing up for a wider variety of terms that they deemed relevant.

That was major event #1 that contributed to the death of the tail in paid search. The cost efficiencies gained from massive keyword lists and low bids were essentially wiped out over a few short months. Why try and manage hundreds of thousands of long tail keywords when you could just punch in your top 20 terms and let Google do the rest?

Don't get me wrong - there's still efficiencies to be realized by utilizing exact matched long tail keywords but those efficiencies have been greatly reduced.

Reason 2: Google Suggest "suggests" shorter queries which from what I've seen in most (all???) cases have sufficient ad inventory to fill the page.

Why search for "used red hybrid cars in illinois" when Google suggests "used cars" instead? Which query do you think stands to make them the most $$$?


Google acknowledges this feature will "help formulate queries"...I would say a better description would be to "help formulate the most valuable queries". While not everyone searching will use the suggested terms I'm sure a fairly high % of people will...enough so that it will make a difference in search volume for both long tail and shorter terms. While I think just about everyone would agree that Suggest will drive up keyword costs on certain terms I do think this opens some new opportunities and efficiencies within the paid search space.

At minimum, look at what Google Suggest is "suggesting" for your base terms and make sure you have visibility for those terms. If you have a very extensive campaign in place already you're likely covered already but I have seen some some interesting suggestions for keywords I watch. You can also use the suggest feature to further refine your negative keyword lists.

I also think this is a great opportunity to rethink the structure of your account(s). I've taken "long tail" accounts with thousands of keywords and consolidated them down to a few hundred keywords. This became more of a common practice for me in 2007 as expanded broad match kept expanding. Why try and manage 20,000 keywords when you can get the same (or better) with 200 keywords? The time savings alone is huge and almost without fail I've been able to maintain or improve ROI using a smaller set of keywords.

As paid search professionals we're use to big changes from Google with little or no advanced notice that can change the way we do business overnight. This is one of those changes and I'm sure we'll see even more changes in the coming months. I think with this particular change there's an opportunity for paid search pros to do more with less...I guess we'll see how it plays out over the coming weeks.

Others are commenting on this change as well, here are some links to other discussions that you may find useful:

How Google Suggest Suggest Will Affect Search
~ Bigmouthmedia

Google.com Finally Gets Google Suggest Feature ~ Search Engine Land

New "Google Suggest" Tells You Where to Go ~ PC World

Google Suggest(s) a New Reputation Management Nightmare ~ Martin Bowling.com

Google Suggest - Long Tail Killer or New Opportunity? ~ 5 Star Affiliate Programs

9 Ideas How Google Suggest Could Change Search Marketing ~ Scott Clark

Google Suggest to Change the Ways of Search Engine Optimization
~ SER


How do you think Google Suggest will impact paid search and paid search marketers?

Labels: ,


6 Comments:

At 12:25 PM, Anonymous billy said...

I am curious on your long term study of how cutting down the keywords from managing over 20k keywords to a few hundred is improving ROI, and time efficiency. It will certainly cut down a lot of time for everyone.

 
At 6:06 PM, Anonymous Anonymous said...

Just squeezing out the "little guy" who can't afford to compete with the big spending competition...

Quite a shame.

 
At 7:18 AM, Blogger Jeremy Mayes said...

Billy -

I'm still working with a lot of smaller (in terms of keywords) accounts that are doing great. If you can tame EBS you can do great things with small keyword lists.

 
At 11:53 AM, Blogger Gregory said...

This is a great picture of how AdWords has changed over time. When you use the technique you describe, of reducing the number of keywords, eliminating lots of long tail ones, do you ever run into a problem of not being able to exact match on specific phrases, and instead your smaller number of broad match keywords not only bring in the people who are interested, but also people who broad match on queries that show the people have no interested in your client's site? I swear, I think lots of people only read the headline of an ad, and many seem to read nothing - they just click. Here negative keywords are needed, but sometimes there are too many query variations and you just can't form an effective negative keyword list. What options are there for dealing with the situation? Do you think it would be best to switch a campaign like this to containing only a lot of exact matches? Are there other strategies?

 
At 2:54 PM, Blogger Jeremy Mayes said...

Gregory, thanks for the comment, I'll try and respond the best I can.

"When you use the technique you describe, of reducing the number of keywords, eliminating lots of long tail ones, do you ever run into a problem of not being able to exact match on specific phrases,"

Sometimes, and in those cases I'll add the exact matched term of phrase to a separate ad group.

"and instead your smaller number of broad match keywords not only bring in the people who are interested, but also people who broad match on queries that show the people have no interested in your client's site?"

I use negative keywords right from the start and add more daily/weekly as they surface in search logs and reports.


"I swear, I think lots of people only read the headline of an ad, and many seem to read nothing - they just click."

Agreed. Where possible I try and qualify people with the ad headline...sometimes qualifying via the body text doesn't cut it.

"Here negative keywords are needed, but sometimes there are too many query variations and you just can't form an effective negative keyword list."

It's not unusual for me to add 50 - 100 negative keywords per broad matched keyword right from the start...with many more following in the subsequent days/weeks. If you're using broad match plan on dedicating some time fro adding negative keywords for the entire life of the campaign.

"What options are there for dealing with the situation? Do you think it would be best to switch a campaign like this to containing only a lot of exact matches? Are there other strategies?"

I'm still an advocate of exact match, just on a smaller scale...especially at the onset of campaign. Most of my exact match terms are proven winners with track records of high conversion rates.

Everyone should use exact match - the change for me was how I used it. The days of starting a new campaign with 5,000 exact matched terms is a thing of the past if you ask me. Not to say you can't get to a point where you have that many EM'd terms...just saying it doesn't make sense to start a campaign that way anymore.

 
At 8:51 AM, Anonymous Anonymous said...

I think the so called benefits, obtaining new negatives or running smaller campaigns that are easier to manage are peripheral compared to the amount of extra $$$ we are all gonna have to pay to appear on generic terms we don't normally chase.

 

Post a Comment

PPC Discussions Home

Subscribe to this site About PPC Discussions

PPC Discussions is a blog about paid search authored by Jeremy Mayes, a search marketer from Illinois.

Subscribe to PPC Discussions via e-mail

Advertising

PPC Search Engines

The Big 3 in PPC

Subscribe to this site Previous Posts

PPC Discussion Archives

Subscribe to this site Miscellaneous

Add to Technorati Favorites

Powered by Blogger