PPC Analysis: Beyond Keyword Performance

Most PPC account managers would agree that an AdWords account built today would contain fewer keywords than the same account built several years ago. With modified broad match offering an alternative to handfuls of similar low volume phrase match keywords, and Product Listing Ads requiring no keywords at all (at least not in the traditional sense), our ability to generate lists of thousands of "long tail" keywords has less impact on our accounts' performance.

In fact, Google dissuades advertisers from bidding on a large number of low-volume terms; if a keyword is considered to be of “low search volume” it won’t even be eligible to enter the auction if it is searched for. As a result, we find ourselves either bidding on broader variants of our desired long tail keywords or accepting that we won't see impressions at all. 

So, with our keyword choices (or lack thereof) holding less sway in the success of our PPC advertising, where else must we focus our efforts?

Search Query Reports

If we find ourselves using more modified broad keywords and fewer phrase or exact match keywords, we need to check that the queries generating clicks to our ads are still relevant to our offering. Visit the keywords tab and select the option to view user search terms under the “details” drop down. Add any irrelevant queries to your negative keywords list.

Similarly, the search queries prompting clicks to your Shopping (Product Listing Ad) campaigns can be found in the “search terms” view of the dimensions tab.

Bid Modifiers

We have the ability to increase or decrease our bids depending on how likely a particular searcher is to convert, which we can surmise based on our knowledge of how similar users have converted historically (in light of their device type, location, or the time of day that they are searching, for example).

Device Type

A common and effective bid modifier tactic is to set bid adjustments according to the value that our mobile advertising offers relative to our desktop and tablet targeting. Advertisers without mobile-optimised websites, for example, would set decreased mobile bids if their website converted mobile visitors at a lower rate. 

Conversely, if campaign performance on mobile trumps desktop, we can set an increased bid for searches made on a smartphone.

Location

Bids can also be modified depending on user location. For example, SilverDisc is currently in the process of building an AdWords account for the car breakers U-Pull-It.

We have set our location targeting radius around their postcode, but have modified the bids to increase when the user is searching from an area closer to the yard (we presume that users nearer more likely to visit U-Pull-It's yards, i.e. will have a higher "conversion rate").

Time of Day

Bids can also be increased during particular times of the day or days of the week; try segmenting your dimensions tab by "hour of day" to see whether your conversion rate spikes at particular times of the day (but do be careful with this one as the time that user converted might not be the same as the time of their click).

Performance by Network

Segment your campaign data by "network (with search partners)" to see whether Google's search partners are delivering conversions at the same rate as Google search itself. If the performance isn't up to scratch, visit the campaign settings and switch off "search partner" targeting.

Similarly, if your display network targeting isn't already split out from your search advertising, check to make sure that you are seeing an acceptable performance. If not, visit the "placements" area of the "display network" tab and exclude placements that are generating lots of clicks but no conversions.

Of course, this doesn't mean that we shouldn't choose our keywords carefully or evaluate their performance regularly, but we do need to make sure that we consider all of the other factors that are influencing the performance of our accounts.

Feel free to get in touch if you need any help.