Have you ever decided to pause certain parts of your AdWords campaigns – whether it’s a keyword, ad group, or entire campaigns – due to the element seeming to underperform?
Turning off poor performing campaigns, ads, and keywords is the easiest part of our jobs. The saddest thing you can ever do in AdWords is pausing a seemingly relevant keyword that brings traffic and sales just because it’s not producing a high enough ROI right off the bat.
Finding ways to turn around the underperforming parts of your account is the key to high-performing, scalable AdWords accounts.
Pausing Is Easy, Not Smart
Let’s look at an example. This advertiser had paused a campaign for the product silicone iPhone cases because the ROI was too low:
The target ROI for the advertiser was 500 percent. With a mere 218 percent ROI, this campaign was more or less dead in the water. They determined it should be paused in order to use the funds elsewhere.
However, with a quick review of the performance by device (computer, mobile, tablet), it was evident that the poor results mainly came from mobile:
- Desktop was performing almost twice as well as the overall campaign with a 362 percent ROI.
- Mobile had a 0 percent ROI, despite requiring 30 percent of the total spend.
- Tablets had a slightly better return than the campaign average.
By simply turning off mobile, we were able to increase the ROI of the campaign to 320 percent. With further optimization (ad testing, nuanced bid management, RLSA implementation, ad scheduling) we were able to get to the 5x mark. The campaign now drives more than 5,000 clicks to the e-commerce store every month with a >500 percent ROI.
This is how AdWords accounts grow — not by pausing underperforming campaigns.
It’s easy to pause keywords, ad groups, and campaigns. However, if all you do is pause low-performing parts of your accounts, you’ll end up in a downward spiral.
Most Poor-Performing Areas of AdWords Can Be Improved by Optimizing
You have to optimize the poor-performing parts of your AdWords account in order to grow.
This isn’t rocket science. Seriously.
The following seven areas of AdWords are the ones most frequently missed when PPC managers pause campaigns or ad groups they could have optimized instead:
- Devices: desktop, mobile, tablets
- Losing variations in your ad split testing
- Ad schedule (hour of day and day of the week)
- Wrong search terms messing with your keyword performance
- Search partner network
- RLSA: Don’t bid too high, don’t bid too low
- Attribution models
Let’s dig into each of these areas.
1. Devices: Desktop, Mobile, Tablets
This might be one of the most important segments you’ll review. This segment can completely ruin your chances of scaling your AdWords campaigns.
It goes for every single element of your campaign: campaigns, ad groups, ads, and keywords.
You can have an ad you think is performing poorly in an ad test, but it’s actually just because it performs poorly on mobile devices. If you review mobile vs. desktop performance, then the ad is performing better.
A lot of ad groups or campaigns are also unnecessarily paused because the ad/group keyword is showing a low ROI. As PPC managers, we tend to pause underperforming elements, when we should get into the habit of stopping and thinking first.
What to do:
- Before pausing a keyword/ad group due to low ROI, check the device segment and try to set the necessary bid adjustments. This might include completely turning off a specific device. I’d rather see you turn off just one device than the entire keyword across all devices.
- If the campaign is in the top 20 percent of conversion/traffic volume, consider breaking out the campaign into mobile, desktop, and tablet campaigns.
- Have realistic expectations for your mobile performance. Depending on your products, it may not perform at the same level as your desktop traffic.
2. Losing Variations in Your Ad Split Testing
When you’re reviewing your keywords for bid changes, see if there are any keywords you need to pause.
We often sit with blinders on, just looking at keywords and then look to the right for its performance data. Or as many of us do nowadays, export all keywords to Excel and apply some fancy filters, conditional formatting, or the famous pivot tables.
The issue with this approach is that you’re ignoring all the other potential influences there can be on a keyword than just the bid.
Let’s say you’re running an ad test with two ads. One ad converts at a 10 percent conversion rate ($10 CPA, 100 conversions) and ad B converts at 2 percent ($50 CPA, 20 conversions). The average conversion rate is 6 percent with an average CPA of $16. Any keyword in this ad group will theoretically see a decrease in their CPA by 62.5 percent if you pause the poor-performing ad.
Instead of solely looking at your keywords and changing your bids, try to look more holistically and include your ad tests in your bid management.
3. Ad Schedule (Hour of Day & Day of the Week)
For most accounts you’ll be able to find varying performance based on two time segments:
- Hour of day
- Day of the week
Go to Dimensions –> Time –> Hour of Day or Day of Week to see your own performance.
This isn’t an exact science. You should always be careful with shutting off a specific hour during the middle of the day if there isn’t a good reason for it. You might be getting clicks at lunch time that consumers complete at their computer at home.
However, there are some typical trends in e-commerce you can review for. In some industries, Sundays perform incredibly poorly (for others, they’re key days). Midnight to 7 a.m. is typically a dead zone — very little happens.
With ad scheduling, you’re able to set bid adjustments based on the time of day (i.e. from 1 p.m. to 6 p.m. I want to bid higher/lower) or on a specific weekday (I sell little on Sundays, so I want to bid 50 percent lower).
4. Wrong Search Terms Messing With Your ROI
Are the wrong search terms being triggered by a keyword? Although this issue isn’t as common, it’s still worth investigating.
Any keyword in either phrase match, broad match modified, or plain broad match has a chance of triggering searches that are irrelevant for the audience you’re trying to attract.
An example is the keyword queen size adjustable bed. It’s not uncommon for such a keyword to trigger for the following search queries:
- queen size adjustable beds – 🙂
- adjustable beds – :/
- king size adjustable beds – 🙁
- adjustable bed frames – :’(
- mattress for adjustable beds – :@
Your keyword (queen size adjustable beds) might convert at 10 percent, but because you have so many other search terms stealing budget, the overall average conversion rate for your keyword will be low and could push you to pause an otherwise profitable search term.
Not Seeing Any Conversions? You Might Have the Same Issue
In the example above, we had one search term that was profitable and others that weren’t.
However, you can still experience this problem of other search queries keeping your main keyword from performing well even if you don’t have any conversions. In the example above you could “easily” see that if you just turned off the rest of the search terms, then your main keyword could be profitable.
If all the other search terms are hogging all the budget from the campaign they’re associated with, then your actual keywords might never get a chance to convert.
5. Search Partner Network
My good old friend, the search partner network, the network of mysteries and myths.
There is only one reason why you should ever turn off the search partner network: low ROI.
A lot of people turn off the search partner network due to “low CTR.”
NO! Just no.
There are two different quality score calculations for the search partner network and the Google search page, so you shouldn’t worry about the sometimes lower CTR you find on the search partner network.
It’s always wise to review whether the search partner network is profitable for you. It takes three seconds and it’s done. Check every three months just in case.
Note: There is no need for you to check the search network at any other levels than your campaign level.
6. RLSA: Don’t Bid Too High, Don’t Bid Too Low
A rare but costly mistake is bidding too high for your RLSA audience.
One of the standard recommendations from Google’s account managers (at least at the time of this writing) is to assign a 100 percent bid adjustment to your RLSA audience.
In essence, this is not bad advice. It’s excellent to bid aggressively to ensure a top position for users who have already visited your site.
However, the issue appears when your keywords/ad groups are already in the first position. For many of the e-commerce accounts I run, we have 10 to 40 percent of our keywords we’ve honed in on (often exact match) that we have squared away solidly in the 1.5 to 2.2 average position range. Adding a 100 percent bid adjustment to that is simply unnecessary and will needlessly increase your CPA and lower your ROI.
7. The Magic Elixir: Attribution Models
Last but not least, you should always review various attribution models to ensure you’re not pausing campaigns or ad groups that are contributing at a different level than just last-click.
A quick explanation about attribution:
- Last Click Attribution means the source that receives the last click before a transaction gets credited the revenue.
- First Click Attribution means the source that receives the first click before a transaction gets credited the revenue.
If your account is built correctly, you will without a doubt have ~20-40 percent more conversions scattered on your not-so-well-performing segments than you realize:
In the example above, I’ve reviewed the last 30 days for one account.
- The revenue that direct traffic contributes is cut in half when I compare directly to a first interaction model.
- Revenue contributed to paid search increases 27 percent.
- Revenue contributed to SEO increases 155 percent!
If you’ve ever tried pausing low-performing parts of your Google AdWords campaign expecting your ROI to increase in the following weeks but were met with a decreased ROI instead — this is the reason. You shot down areas of your campaign that didn’t produce last-click conversions but helped capture consumers earlier in the buying funnel.
Don’t Forget About Ad Testing
While we’re on the subject of pausing AdWords campaigns, it’s important to bring up the subject of ad testing.
If you’re seeing keywords or ad groups that aren’t converting at a satisfactory level but are otherwise bringing in traffic and have healthy search volume (i.e., search volume is in the top 25 percent of your account), then testing your ads before blindly pausing the keyword/ad group should always be considered.
Again, the easiest thing to do in AdWords is to pause. It really doesn’t take much to see a keyword or ad group is not doing well if you’re not getting any conversions from it. But researching what your competitors are doing and writing new, better ads that steal customers away from your competition — that’s where the true magic of PPC lies.
Takeaway Tips to Implement Today
- Segment your most important campaigns in mobile, tablet, and computer. This will already give you a fighting chance before you get started.
- Get in the habit of checking your segments prior to pausing anything.
- Check your segments habitually. Schedule it in or you might forget about them.
The Whole Is Greater Than the Sum of Its Parts
It’s important that you see these areas as supplementing each other. If you get a 20 percent lift from mobile segmentation or a 15 percent lift from pausing a poor ad, customize your bids to get more clicks from higher performing ad groups a bit, then you’ll see that it all adds up in the end.
So stop pausing underperforming campaigns and tweak them instead. Your clients, bosses, and salary will thank you for it.
Featured Image: DepositPhotos
All screenshots taken by the author
Subscribe to SEJ
Get our daily newsletter from SEJ's Founder Loren Baker about the latest news in the industry!