One common question I’m often asked is: how do negative keywords in Adwords / Google Ads interact with display / programmatic targeting?
The very short version of the answer is: in the most annoying way imaginable–so use with care!
Let’s backtrack, because to understand how this functionality got annoying–and what it even is–we need to understand where it came from.
Long, long ago, Adwords/Google Ads only had search targeting, including you putting in the keyword that would trigger and ad, and “negative keywords” (which would ensure that an ad is not triggered should the search term the person uses include those words.) This works fine and, fundamentally, that system is still in place.
Then, Google added in display/programmatic campaigns–and just used the same targeting options! I remember when I first did display/programmatic ads via Google, it was just the identical targeting as in search: just put in keywords and negatives, and voilà! Since there are “search terms” on display campaigns, the keywords were basically just words used on the page. And negative keywords, unsurprisingly, were just words that, if used on the page, would ensure that your ad was not displayed.
Over time, that evolved into the keywords not just being the exact words on the page, but more conceptual. It could be a page about that topic. Like everything was a broad keyword. As the positive targeting changed, however, the negative targeting remained unchanged: any use of that negative on the page, would disqualify your ad from being triggered.
Then, over the last decade, Google has been slowing making a transition to a different type of targeting: more “audience” based. You define an audience, with a bunch of characteristics, and voila, those are the types of people who are targeted (or so they say). This was a separate type of targeting, and in addition to the display keyword targeting methods discussed in the last paragraph; they coexisted side by side. But the negative keywords? That functionality stayed the same: if that word appears on a page, the page won’t be shown.
More recently, Google has been expanding the audience targeting and increasingly de-emphasizing the original keyword targeting style. If I were to place money and make a bet, the trend continues and the original keyword targeting disappears completely. It’s like a legacy feature that is still there just because no one prioritized taking it off the interface just yet.
But the negative keywords? Still there, and still exactly like it was in the original version eons ago: you use any word, and that whole page is disqualified from your ad being shown on it!
As a result of all this, the most critical fact about negative keywords in display and programmatic campaigns to keep in mind is this: use negative keywords on display and programmatic campaigns with a lot of care, because, well, if any page has any of the keywords on it, then it won’t be shown.
A secondary caution to keep in mind is to remember how broad the broad keyword type is–even in the negative incarnation. So if you have a negative keyword “New York” because you don’t want it listed on any pages about New York, then remember that, if the keyword type is “Broad,” any page with the word “new” on it would be disqualified.
This is why I use my own personal best practice: I only use “Phrase” for negative keywords, because, in the above example, it would exclude “New York” but not “new.” (And note that “Exact” is meaningless in a display context, although you can still use it. I would assume it is treated as a phrase, but you should test that yourself.)
But what’s going to happen with negative keywords? Who knows, anything is possible! But here’s my guess: along with keywords, they’ll be slowly deprecated.
But they do serve a useful function. Sometimes, if you are advertising “java” services, you want no page that mentions “coffee” or an “island” anywhere on the page! Google is counting on its own AI doing a better and better job of the targeting, so taht negative keywords will be less needed.
That’s what Google wants, and is pushing. The reality? There AI’s reality is far behind their AI’s marketing. So the best we can do is hope that the negative keywords aren’t deprecated until there is an adequate replacement.
And the adequate replacement would seem to be audiences, but used as negatives–to define audiences specifically to exclude. You can do that to some degree, such as excluding audience lists of customers. But many audiences aren’t allowed to be used as negatives. (It’s even worse on Facebook, where you can target by, things like “Interest in African American culture” but you can’t exclude anyone based on it; I’m guessing they want to prevent people from not targeting certain groups due to racial preferences.)
In conclusion, this is more of the general pattern: increasingly limited options. The result? The options we are given, we need to use with even more sophistication than we did before.
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.