Poor account structure and bad campaign settings can cost you hundreds or thousands of wasted dollars every week. Like a physical for your account, good auditing helps to catch bad habits and some warning signs before costs spiral out of control. Here are 4 ways to uncover common mistakes in less than an hour of review.
1. Segment by Device
Is your website mobile ready? Is your product something you’d buy on a phone? Is the checkout easy to do with limited space and time?
If you answered no to any of those questions, it may not make sense to target mobile devices. But, you might be pleasantly surprised, especially on brand terms.
To find out for yourself, use the Segment menu and choose Device.
The result is a report that breaks down whichever segment you’re looking at, say Campaigns, into lines for each of the different devices:
If you don’t have AdWords conversion tracking set up, you can also get data by device in Google Analytics. Go to:
Visitors > Mobile > Devices
The report will separate the data more granularly, but you can use the sub-navigation above the report to roll up or drill down as you like:
Just as you shouldn’t mix targeting for display network and search network in one campaign, you should separate devices (mobile vs. tablet vs. desktop), so you can more directly target, measure, and value them.
2. Do a Bidding Sanity Check
There are a lot of layers to bidding, but the core is that you should you should pay more when you’re more likely to make more money.
The bidding hierarchy looks like this:
Exact Match > Phrase Match > Modified Broad Match > Broad Match
You should target known converting search queries that have the most data using exact match and generally be willing to pay more for them.
Broad match, which will attract a wide range of terms of varying value, is inherently less valuable, because you don’t know what you’re getting.
If you’re bidding more for broad match terms than exact match terms, you’re wasting money.
The same is true on the display network (formerly content network), which looks something like this:
Retargeting > Managed Placements > Automatic Placements
Automatic placements are like the (very, very) broad match of the display network. Managed placements are similar to exact match. Retargeting is often the most profitable display campaign, though not always.
3. Analyze Brand vs. Non-Brand Terms
Conversions on your brand terms are earned through other marketing. Sure, there are plenty of important things you need to do to make brand terms more successful, such as grouping them properly and implementing sitelinks.
I’ve seen more than a few clients who outsource their paid search management and things look profitable until you look past brand terms, which can account for more than 50 percent of conversions, but less than 10 percent of cost.
Non-brand keywords are your growth opportunity. They’re where you’ll be able to find new customers who aren’t as familiar with your brand and they’re a more scalable source of traffic.
Start by analyzing the KPIs of your account segmented by brand vs. non-brand keywords, such as:
It’s common to wonder whether you’re undervaluing non-brand terms with last click attribution (i.e., did the click on a non-brand term, search on a brand term and then convert). Both AdWords and Google Analytics default to the industry standard last click attribution.
Attribution is a complex topic and many of the specialized tools to tackle it are out of reach for most advertisers. However, you can use two free tools.
AdWords Search Funnel
This hard-to-find section of AdWords can show you the average clicks to conversion, top paths, conversions under First Click attribution and more. Go to:
Reporting and Tools > Conversions > Search Funnels (lower left corner).
The Top Paths report segmented by search query is a particularly good place to start.
Multi-Channel Funnels in Google Analytics
If you want to see inter-channel assists, the new version of Google Analytics has a similar report.
If you’re in the old version, click to the New Version in the top right. Then:
Multi-Channel Funnels > Top Conversion Paths
This is the result.
Chances are that you’ll see self-referrals are most common and that it doesn’t take that many clicks to drive conversion. If so, you can rest assured that your ROI metrics from last-click attribution are reasonable enough. You can start optimizing non-brand terms without worrying about collateral damage on your brand terms.
4. Examine Ad Group - Text Ad Harmony
The two most common issues with ad groups are that:
- Grouping isn’t tight enough.
- Ads don’t address target search queries and their goals.
Grouping Isn’t Tight Enough
Your text ad can’t appeal to everyone. But, you greatly increase your chance the more narrowly you cast the reach of that text ad.
Put another way, you’ll get much further writing specific ads that target “cat food”, “dog food”, and “pet food” keywords individually than as a group.
David Szetela shared a good rule of thumb: all of the keywords in the ad group should have at least two words in common.
Ads Don’t Address Target Search Queries & Their Goals
Great PPC ad writing attracts as many people as possible who are as qualified as possible. One of the easiest ways to improve click-through rate is to repeat the search query in the text ad, possibly in the headline.
Which search query or keyword should you target? Sort by the impressions and look for common terms in the top 3 on the list. If you can’t find enough common ground, split up the ad group accordingly.
Cut Wasted Spend with an Audit
It’s easy to forget the fundamentals or overlook some best practices when you’re building out new campaigns. Whether you’re taking over an account, reviewing someone else’s work, or just need a refresh, a regular checklist and audit can easily trim wasted spend.
Search and traffic sourcing are both crucial to luring shoppers to your website. In this article, "2 Successful Holiday Strategies for Online Retail", you'll learn how to use a two-pronged approach for your holiday search campaigns that combine top keywords with the best referral sites. Data in this article comes from SimilarWeb.