Here at Hanapin, we’re always trying to find the best way to do things. If there’s a method that generates the best results, we’ll call it a best practice and teach it to all the Account Managers as the best way to do that thing. One such thing we’ve been working on for a while is account structures. It seems there are a number of different methods that different people in the office swear by, so it was a bit of a heated debate when we finally decided to tackle it. Since we’ve made so much progress on what we think could be the best account structure, I decided to share.

Match Types

One of the biggest factors that influences account structure is how you handle sorting match types. David Rodnitzky does an amazing job of explaining why sorting your match types at the campaign level is important in this preso. I’m going to assume you went over and checked it out, and I don’t need to cover the same material. Like Rodnitzky, I feel that in most cases you really only need Exact and Modified Broad match types. I even wrote an article about it with case studies that pretty much prove it’s all you need, most of the time.

So now that you’ve read all about Rodnitzky’s Alpha Beta structure ideas and my article about how Phrase match isn’t needed most of the time and plain Broad is crazy for most accounts, here’s the summary of where we’re at with this: You want one campaign that has all your keywords for that campaign in modified broad match, and one that has all your keywords in exact match. A vital part of this is making sure you have all your exact match keywords as embedded negative exact match keywords in your modified broad match campaigns. I would also suggest lower bids on the modified broad match keywords, using different ad testing strategies, and using less budget on these campaigns if you’re restricted in your budget at all. 

SKAGs

No, not the crazy monsters from Boarderlands. Rodnitzky uses this acronym to refer to single keyword ad groups. This strategy has some benefits, and some cons. Rodnitzky cites being able to make super targeted ad copy and using precise landing pages as the pros. I’ll cite the cons as how time consuming it would be to do on larger accounts and how many ad groups it would create. Things could get crazy.  For us, we decided this was a good strategy for small accounts and super awesome top performing keywords. I would also say it’s better to have a SKAG than try to throw a keyword into an ad group where the ad copy and landing page might not make total sense. Don’t be lazy! So, while this strategy didn’t make it into our perfect account structure strategy, Rodnitzky’s pros for it are still awesome, so just use your personal judgment to decide if it’s right for your account based on whether those pros outweigh the cons.

Top Performers

One of the times SKAGs make total sense is with top performers. If you did an analysis on your account, I bet you’d find a handful of search queries bring in a larger percentage of your conversions. We call these the “top performers”. One of the first things we look at when taking over an account that isn’t a total disaster is an analysis of the top performers. It’s likely that there are some structure issues that could be dampening their greatness like cross-traffic pollution, ad copy and landing pages that aren’t as targeted as possible, and bids that might not be getting them the most conversions they could get.

Putting top performers into single keyword ad groups allows you to be sure their ads and landing pages are as targeted as possible, helps you pay more attention to their bids, you can give the campaign more budget, you can have the setting for ad delivery on accelerated, and whatever else you weren’t able to do that was limiting the keyword because it was mixed in with a buncha junk. Remember to target these keywords in exact match and add it as a negative exact match to all other campaigns!

Naming Conventions

One of the first things we all agreed on was naming conventions.  We decided the name of a campaign needed to tell anyone looking at it what network it was running on, what the topic/service/product/whatever was, and what match type or targeting strategy it’s using. Sometimes it’s not that cut and dry, like with remarketing. The main point is that people should be able to tell what that campaign is all about without having to dig into the settings and stuff.  You should also stay as consistent as possible, because that will make for easy filtering and sorting when people on the account are looking for certain things like all campaigns dealing with a certain product, etc. Here are some examples:

Search | Branded | Exact

Search | Remarketing | Doormats | Exact

Display | Large Doormats | Placements

Display | Remarketing | Cart Abandoners

Display | Search Companion | Large Doormats

Settings

Some settings aren’t exactly thinkers, like network, bid strategy, etc. You probably have a clear idea of what you’re trying to do and what setting to select for these settings based on the campaign type and what you’re targeting. However, a couple can be trickier. Specifically, ad rotation and ad delivery settings can be a much-debated topic.

We decided that since you should be doing ad testing, you should always pick the evenly rotate ad rotation setting. However, if you forsee yourself forgetting to keep up with ad tests or you’re not ad testing, pick either the one that will optimize after 90 days or one of the optimize options from the get-go. If you have a lot of conversion data and aren’t ad testing, pick the optimize for conversions options. Otherwise, optimize for clicks.

If your campaign is limited by budget, you’ll want to pick the standard ad delivery method. If it’s not limited by budget, you’ll want to pick the accelerated method. However, both methods can perform better or worse with a day parting strategy. If you are limited on budget and pick the standard delivery, you may never really know how your ads can perform at all hours, and it will be difficult to have a statistically sound analysis for that. It may behoove you to save budget somehow elsewhere in the account to test running at full blast, so you could at least have proper day parting data.

Fact or unicorn?

What do you guys think? Does this account structure strategy seem like the best one for most accounts? I’m sure there are a lot of unique instances where there could be a better idea for a specific situation or specific site. I’d love to hear feedback on what has worked best for you across your accounts or agency!