Ever since device bidding became uniquely available (sometime between 2005 and 2011), I’ve been an ardent advocate of separating PPC campaigns by device. (Well, except for tablets, because no one cares about them anymore.)
For starters, there is (at least) a 50% chance that you are reading this article on a phone. And you probably know that Internet and search consumption is dominated by mobile platforms.
Case in point: Looking at an eCommerce brand that literally sells bacon online, I compared September 2022 website traffic to September 2009 by device and found that:
- In 2009, mobile devices accounted for less than 1% and desktop computers accounted for 99%.
- In 2022, mobile accounts for 49% and desktop for 46% (total visitor growth increased by 504%).
So yes, mobile is a thing and should be treated as such.
Why it is important to do PPC by device
As mobile Internet usage has grown over the past 10 years, so has it from a search market share perspective.
So much so that most verticals (not brands, but complete verticals) you’ll find that mobile drives the majority of traffic and spend when it comes to search. (Not you, B2B, you’re still good with the old desktop.)
This is incredibly evident in consumer goods, brick-and-mortar locations, and brands with low price points and/or short buying cycles.
Before 2017, you would even see up to 50% lower CPC on mobile than desktop thanks to less competition.
In some scenarios, you still see that. But the volume coming from mobile was so high that it would offset the cost savings on the CPC side.
These days (ignoring Pmax), the level of discount on mobile is lower, while the volume coming from mobile has largely outpaced and outpaced that from desktop.
In the case of a Caribbean QSR brand we work with, mobile CPC is around 32%. But mobile drives 17 times more impressions and clicks than desktop, leading mobile to spend 11 times more than desktop.
Next, we look at the environment of users looking at mobile devices. Consumer behavior is different than desktop.
Mobile devices trigger immediate engagement (ie “Hurry up and get what you need”).
On the other hand, the desk implies a seated experience, where people research and compare while taking their time. (I should note that this mindset has completely changed in the last 10 years, largely due to consumer wallets.)
Because of this, the contribution of mobile devices to e-commerce and direct response campaigns has increased dramatically.
Gone are the days when you didn’t want to pull out your credit card and punch it into your phone while out in public.
As we approach a potential recession, the demand to ramp up any performance effort skyrockets.
It is clearly evident that the mobile must be isolated from the desktop due to two big factors:
(Sorry tablet, but with its market share shrinking since 2019, it’s bundled with desktop in my opinion.)
Get the daily newsletter marketers trust.
Won’t search engines allocate my spend and optimize for performance?
Technically, they can optimize and allocate spend to a device.
But, if you trust the engines to do what’s best for you, then you’re more doomed than any New York Jets quarterback since 2007.
Think of it this way:
An engine will bid up or down per device, when a bidding strategy is implemented, to achieve the best possible result for the rule.
You can also retain budget from tablets or desktops to spend on mobile devices. but that can No be in your best interest.
Let’s look at a working scenario that I tested in a lead generation campaign using a max-click bid strategy.
We were too confident in the site and the creative, so we realized that optimizing for more traffic would lead to more leads than specifically optimizing for more conversions.
All consolidated devices:
- $10,000 total expense
- $3,000 desk
- $6,750 mobile
- $250 tablet
- 4,150 total clicks ($2.41 CPC)
- 1,150 desktop ($2.61 CPC)
- 2,900 mobile ($2.32 CPC)
- 100 tablets ($2.50 CPC)
- 850 leads ($11.64 CPL / 20.5% lead rate)
- Desktop 403 ($7.44 CPL / 35% interest rate)
- 414 Mobile ($16.30 CPL / 14% interest rate)
- 33 tablets ($7.57 CPL/33% lead rate)
Then we separate the devices into two groups:
- Desktop and tablet together.
- Mobile by itself.
We split the budget to be 65% mobile and 35% desktop and tablet:
- $10,000 total expense
- $3,750 desktop + tablet
- $6,250 mobile
- 4,400 total clicks ($2.27 CPC)
- 1,426 desktop + tablet ($2.63 CPC)
- 2,974 mobile ($2.10 CPC)
- 945 leads ($10.58 CPL / 21.4% lead rate)
- 499 desktop+tablet ($7.51 CPL / 35% interest rate)
- 446 Mobile ($14.01 CPL / 15% interest rate)
Both scenarios have the same bidding strategy.
The consolidated version was the last four weeks of the campaigns, while the separated version was the last four weeks of a six-week trial. (We discard the first two weeks for the bid strategy to learn correctly.)
Both time periods were not in high season.
The reason the budget for desktops and tablets was set at 37.5% instead of 32.5% (level of spending when consolidated) was that it was theorized that there was cannibalization of devices in the mobile budget. Therefore, we choose to give you an additional $500 to test that belief.
The result proved a number of theories:
- In fact, mobile was cannibalizing the desktop, leading to fewer desktop clicks and therefore conversions.
- Mobile CPC likely drifted higher because the allocation of the engines to desktops and tablets didn’t give them their real chance.
- Not allowing devices to impact each other in bid strategy or budget allowed for higher click volume and therefore leads (lead rate was largely unchanged), but this lowered cost per lead .
This also illustrates the importance of human supervision and intervention in bidding strategies.
While not used in this test, there are studies that have shown the benefit of citing the mobile experience in a mobile creative (i.e. “Sign up from your phone”) to improve CTR and lower CPC.
How will this work on a low budget?
Here lies the real decision of the advertiser. This approach is feasible if:
- You have a substantial budget to work with so no campaign is considered restricted. (This depends on your account structure/number of campaigns and how much daily budget you have.)
- And you already know the demand and spend per device.
Who shouldn’t do this?
There are arguments for not doing this approach, the most common are:
- Smaller budgets, the reality is that this may not be possible without causing a campaign to be budgeted incorrectly.
- If you’re happy with your current performance in device consolidation, keep doing what you’re doing.
- Lack of resources to manage more campaigns.
This isn’t for everyone, but never dismiss the concept of device segmentation until you’ve reviewed the data.
Why is device segmentation more important than SKAG?
Single keyword ad groups (or SKAG) have been a common model For many years in the search industry, it has been used largely for exact and few-phrase ad groups. You needed a strong negative game to do it.
But the fact is that the SKAGs are now heavily biased by the “relaxation of party rates” in recent years.
The amount of time and effort to put together a strong SKAG ad group is likely to yield less return than a solid device targeting build.
Blame the engines for that.
The opinions expressed in this article are those of the guest author and not necessarily those of Search Engine Land. Staff authors are listed here.
New to Search Engine Land