Andy Black
Help people. Get paid. Help more people.
Staff member
FASTLANE INSIDER
EPIC CONTRIBUTOR
Read Fastlane!
Speedway Pass
(Make sure you read this post first.)
Hidden Dangers of Wide Funnels
Once upon a time, in a small office in Dublin, we were sending traffic to a London Hotels landing page by bidding on the broad match keyword “london”.
Even with very low bids, the keyword was unprofitable, and the AdWords specialist running the campaign had been repeatedly dropping bids to stop losing money.
She had gotten into a vicious circle of dropping bids and losing traffic, until eventually the traffic dried up completely.
Of course, we were eligible to show for search terms like “plumbers london” or “cars for sale in london”, but our London Hotels ad had such a poor CTR that Google decided we weren’t relevant enough, and therefore not worth showing very often.
And rightly so!
Even worse, since we were bidding in broad match, our ad was eligible to show when the searcher typed in a search term that contained none of the words in our keyword.
In particular, since we were using a one word broad match keyword, we weren’t helping Google much by telling them when we wanted our ad to show.
And we were very likely showing for search terms that didn’t even contain the word “london”.
When we looked at the Search Term Report we found loads of search terms used by people blatantly not looking for a hotel in London.
One search term was “british army”. So at least one person searching for “british army” clicked on an ad for “London Hotels”.
Go figure.
Another search term was “uk sexy girls”, and we didn’t want to guess what the intention was there.
The weirdest search term we saw was “you tube”.
Why on earth would someone click on an ad for London Hotels if they typed in “you tube”?
Maybe it was just a misclick?
Then again, why would Google even show our ad when someone typed in “you tube”?
Then we remembered that the London underground is called the Tube.
Which made us realise how tenuous broad match can be.
Bear in mind that the Search Term Report only shows the search terms that brought you clicks if someone using that search term clicked on your ad.
So our ad was probably showing for even less relevant search terms, but we couldn’t see what they were because our ad was so irrelevant that no-one had even clicked on it.
I eventually persuaded the AdWords specialist to change the broad match keyword to a phrase match keyword.
Meaning that at least our ad would only show when people had “london” somewhere in their search term.
Which was still pretty poor, but better than what we had previously.
We left the campaign for two weeks without touching it.
We started to see the CTR rising, the ad position rising, and the CPC falling.
Our impression share had always been “< 10%”.
Meaning our ad showed less than 10% of the time it was eligible to show.
But one day it pop up above the 10% mark, and we watched it rise daily through the teens up to the dizzy heights of 20%. (High given we were bidding something like €0.20.)
The campaign started being profitable, and we were even able to push bids a little bit to bring in more traffic at the expense of eating into our profit margin.
I’d had a very hard time convincing anyone in the team to move away from broad match.
The common worry was that we would get less traffic.
My argument was and still is that, yes, we’re closing the top of the funnel and are eligible to show for less search terms.
But our CTR is going to rise, and Google is going to reward us with a higher impression share.
Meaning we’re going to get a higher volume of more relevant traffic.
If we’re only getting a 10% impression share with a wide funnel, then you could visualise it like this:
If we narrow the funnel then we should reduce wasted impressions, increase our CTR because our ad is more relevant, and be rewarded with a higher impression share.
You could visualise it like this:
As mentioned before, I’d recommend having the narrowest funnel possible, and maxing out your CTR to get the full love from Google.
But there’s more conclusions that we can derive from the two visuals above.
In the wide funnel, we’re assuming that a 10% impression share means we had a 10% impression share for all the search terms, no matter how valuable.
That just doesn’t wash.
If we imagine that the search term “cheap hotels london” is the most valuable to someone advertising hotels in London.
Maybe you have to bid €2.00 to even appear on the bottom of the first page of Google.
Then if we were bidding, say, €0.40 to hoover up the (ahem!) “cheaper broad match traffic”, I would be willing to bet that our impression share for “cheap hotels london” would approach 0%.
Meaning our 10% impression share actually comes from all the junk keywords at the side of the funnel, and that thin slice of traffic is actually a curve with thick sides and a very pronounced dip in the middle.
So if our conversion rate is too low and we have to drop bids even more, then our ads are less and less likely to show for the search terms that are likely to convert (the valuable stuff in the middle).
So dropping our bids can have the effect of showing our ads more often for the search terms at the side of the funnel.
Which drops our conversion rate even further.
And we crash and burn … again.
Bummer.
So to summarise…
Using broad match unwisely means our ads don’t show for the search terms we intended them to show for, allows traffic to go to different keywords in our accounts depending on Google’s preference rather than ours, and now can also change our conversion rates if we amend our bid prices.
That just sucks…
----------
Want to come back up for air and apply what you've learnt?
Read this post.
Hidden Dangers of Wide Funnels
Once upon a time, in a small office in Dublin, we were sending traffic to a London Hotels landing page by bidding on the broad match keyword “london”.
Even with very low bids, the keyword was unprofitable, and the AdWords specialist running the campaign had been repeatedly dropping bids to stop losing money.
She had gotten into a vicious circle of dropping bids and losing traffic, until eventually the traffic dried up completely.
Of course, we were eligible to show for search terms like “plumbers london” or “cars for sale in london”, but our London Hotels ad had such a poor CTR that Google decided we weren’t relevant enough, and therefore not worth showing very often.
And rightly so!
Even worse, since we were bidding in broad match, our ad was eligible to show when the searcher typed in a search term that contained none of the words in our keyword.
In particular, since we were using a one word broad match keyword, we weren’t helping Google much by telling them when we wanted our ad to show.
And we were very likely showing for search terms that didn’t even contain the word “london”.
When we looked at the Search Term Report we found loads of search terms used by people blatantly not looking for a hotel in London.
One search term was “british army”. So at least one person searching for “british army” clicked on an ad for “London Hotels”.
Go figure.
Another search term was “uk sexy girls”, and we didn’t want to guess what the intention was there.
The weirdest search term we saw was “you tube”.
Why on earth would someone click on an ad for London Hotels if they typed in “you tube”?
Maybe it was just a misclick?
Then again, why would Google even show our ad when someone typed in “you tube”?
Then we remembered that the London underground is called the Tube.
Which made us realise how tenuous broad match can be.
Bear in mind that the Search Term Report only shows the search terms that brought you clicks if someone using that search term clicked on your ad.
So our ad was probably showing for even less relevant search terms, but we couldn’t see what they were because our ad was so irrelevant that no-one had even clicked on it.
I eventually persuaded the AdWords specialist to change the broad match keyword to a phrase match keyword.
Meaning that at least our ad would only show when people had “london” somewhere in their search term.
Which was still pretty poor, but better than what we had previously.
We left the campaign for two weeks without touching it.
We started to see the CTR rising, the ad position rising, and the CPC falling.
Our impression share had always been “< 10%”.
Meaning our ad showed less than 10% of the time it was eligible to show.
But one day it pop up above the 10% mark, and we watched it rise daily through the teens up to the dizzy heights of 20%. (High given we were bidding something like €0.20.)
The campaign started being profitable, and we were even able to push bids a little bit to bring in more traffic at the expense of eating into our profit margin.
I’d had a very hard time convincing anyone in the team to move away from broad match.
The common worry was that we would get less traffic.
My argument was and still is that, yes, we’re closing the top of the funnel and are eligible to show for less search terms.
But our CTR is going to rise, and Google is going to reward us with a higher impression share.
Meaning we’re going to get a higher volume of more relevant traffic.
If we’re only getting a 10% impression share with a wide funnel, then you could visualise it like this:
If we narrow the funnel then we should reduce wasted impressions, increase our CTR because our ad is more relevant, and be rewarded with a higher impression share.
You could visualise it like this:
As mentioned before, I’d recommend having the narrowest funnel possible, and maxing out your CTR to get the full love from Google.
But there’s more conclusions that we can derive from the two visuals above.
In the wide funnel, we’re assuming that a 10% impression share means we had a 10% impression share for all the search terms, no matter how valuable.
That just doesn’t wash.
If we imagine that the search term “cheap hotels london” is the most valuable to someone advertising hotels in London.
Maybe you have to bid €2.00 to even appear on the bottom of the first page of Google.
Then if we were bidding, say, €0.40 to hoover up the (ahem!) “cheaper broad match traffic”, I would be willing to bet that our impression share for “cheap hotels london” would approach 0%.
Meaning our 10% impression share actually comes from all the junk keywords at the side of the funnel, and that thin slice of traffic is actually a curve with thick sides and a very pronounced dip in the middle.
So if our conversion rate is too low and we have to drop bids even more, then our ads are less and less likely to show for the search terms that are likely to convert (the valuable stuff in the middle).
So dropping our bids can have the effect of showing our ads more often for the search terms at the side of the funnel.
Which drops our conversion rate even further.
And we crash and burn … again.
Bummer.
So to summarise…
Using broad match unwisely means our ads don’t show for the search terms we intended them to show for, allows traffic to go to different keywords in our accounts depending on Google’s preference rather than ours, and now can also change our conversion rates if we amend our bid prices.
That just sucks…
----------
Want to come back up for air and apply what you've learnt?
Read this post.
Dislike ads? Remove them and support the forum:
Subscribe to Fastlane Insiders.
Last edited by a moderator:
