Simplify PPC management
Learn moreOne of the most crucial PPC tasks is negative keyword management. In the past 3-5 years, managing negative keywords has increased in importance and complexity as much, or more, than any other PPC management tactic. There are a few reasons for this.
Looser Match Types
With the changes to exact and phrase match over the past few years, the number of search terms your keywords can match has expanded dramatically.
We commonly see 2-3 word exact match terms match many 5+ word search terms. While some of these terms are useful, they are a waste of money in many cases.
The examples and ideas about exact match changes and phrase match inaccuracies have not changed in several years.
Smart Bidding Often Does Not Learn
We commonly see terms that can spend a considerable amount of money and receive a tremendous number of clicks (enough for a machine to learn), and yet smart bidding does not accommodate these trends.
An easy to illustrate example (more info here) comes from an advertiser who spends millions each month. Their average CPA is around $300. They were using target CPA bidding.
When the search term contains the word ‘how,’ their CPA is over $1000.
After hundreds of thousands of dollars and more than 10,000 clicks in just 90 days, Google never adjusted the bids on these types of words to bring the CPA of question words in-line with the account’s CPA target.
Duplicate Search Terms
The other issue that has repeatedly happened the past few years with the looser match types is Google showing the exact same term from multiple ad groups. When one ad group converts at 10% and another at 1%, you know which ad group should be showing for a search term.
While Google tries to show the ad group that has a keyword that matches the search term exactly, any variation, such as a misspelling, singular vs. plural word, or different word order, Google may ignore that trumping logic. You need to step in and take action, such as adding the new keyword variation or adding a negative keyword at the ad group level.
Putting Guardrails in Place
One of your tasks is to put guardrails in place to keep the machine driving in your desired direction. This helps keep the machine learning within your desired framework and not going out on its own into areas where you do not want it to travel.
One of the best ways to do this is with negative keywords. It is generally better to use n-gram analysis for these tasks as you can examine much more data in a smaller amount of time than reading each individual search term.
Make Sure Your Guardrails are Not Blocking Desired Traffic
The biggest issue to watch when adding negative keywords is negative keyword conflicts. Google’s recommendations do not consider match types or campaign negative lists when showing keyword conflict recommendations.
On a side note, Adalysis does look at match types, ad group organization, and campaign negative lists when determining keyword conflicts.
Steering the Machine
Most accounts are using some smart bidding, which will not change. Most accounts are using multiple match types, which will not change either.
Your job is to steer the machine and put guardrails in place to keep the machine from making these repeated mistakes. The machine will try to learn, but sometimes it’s not asking the correct questions. When the machine is not built to learn something, the human must step in and guide the machine in the proper direction, and that’s your job.
The first step towards steering the machine is the proper usage of n-gram analysis with negative keyword management. You can follow this up by managing duplicate search terms and negative keyword conflict resolution.
As match types continue to get loser, more accounts move toward smart bidding; negative keyword management has become one of the primary tasks of any successful PPC manager.
If you are looking for software to help you steer the machine in the proper direction, take a look at Adalysis. It is built by PPC marketers for PPC marketers who want a workflow that combines the knowledge and creativity of humans and the technology and scale of machines. You can take it for a free 2-week trial to see how it can improve your account.
Additional reading (with video): Everything You Need to Know about Negative Keywords to Improve Your PPC Results.
Impression Share can be used as a diagnostic tool, along with other analyses, to determine the steps you need to…
Target and Max CPA/ROAS bidding are the most common automated bid solutions used by Google Ads accounts. There are some…
It's that time of the year when we start thinking about what's coming next. While making predictions for the upcoming…
SMX Next is fast approaching, and best of all - it is free! Brad will be speaking in the session:…
Most of the time, accounts that are using smart bidding (i.e. automated bidding) ignore bid adjustment since most of them…
Target and Max bidding are separated by a single checkbox. However, the implications of that single check can be quite…
View Comments
Thanks for the article, Brad!
Your articles and Adalysis are a good help in optimizing advertising campaigns.
An important addendum to Googles ongoing push for "smartness". It should be noted that a growing number of campaign types will not have (insight into) keywords or negatives at all, foremost with Shopping. Thus, steering the machine with carefully sculpted lists of negatives will go hand in hand with sticking to "dumber" campaigns that offer the required insight and control.
It will take experience and stubbornness to withstand the many invitations to just leave it all to the machine.
that’s an amazing article
Interesting read and recognize the difficulties. Keywords and negative keywords are very intertwined with ad group organisation. Do you have any thoughts on a methodology approach for how keywords are placed in ad groups? Leaving beside obvious ad groups that should be different (trousers vs shirts), there are many mixed instances where you could do both (e.g. split plural and singular in different ad groups, or combine them in on ad group, split t-shirt and shirt in different ad groups, or combine them in one ad group). etc. etc. Once on route with splitting up ad groups, and knowing Google is ever expanding search terms, more and more negative keywords are needed at the individual group level to "steer the machine". The less you split up, the less steering maintenance needs to be done.
Hi Jos,
There are really two thoughts here. The first is to use ad customizers to manage ad lines for specific terms. With ad customizers, you can use less ad groups.
The second thought is really to keep the ad group organization the same as it's always been (very common, and generally leads to good results) > https://adalysis.com/blog/follow-easy-flow-chart-properly-organize-ad-groups/.
The biggest issue we see with RSAs, addressed in this video https://adalysis.com/blog/rsa-optimization-series-part-2-learn-to-create-and-test-rsas-in-google-ads/ is that a lot of agencies are collapsing their structure with RSAs and then adding a single ad line that's useful for only a subset of keywords and assuming Google will always match the proper ad line to only the keywords specific to that particular ad asset. This isn't happening right now, so you might have a single ad asset that says shirts and another ad line with the asset that mentions pants, and you end up with an ad for the keyword pants that talks about shirts.
Google is not doing a good job of doing asset to kewyord matching. So your ad really needs to be useful for all the keywords within the ad group regardless of which assets are displayed.
Brad, what was the match type of the keyword in your “how” example?
It was exact match, which is why it's kinda crazy what it will show for.