When it comes to managing AdWords, something I’ve come across time and again is that lots of accounts are oversegmented. I’ve seen AdWords accounts with almost awe-inspiring intricacy. No dimension unsegmented; no campaign setting untweaked; no minute negative keyword unadded. I think we’ve officially reached the point at which some of you are too good at your jobs.
Automation is better than it’s ever been. It’s so good, in fact, that it often beats oversegmented, even overengineered accounts. Campaigns can be segmented by devices, match types, audiences, geographies and more. Campaigns shouldn’t be segmented by all of those.
AdWords Smart Bidding looks at specific queries (and the context of those queries)
Overly segmented account structures are attempting to approximate something that Smart Bidding already does: bid to a user’s specific search query and adjust bids for devices, time of day and audiences to control the impact on the advertiser’s objective. AdWords Smart Bidding considers dozens of additional signals and the combinations of them, like mobile devices at nighttime in a specific area.
You don’t need to manually define each segment’s value if you’re accurately tracking your conversions in AdWords. Tell Smart Bidding what your end goal is, then track performance. You can stop using CPC bidding as a proxy for value; Smart Bidding can boil everything down to what really matters for you, whether that’s a CPA or ROAS goal.
A simpler proposal
Your default campaign structure should be a lot more straightforward. This may sound insane, but here’s how I think campaigns should be organized:
- Organize your ad groups around what ads you want to serve to groups of users.
- Organize your campaigns around your objective and KPI.
Some aspects of campaign setup warrant separate campaigns — such as budgetary control and the countries/territories where you can actually sell your goods and services. There is no longer a need for additional campaigns to work around long-gone AdWords limitations surrounding bidding and messaging. Bidding has Smart Bidding. Ad text has ad customizers. Audience targeting benefits from both Smart Bidding and ad customizers.
One thing in particular that I want to highlight is separate campaigns or ad groups by match type. The AdWords system is set up to prefer the more specific keyword, and in those rare cases when it doesn’t, it’s to your benefit. A less specific keyword will trigger only if you’re projected to have a higher Ad Rank and a lower CPC. (You might even consider de-duplicating your match types of the same keyword in your account. That’s too big a topic to cover here.) Ultimately, what may seem like sloppy structure is actually saving you money.
If you have a set of high-performing keywords that deserve their own budget, you should break those out. That’s a case where it makes sense to make such a management decision. But that should be for your best stuff. Let performance dictate what gets priority.
The benefits of aggregated campaigns
An overly segmented AdWords campaign structure can actually be a serious barrier to performance.
1. Automation works better on large sets of information.
AdWords’ Smart Bidding can work on pretty paltry data. But it works even better when it has large amounts of insight to feed into its machine learning. Larger campaigns, including data across all different types of cross-device, cross-user-list, cross-time insights, tend to perform better. A bigger campaign is actually more likely to perform better when you fully embrace automation.
2. There are fewer ads to maintain.
Keeping up with your ads is a lot of work. The smarter you are about the amount of work you create for yourself, the more your work time is spent on finding and deploying great messaging. And the less time you have to spend making sure that you’ve copied/pasted your ads across all nine device-specific campaigns that advertise women’s tankinis to previous customers who reside in New England.
Some opinions expressed in this article may be those of a guest author and not necessarily Marketing Land. Staff authors are listed here.