Single Keyword Ad Groups–known as “SKAGs”–seem to be the greatest PPC search strategy since sliced bread, if sliced bread was a PPC strategy. Yes, Virginia, today we’ll talk about search PPC, even though this blog focuses on display.
SKAGs, in short, are what the name says they are: it’s a Google Ads/AdWords strategy that revolves around creating ad groups with one keyword per ad group.
The pro-SKAG argument is, it gives you the most granular control possible. You can include the one phrase you want and exclude all the other phrases (via negatives). While it’s a general best practice to make ad groups specific and targeted, SKAGs take it to the next level. A truism in marketing is that the more targeted the marketing, the more potential the marketing has to resonate (ceteris paribus). Thus, this should help improve your campaigns.
Imagine: instead of a “cheap react.js software dev” ad group where you have lots of synonyms broken out: “affordable react.js software dev” into one and “cheapest react.js software dev in India” into another, and so forth. And then, imagine putting different keywords into separate ad groups–imagine those same phrases, but separate ad groups for the exact, broad, modified broad, and phrase versions–and bam! You don’t have the broad distractions polluting the pristine waters of the exact keywords, and so forth.
So, the three key benefits of SKAGs are:
- More granular keyword control.
- More granular negative keyword control.
- More granular ads as a result.
These three, taken together, let you run with even more granular targeting. That’s a lot of granularity, but granularity in marketing is a good thing!
It sounds like the heavenly Bali of AdWords.
Or is it?
In a recent discussion with a few PPCs, all of us observed that we’ve all come up to the anti-SKAG camp, although all agree there is a time and place. Here are some reasons why:
First, it adds a very serious overhead/management time burden. What you could analyze in one thematic ad group is now spread out over dozens of ad groups, thus increasing the time necessary to manage a campaign. Yes, this could be minimized with automated tools, but those only take you so far. No matter how great your tools are, good campaigns need Tender Love & Care (like everything else in life).
We can’t underestimate the importance of the extra management time it will take to run it, even with software. Ronald Coase won a Nobel Prize for what is called–shockingly–the Coase Theorem. The Coase Theorem boils down to transaction costs matter. We tend to discount them. Yeah, it’s just an extra 2% but once you factor in this 2%, another 3%, another 1%, and all the non-financial transaction costs like time, what once looked very promising suddenly looks barely promising at all.
Second, with the ad groups having only one keyword per ad group for all campaigns it will take a very long time to get enough data for you to analyze that ad group on its own. If every longtail keyword turns into a keyword, then almost all your keywords would get something like 2 impressions per day. And how many months is that before you can do something as simple as see if Ad [A] or Ad [B] is performing better? It will take Q1 through Q3 to start to get a feel on which ads are doing better or not. Not to mention any other aspect of the targeting of that ad group you may be testing out.
The third problem with SKAGs is that there’s no good way to total ad data across ad groups. Let me explain. In practical terms, you’ll have almost-identical ads in almost every ad group within a campaign. If it’s one ad group only, you could see, “Ah, this ad works best!”. But across all ad groups, you’ll want to know information like, “okay, in some ad groups ad [A] works better but in others, ad [B] works better, but which one is working better? What should be my default go-to ad? What are the patterns between seen when it is working vs. when it is not? When ads are so granular, finding these general patterns becomes harder. Nathan points out that you can use ad labels and pivot tables, but those two need more work than looking at the “cheap react.js software dev” ad group.
And the fourth problem is the kicker: the benefits of the extra granularity are often tiny. In a broad sense: yes, the more granularly targeted any marketing is, the better. But there is a point of diminishing returns. And often, in AdWords at least, that point is very early. Think of this demographic targeting: compare targeting “women, 30–35 years old, with 2 kids, and self-employed as a react.js software developer” to targeting “women, 30 years, 7 months, and 1 day old, with 2 kids, and self-employed as a react.js software developer”. Yes, you get more granularity in the latter case but would you be able to improve your end conversion goal by a non-trivial amount by doing so?
The general answer to that is, “It Depends”. When we take into account the first three challenges with SKAGs, the “rule of thumb” I’d suggest would be: “not unless you have large campaigns that are also largely-automated”. So, instead of SKAGs being the secret sauce of your Google Ads campaigns, they can be one weapon you may want to use in that one situation when needed. No use carpet bombing Nagasaki to kill one butterfly there, nor putting the world on lockdown home arrest to stop one stronger strain of the flu. Hypothetically, of course.
Morgan Friedman has been building and running Display campaigns on top of GDN Network of Adwords, err, he means "Google Ads," for almost 15 years. Friedman is, by nature, an obsessive optimizer, and has been A/B testing every obscure option, configuration, strategy, and tactic on Display Ads. Oh and search ads, as well as figuring out how to grow companies and politicians from just the seed to hundreds of thousands of users, or voters, as well. His favorite number is eleven. He enjoys writing about Managed Placements.