It's the dream. To be able to log into AdWords and instantly find the changes you need to make. Well that's the trick that long-time paid search managers have been using to allow them to keep an eye on so much material so easily.
If you save the filters below in your own accounts (at keyword level) I guarantee you'll save time and always know what needs attention in your PPC campaigns.
How to set it up:
Spend > $X
CPA > $Y
Avg Position better than Z
Sorting: Conversions from high to low
What it's for: This filter will identify the keywords in your account that contribute most to your total conversions, but are costing you more than you can really afford per conversion. The keywords caught by this filter are high enough up the page that you have scope to reduce the CPA by lowering your positions/bids.
Tips: Set the CPA limit to a tolerance just a little higher than your target. You want this filter to identify keywords whose cost is above what is profitable for you. Set the average position to a slot above your average, where there is plenty of scope for keywords in those positions to be moved down and still attract reasonable traffic.
Actions: Reduce bids. Any keywords showing up in this filter are important to your account, but too expensive at the moment. They are high enough up the page that you can reduce the bids without demolishing traffic volumes.
How to set it up:
Spend > $X
CPA > $Y
Avg Position worse than Z
Sorting: Spend from high to low
What it's for: This filter will find the keywords in your account that are too expensive but don't have much scope to be reduced further. Decreases in bids here will likely move these keywords off the first page more and more often, reducing impression share down very low. These keywords will be expensive no matter what position they are in.
Tips: You can set up this filter with a position tolerance at the edge of the banner (usually top three positions) or down at the edge of the first page (usually nine or 10). In our experience the positions on a side are fairly similar in terms of performance. CTR won't vary significantly, but impression share will drop off as you move lower.
Actions: Reduce bids, but don't be afraid to pause. Sometimes the keywords in this filter simply won't work. If any keyword shows up in this filter for period after period you'll probably find that even when it shows mostly off the first page it still won't start converting well for you.
How to set it up:
Spend > $X
Conversions = 0
Sorting: Spend from high to low
What it's for: Finding keywords that simply don't work.
Tips: Check your timelines. Just because a keyword didn't work in this period doesn't mean it never did. It could have been through a blip. Always check keywords in this bracket across multiple timelines.
Actions: Bearing in mind the disclaimer above about timelines, you can usually safely pause these keywords. These are spending your money and showing you nothing in return. If your business takes a large proportion of leads over the phone then don't necessarily be so hasty. Consider phone-tracking methods if you're in this situation.
How to set it up:
Spend > $X
CPA < $Y
Avg Position worse than Z
Sorting: Conversions from high to low
What it's for: These keywords are make or break for your campaign. These are the keywords that have good traffic volumes, convert at below target CPA, and have scope to move up the page and get more daily impressions.
Tips: Set your average position pretty high on this one. The idea is to find keywords that can get more impressions if they're moved up the page. Almost anything not right at the top could potentially improve so be bold here.
Actions: Increase the bids. For as long as the keywords keep their low CPA, keep pushing those bids up. You want as much traffic as you can get from these keywords.
How to set it up:
Impressions > X
Quality Score < Y
CPA > $Z
Sorting: Impressions from high to low
What it's for: Identifying keywords that are having a negative impact on your overall campaign and account level quality scores. These keywords hurt the performance of all the other keywords in the campaign. But monetary performance has to come first.
Tips: Set your CPA limit such that only keywords that are not profitable for you are included. Any keywords with low quality scores but good CPAs should be kept, and you need to look at your ads to try to improve the quality score.
Actions: Pause these keywords. Run down the list, dealing with the highest impression volume keywords first. These are the terms that are having the biggest impact on your campaign. The more impressions you get with low quality scores, the more you're going to have to pay for everything else in the long run. If a particular keyword looks nicely relevant and targeted and looks like it should work, then give it a stay of execution for now. Work on your ad text and your landing pages to improve your CTR and relevancy.
There are two pieces of information we'd love to be able to act on but currently can't.
One example of this would be a filter to show all keywords that contain a specific word or phrase that is particularly important to you, and where average position is below X. This type of filter would help you to see keywords of particular relevance that should be getting as much traffic as possible but aren't.
By: Alistair Dent,
If you save the filters below in your own accounts (at keyword level) I guarantee you'll save time and always know what needs attention in your PPC campaigns.
Filter 1: Great But Expensive
How to set it up:
Spend > $X
CPA > $Y
Avg Position better than Z
Sorting: Conversions from high to low
What it's for: This filter will identify the keywords in your account that contribute most to your total conversions, but are costing you more than you can really afford per conversion. The keywords caught by this filter are high enough up the page that you have scope to reduce the CPA by lowering your positions/bids.
Tips: Set the CPA limit to a tolerance just a little higher than your target. You want this filter to identify keywords whose cost is above what is profitable for you. Set the average position to a slot above your average, where there is plenty of scope for keywords in those positions to be moved down and still attract reasonable traffic.
Actions: Reduce bids. Any keywords showing up in this filter are important to your account, but too expensive at the moment. They are high enough up the page that you can reduce the bids without demolishing traffic volumes.
Filter 2: Good, But Not Much Scope
How to set it up:
Spend > $X
CPA > $Y
Avg Position worse than Z
Sorting: Spend from high to low
What it's for: This filter will find the keywords in your account that are too expensive but don't have much scope to be reduced further. Decreases in bids here will likely move these keywords off the first page more and more often, reducing impression share down very low. These keywords will be expensive no matter what position they are in.
Tips: You can set up this filter with a position tolerance at the edge of the banner (usually top three positions) or down at the edge of the first page (usually nine or 10). In our experience the positions on a side are fairly similar in terms of performance. CTR won't vary significantly, but impression share will drop off as you move lower.
Actions: Reduce bids, but don't be afraid to pause. Sometimes the keywords in this filter simply won't work. If any keyword shows up in this filter for period after period you'll probably find that even when it shows mostly off the first page it still won't start converting well for you.
Filter 3: Doomed
How to set it up:
Spend > $X
Conversions = 0
Sorting: Spend from high to low
What it's for: Finding keywords that simply don't work.
Tips: Check your timelines. Just because a keyword didn't work in this period doesn't mean it never did. It could have been through a blip. Always check keywords in this bracket across multiple timelines.
Actions: Bearing in mind the disclaimer above about timelines, you can usually safely pause these keywords. These are spending your money and showing you nothing in return. If your business takes a large proportion of leads over the phone then don't necessarily be so hasty. Consider phone-tracking methods if you're in this situation.
Filter 4: Gold
How to set it up:
Spend > $X
CPA < $Y
Avg Position worse than Z
Sorting: Conversions from high to low
What it's for: These keywords are make or break for your campaign. These are the keywords that have good traffic volumes, convert at below target CPA, and have scope to move up the page and get more daily impressions.
Tips: Set your average position pretty high on this one. The idea is to find keywords that can get more impressions if they're moved up the page. Almost anything not right at the top could potentially improve so be bold here.
Actions: Increase the bids. For as long as the keywords keep their low CPA, keep pushing those bids up. You want as much traffic as you can get from these keywords.
Filter 5: Quality Score Duds
How to set it up:
Impressions > X
Quality Score < Y
CPA > $Z
Sorting: Impressions from high to low
What it's for: Identifying keywords that are having a negative impact on your overall campaign and account level quality scores. These keywords hurt the performance of all the other keywords in the campaign. But monetary performance has to come first.
Tips: Set your CPA limit such that only keywords that are not profitable for you are included. Any keywords with low quality scores but good CPAs should be kept, and you need to look at your ads to try to improve the quality score.
Actions: Pause these keywords. Run down the list, dealing with the highest impression volume keywords first. These are the terms that are having the biggest impact on your campaign. The more impressions you get with low quality scores, the more you're going to have to pay for everything else in the long run. If a particular keyword looks nicely relevant and targeted and looks like it should work, then give it a stay of execution for now. Work on your ad text and your landing pages to improve your CTR and relevancy.
Honorable Mentions
There are two pieces of information we'd love to be able to act on but currently can't.
- Google-only CTR. When Google revamped the reporting interface they removed the ability to segment keyword data by network, so you cannot see the effect that the search partners are having on your stats. You would definitely want to be able to see your keywords that have low CTRs but search partner activity can mask the true nature. There are two main ways around this:
- Segment by Top vs Side. This new segment will let you see that data separately, but it's not ideal for our purposes (screenshot above).
- Use the API. The data is still available (for now) in the API for those of you who really want to see it. This isn't the most straightforward thing to set up for most campaign managers though. If you have API access then I suggest you take a look at this data. If you don't, then just use the method above.
- Impression share per keyword. Google give us impression share data at campaign level, but not at keyword level. The ability to see how much additional traffic each individual keyword missed out on due to its rank on the page would be of great benefit. The filters above use average position as an analogue to impression share, to get an idea how much more traffic would have been available if we'd been more aggressive. If we really knew how much we'd missed we'd be able to make a better estimate of the scope still available to us.
Using Filters
The filters in this article are the types of filters used by experienced campaign managers to quickly get to the source of potential problems in your account and make rapid improvements to your performance. There are many others you can use to identify items of key importance to your own specific account.One example of this would be a filter to show all keywords that contain a specific word or phrase that is particularly important to you, and where average position is below X. This type of filter would help you to see keywords of particular relevance that should be getting as much traffic as possible but aren't.
By: Alistair Dent,
No comments:
Post a Comment