Andy Black
Help people. Get paid. Help more people.
Staff member
FASTLANE INSIDER
EPIC CONTRIBUTOR
LEGACY MEMBER
Read Fastlane!
Speedway Pass
Don’t Let Google Control Where Your Traffic Goes
We were running two similar campaigns, one for Cooker Repairs, and one for Oven Repairs.
Each campaign contained two adgroups, and each adgroup had only one keyword, and only one ad.
The only difference between the keywords within a campaign was the match type.
You can imagine it like this:
The ads were a very good match for the keywords we were bidding on.
Even the domain name was customised to the keyword.
After 200 impressions, we’d had 15 clicks for a CTR of 7.5%.
We were curious to see whether people searched for Cooker Repairs more than Oven Repairs.
And surprised to see that all 200 impressions went to the Cooker Repairs campaign, with none going to the Oven Repairs campaign!
However, when we brought up the Search Term report for the Cooker Repairs campaign, we saw that a lot of the impressions were caused by people searching for “oven repairs dublin”.
This seemed strange since we would have expected this traffic to have gone to the [oven repairs dublin] keyword in the Oven Repairs campaign.
Even though we were bidding in exact match on that search term, the traffic was heading towards a broad match keyword in a different adgroup in the account.
The ad that these searchers were seeing showed a headline of “Dublin Cooker Repairs”, which didn’t get a good CTR because they were looking for oven repairs.
Actions Taken
We added “oven repairs” as a negative keyword into the Cooker Repairs campaign to force the traffic to the right adgroup.
And also added “cooker repairs” as a negative keyword in the Oven Repairs campaign to be consistent.
Our campaign now looked like this:
The Results
Adding the negative keywords now meant that people looking for “oven repairs dublin” saw an ad for “Dublin Oven Repairs”, rather than “Dublin Cooker Repairs”.
Our CTR almost immediately doubled.
All the searchers that were previously being shown the wrong ad, suddenly got shown the ad that was a perfect match for what they were looking for.
The increased CTR had a knock-on effect of increasing our impression share slightly (meaning our ads showed more often when they were eligible to show), and Google rewarded us with a lower CPC.
It was only when we looked at the actual search terms people were typing in did we see the big difference between what we thought we were doing, and what was actually happening.
Conclusions
You already know this one… be aware that search terms are NOT the same as keywords.
Structure your account accordingly if you want to maximise your CTR and conversion rates.
A more important conclusion might be to not trust the AdWords algorithm to do what you think it will do.
If traffic has the option to go to different keywords within your account, then you are giving Google the option to arbitrarily decide where your traffic can go.
I’ve seen instances where dropping bid prices in an unprofitable adgroup has caused the traffic that was going to that adgroup to “pop” up somewhere else in the account, where there happens to be a broad match keyword that has a higher bid price.
Imagine if we’d paused the Cooker Repairs campaign because it didn’t seem to be running well. And then we suddenly get 200 impressions to the Oven Repairs campaign.
Would we conclude that there was a sudden breakdown of ovens in Dublin that had been uncannily fault free up until then?
What effect does this moving around of traffic in your account do to all your ad and landing page testing?
What if you dropped bids in one corner of your account, and the traffic got picked up by some other corner of your account. It’s like some weird water bed where you push down on bids in one location, for the traffic to pop up somewhere else that you’re not looking at.
Maybe bid adjustments somewhere else in your account cause ads you are optimising to suddenly perform badly, or their landing page to no longer convert as well?
…when in fact all that has happened is that your broad match keywords are suddenly matching a larger set of search terms than previously, and the new search terms they are matching are less relevant?
Unless you start analysing your account at the search term level and see how your traffic is moving between campaigns and adgroups, then you aren’t going to spot how this can affect all the test results and conclusions you have reached to date.
This is particularly insidious in that you don’t even know you’ve built your house on sand, and another reason why you shouldn’t mistake keywords for search terms.
Not if you want to make the most profit you can out of AdWords that is.
----------
Want to keep going down the rabbit hole?
Read the next post.
We were running two similar campaigns, one for Cooker Repairs, and one for Oven Repairs.
Each campaign contained two adgroups, and each adgroup had only one keyword, and only one ad.
The only difference between the keywords within a campaign was the match type.
You can imagine it like this:
The ads were a very good match for the keywords we were bidding on.
Even the domain name was customised to the keyword.
After 200 impressions, we’d had 15 clicks for a CTR of 7.5%.
We were curious to see whether people searched for Cooker Repairs more than Oven Repairs.
And surprised to see that all 200 impressions went to the Cooker Repairs campaign, with none going to the Oven Repairs campaign!
However, when we brought up the Search Term report for the Cooker Repairs campaign, we saw that a lot of the impressions were caused by people searching for “oven repairs dublin”.
This seemed strange since we would have expected this traffic to have gone to the [oven repairs dublin] keyword in the Oven Repairs campaign.
Even though we were bidding in exact match on that search term, the traffic was heading towards a broad match keyword in a different adgroup in the account.
The ad that these searchers were seeing showed a headline of “Dublin Cooker Repairs”, which didn’t get a good CTR because they were looking for oven repairs.
Actions Taken
We added “oven repairs” as a negative keyword into the Cooker Repairs campaign to force the traffic to the right adgroup.
And also added “cooker repairs” as a negative keyword in the Oven Repairs campaign to be consistent.
Our campaign now looked like this:
The Results
Adding the negative keywords now meant that people looking for “oven repairs dublin” saw an ad for “Dublin Oven Repairs”, rather than “Dublin Cooker Repairs”.
Our CTR almost immediately doubled.
All the searchers that were previously being shown the wrong ad, suddenly got shown the ad that was a perfect match for what they were looking for.
The increased CTR had a knock-on effect of increasing our impression share slightly (meaning our ads showed more often when they were eligible to show), and Google rewarded us with a lower CPC.
It was only when we looked at the actual search terms people were typing in did we see the big difference between what we thought we were doing, and what was actually happening.
Conclusions
You already know this one… be aware that search terms are NOT the same as keywords.
Structure your account accordingly if you want to maximise your CTR and conversion rates.
A more important conclusion might be to not trust the AdWords algorithm to do what you think it will do.
If traffic has the option to go to different keywords within your account, then you are giving Google the option to arbitrarily decide where your traffic can go.
I’ve seen instances where dropping bid prices in an unprofitable adgroup has caused the traffic that was going to that adgroup to “pop” up somewhere else in the account, where there happens to be a broad match keyword that has a higher bid price.
Imagine if we’d paused the Cooker Repairs campaign because it didn’t seem to be running well. And then we suddenly get 200 impressions to the Oven Repairs campaign.
Would we conclude that there was a sudden breakdown of ovens in Dublin that had been uncannily fault free up until then?
What effect does this moving around of traffic in your account do to all your ad and landing page testing?
What if you dropped bids in one corner of your account, and the traffic got picked up by some other corner of your account. It’s like some weird water bed where you push down on bids in one location, for the traffic to pop up somewhere else that you’re not looking at.
Maybe bid adjustments somewhere else in your account cause ads you are optimising to suddenly perform badly, or their landing page to no longer convert as well?
…when in fact all that has happened is that your broad match keywords are suddenly matching a larger set of search terms than previously, and the new search terms they are matching are less relevant?
Unless you start analysing your account at the search term level and see how your traffic is moving between campaigns and adgroups, then you aren’t going to spot how this can affect all the test results and conclusions you have reached to date.
This is particularly insidious in that you don’t even know you’ve built your house on sand, and another reason why you shouldn’t mistake keywords for search terms.
Not if you want to make the most profit you can out of AdWords that is.
----------
Want to keep going down the rabbit hole?
Read the next post.
Dislike ads? Remove them and support the forum:
Subscribe to Fastlane Insiders.
Last edited by a moderator:
