There is a significant performance difference between desktop and mobile traffic on the display network (just like there are differences between text and banner ad performance).
By running desktop and mobile traffic in separate ad groups you can analyse and optimise performance much faster. If you are running text ads and banner ads on the display network you should combine this strategy with the “banner vs text ad” strategy (as outlined in the previous shortcut here).
Your campaign structure then looks like this:
Display Campaign
– Ad Group 1 [Text] [Mobile]
– Ad Group 2 [Text] [Desktop]
– Ad Group 3 [Banner] [Mobile]
– Ad Group 4 [Banner] [Desktop]
– For each [Desktop] ad group you need to set the mobile bid adjustment to -100%
– There are two ways to ensure that [Mobile] ad groups only generate mobile traffic. One solution is to create a mobile-only campaign (change Settings -> Devices to mobile only). The second solution is a bit of a workaround where you might still receive a tiny amount of desktop traffic. In your desktop and mobile campaign, set device preference to “Mobile” for all ads in the mobile ad groups. Then set the max. cost per click in the mobile ad group lower than in the desktop ad groups and increase the mobile bid adjustment to at least 100%. This will basically set the max. cost per click so low that your mobile ad groups can’t generate any desktop traffic. The mobile bid multiplier raises the mobile bids to a high enough level so that these ad groups will only generate mobile traffic.
– If the mobile ad groups still receive desktop impressions, decrease your cost per click further and increase your mobile bid adjustment further. It’s currently the only workaround to force Adwords to run the ads on mobile but not on desktop devices (if you don’t want to setup a separate mobile-only campaign).
To analyse performance simply click on the campaign -> Ad Groups -> search for “desktop”, “mobile”, “text” or “banner” to apply a filter.
Make sure to add the following columns: converted clicks, cost / converted click, click conversion rate, bounce rate and avg. session duration.
