Well played, Google: Advertisers (finally) get more device bidding controls

google-adwords-green2-1920
Just last week, I submitted this column with the title, “Is it time for Google to rethink Enhanced Campaigns?” The piece I wrote was rife with criticisms of the desktop/tablet-centric campaign structure advertisers have been stuck with, and it proposed some possibilities for an upgrade.

And just like that, before it was time for my original submission to publish, it happened. Google announced Tuesday that advertisers will now be able to specify the device of their choosing as the base bid for keywords, and then layer on separate bid modifiers for the other two device types.

This is great news for advertisers, even if embracing this golden gift did mean that I had to totally rewrite a perfectly good but scathing account of the old campaign structure.

In this freshly written column, I’ll walk through the benefits of the new bidding possibilities and explain how advertisers should be thinking about optimizations in the coming weeks and months.

Bye, Felicia — The marriage between desktop and tablet is no more

Oh, happy day!

Ever since Enhanced Campaigns were rolled out in 2013, advertisers have been forced to set the same bids for desktop and tablet devices, as Google saw them as “converging devices” and lumped them together for keyword base bids.

This wasn’t altogether terrible logic in mid-2013, when our company (now Merkle|RKG) put out a Digital Marketing Report that showed tablet revenue per click just eight percent lower than that of desktop. At the time, relatively affluent early iPad adopters dominated tablet traffic share and converted at a rate comparable to what occurred on desktop computers.

However, as cheaper tablet options became available and the demographic of tablet owners shifted, the performance of these two device types diverged. By Q1 2016, non-brand tablet revenue per click was a full 30 percent lower than that of desktop, according to Merkle’s latest Digital Marketing Report (registration required), while cost per click was a mere five percent to 10 percent lower on tablets than desktop. (It’s not totally clear why CPC differed even that much given the identical bids for both.)

rkg-q1-2016-paid-search-mobile-rpc-vs-desktop

Advertisers can now finally set separate bids for desktop and tablet, as Google unmarried the two with its updates, which should result in more efficient ad spend as each device gets the correct bid to reach efficiency targets.

Mobile gets its due

Aside from splitting out desktop and tablet bids, Google also made it possible for advertisers to target campaigns solely to mobile devices, as campaigns can be set to use mobile as the preferred device for base bids with tablet and desktop traffic shut off entirely using -100-percent modifiers.

With the old campaign structure, desktop and tablet devices couldn’t be eliminated from campaigns, and the only “hack” for achieving mobile-only campaigns was to specify very low base bids for desktop/tablet and then multiply them by +300 percent on phones. Certainly not the cleanest of methods.

Even if advertisers don’t wish to shut off targeting to desktop and tablets, the new campaign options allow base bids to be calculated using the most significant data – regardless of which device type that data is tied to.

As phones have grown to account for 44 percent of all Google paid search traffic (for Merkle advertisers as of Q1), many keywords and Google Shopping product groups now see a plurality, or even a majority of all traffic coming from phones.

Since device modifiers can’t be placed at the keyword or product group level, only at the ad group or campaign levels, it makes sense to have the keyword bid targeted to the device type that drives the most traffic, and to then use bid modifiers to adjust for device types with less traffic.

The great campaign structure debate gets new wrinkles

With this new capability to target different device types for base bids, advertisers should probably now consider how much traffic they’re getting from each device type for different keywords (for text ads) and product groups (for Google Shopping campaigns) in assigning them to campaigns.

By assigning keywords and product groups to campaigns based upon the device that drives the most paid search traffic, advertisers can set base bids based on the most statistically significant data.

Additionally, these changes reopen the possibility of having duplicate campaigns to target to different device types. While most advertisers obviously don’t want to undo all of the good that came from Enhanced Campaigns and their elimination of the need for duplication, this might be a sound strategy for advertisers that, for example, see significantly different performance for geo-modifiers on various device types.

By setting up separate versions of a campaign for each device, advertisers can then set separate geo-modifiers for each.

There are a lot of new possibilities, and it will be fun testing in the months to come.

Still not quite the perfect solution

While I hesitate to bring up anything negative, as these changes really are big upgrades over the old campaign requirements, the new way of bidding still doesn’t account for the fact that bid modifiers are placed for separate but dependent variables.

As George Michie wrote in this space about two and a half years ago, the most effective method of bidding would be to set bids for combinations of variables. For example, an advertiser could set specific bids for each keyword-geo-device-audience-etc. combination, without using bid modifiers.

The benefit of this approach is that it removes the problems inherent in stacking modifiers for variables that aren’t independent of one another. For example, the impact device type has on performance might vary depending on the proximity of a user to an advertiser’s brick-and-mortar store and whether that user is included in a remarketing audience.

To understand better, you should go read Michie’s post.

While this setup would allow for the most precise bids, it would also require that Google completely do away with bid modifiers or create a two-tiered system in which more advanced advertisers can take advantage of bidding by variable combination and less advanced advertisers can continue to use modifiers.

In practice, Google AdWords is accessed via a two-tiered system already, in that some paid search managers use the API and others rely on the AdWords UI for management. But Google would likely want to avoid further segmentation.

In addition, the search giant also only just rolled out the new campaign options, so it will probably be a while before we see any big updates again.

Good on ya, Google

It took a while, but Google listened to advertisers clamoring for separate desktop and tablet bids and came through with campaign settings that allow advertisers to do just that. As soon as these changes are rolled out, paid search managers should quickly take advantage of the capabilities to place the most effective bids for each device type.

Advertisers will also finally be able to cleanly target campaigns only to phones, as well as use phones for base bids in campaigns targeting all three device types. This capability should also lead to smarter bidding, as account managers can set campaigns to use the device with the most data for base bids, as well as play around with potentially duplicating campaigns in an effort to set different non-device bid modifiers for each device type.

Despite having to rewrite an entire column, I, for one, am pretty pumped about how we can use the new settings, though the timeline for how quickly these changes will roll out is still murky. Feel free to pre-empt my challenges any time, Google.

The post Well played, Google: Advertisers (finally) get more device bidding controls appeared first on Search Engine Land.