That’s a long list to check, right? In our experience, every single one of these items has been the culprit of an SEM performance dip at one time or another, so rather than leave any out, we’ve clumped them into handy buckets to help you navigate to the source of the problem.
Let’s dive in!
Unpredictable factors
Check the change history and investigate any suspicious changes
Did the client make any back-end changes?
Are all the pixels and tags functioning properly?
Did the site go down? (We use Uptime Robot to get real-time notifications)
Has your client received any negative/irrelevant news mentions? (I recommend you set up Google News Alerts to stay on top of this)
Keywords
Review low-spend, low impression share keywords that have a history of converting and increase bids
Review and negate poor-performing competitor keywords
Review keywords with high avg. pos and high impression share and decrease bids to limit headroom
Review new keywords trends (minimize the keyword gap)
Review top-converting queries from last year
Review top Shopping queries and add them in as keywords
Pause high-spend keywords that have not converted in the last 6 months
Pull a search query report to update your irrelevant negs and to propose keyword expansions
Demote sub-par alphas and clean up low-volume/superfluous betas to improve data density
Consider adding low-ROAS / high-CPA terms to an RLSA campaign
Check whether new URL redirects or tracking were added (possible broken links)
Website
Implement a link checker script to see if URLs are active
Test sending URLs to a new landing page to improve CVR
Review highest-CVR landing pages in Google Analytics and use insights to inform ad copy recs
Review highest-CVR landing pages in Google Analytics and use insights to inform landing page recommendations
Bidding & Modifiers
If using Smart Bidding, review the bidding report to see if you’re constantly throwing the algorithm into learning mode
Test tCPA or pause tCPA and revert back to manual bidding
Test turning off eCPC to have more bidding control
Review international geo modifiers and cut out bleeding locations
Review device breakdown and adjust bid modifiers
Adjust time-of-day modifiers to make more granular changes
Audiences
Review and clean up audiences (campaign level vs. ad group level)
Apply some cross-channel audiences (e.g. Pinterest visitors who haven’t converted)
If non-brand performance is particularly awful, consider only running non-brand with RLSA
Shopping
Bulk up your Shopping titles with top-volume queries (and make sure keywords are frontloaded)
Keyword-stuff your shopping descriptions
Make sure you have at least three tiers per product type
Check to make sure Shopping feeds are updating regularly
Review last year’s top-converting products during the same time period
Check for diagnostic errors in Google Merchant Center
Review the search query report for irrelevants and add them to the corresponding negative list
Segment poor-performing high-volume queries into query campaigns
Check whether products are falling into new product groups and impacting performance/causing them to be bid incorrectly
GDN
Implement new creative
Test new responsive ads
Test dynamic ads
Do your campaigns need a restructure?
Review settings; make sure you are not set to overly aggressive or conservative targeting
Assess whether high frequency rates on GDN/YouTube are causing ad fatigue
Review/scrub placements
Consider a smart display campaign
Ad Copy & Extensions
Check for ad disapprovals
Review ad copy and pause poor-performing variations
Review number of ads per ad group
Review consistency of ad settings between campaigns
Refresh ad extensions (does the site offer new value propositions to use?)
Review the competitive landscape; check competitive promotions
Review bulk actions/scheduling for errors/failures
Check feed source for failures if you’re using dynamic attributes
Sanity Checks
Perform some test searches and make sure the ads you want to serve are showing properly
Review query mapping
Assess whether you need a restructure
Review past performance to see comparisons
Ask a non-team member to audit account
Other
Review search partners for performance and turn off if they perform worse than the Google search network
Review QS (can be implemented by a script)
Consider launching DSA campaigns (note that these increase query coverage and often can reduce CPL / increase ROAS for non-brand campaigns)
Set up Google alerts so you are aware of potential increases in search volume and can stay on top of new initiatives
Google Analytics – assess whether trends are happening across channels
Ask client about other marketing initiatives happening that could be impacting brand searches
Are there any new betas worth testing?
Phew! If you’ve made it this far and need to identify an issue other than those on the list above, please drop a comment.
