A big part of our job as search marketers is to review the search terms that trigger our ads, and to act on that data. Typically, that means you do some or all of the following:
- Identify high-performing search terms and add them as keywords to your account. This allows you to write relevant ad copy, control your CPC, and ensure the best landing page experience.
- Identify “search term bleed” -- search terms that are triggering ads from multiple keywords, ad groups, or campaigns -- and sculpt the traffic to ensure certain search terms only trigger ads from specific ad groups and keywords.
- Identify low-performing search terms and add them as negatives to your account. This helps reduce the cost of irrelevant, wasteful clicks that are unlikely to convert.
This article focuses on the third use case: analyzing search terms to identify potential negative keywords.
Here’s what Google has to say about that: “If a search term isn't relevant enough to the products or services you offer, add it as a negative keyword instead. This will help keep your ad from showing to people who are looking for something you don’t sell. For example, if you sell eyeglasses, and you see that the search term “wine glasses” is triggering your ads, you might want to add “wine” as a negative keyword.”
The way this process typically works in the real world, is that you look through your search terms for a given time period, and identify the irrelevant terms. “Irrelevant” usually means lots of clicks and/or high cost with few conversions and/or high cost per conversion / low ROI. Most people then exclude that term at the ad group or campaign level from within the AdWords interface. You may also export them to Excel, adjust their match types and locations, and upload them to your account via AdWords Editor. And some people even create shared negative lists that they apply to multiple campaigns or account-wide through the Shared Library.
All good, right?
There is actually a huge blind spot that a lot of advertisers don’t understand.
In Google’s own words: “You'll see data on which search terms triggered impressions and clicks...you’ll only see search terms that were used by people at least 8 hours ago and have either received clicks in the past 30 days or were searched for by a significant number of people. Any search terms that did not meet this criteria will be summed up in the 'Other search terms' row.”
In plain English, what this says is that search terms that don’t get clicked on, probably won’t appear in your search term reports.
Google wants our ads to be as relevant as possible, right? That’s why they assign a Quality Score to each keyword: so we know which are seen as less relevant and worthy of optimization.
But if my ad is irrelevant for a specific search term, it’s unlikely that many people will click on it. If no one’s clicking on it, that search term probably won’t surface in my search term reports.
Sure, I’ll be able to add good search terms as keywords, and will be able to negate non-converting search terms that have clicks. But we’re blind to all the search terms that never lead to clicks.
Some will say: we don’t pay for unclicked impressions, so who cares? But I don’t want my ad showing if it’s not a relevant match, and Google says they don’t either.
In order to exclude the truly irrelevant search terms - the ones that don’t lead to clicks -- we need to see those terms too.
How big of a problem is this, really?
Pretty big, I think.
Here’s a real-world example from one of our clients:
In the last 30 days, about 31,000 different search terms are exposed in this search term report, representing about 1.6 million impressions.
But look at that ugly “Other search terms” row. Almost as many impressions there (1.2 million), and relatively few clicks.
The visible search terms had a 3.8% CTR, while the invisible search terms had a 0.09% CTR.
These are obviously very irrelevant search terms that, quite frankly, we’d love to see so we can make a strategic call: negate them so they don’t keep dragging down our CTR (and presumably, Quality Score), or add them as keywords and figure out what sort of ad copy would get clicks.
Google, if you’re listening...how about doing something about this?
Search marketers...anyone know a good hack to work around this problem?