At this point, keyword targeting can seem a little antiquated and probably the most univocal part of account management. I tend to view keywords as more supplemental to a mixed bag of strategies, campaign types, audience targeting, and more. However, while they may seem straightforward on the surface, we must still remain vigilant as match type evolution can cause some insidious happenings within our accounts.
Currently, we have four match types to choose from: exact match, phrase match, broad match, and modified broad match (at least with Google Ads and Microsoft Ads). Match type evolution is more or less old news, with the most recent update being around exact match being “not so exact” anymore. While this initially caused unrest in the industry, I thought it made sense as we should be giving our keywords different triggers to work with anyway (i.e. audiences, various bid modifiers, etc.). Not to mention, keywords are seemingly becoming less emphasized due to the vast new features advertising platforms continue to roll out.
With all of these changes, I was quick to realize that broad match became broad in every sense of the word, and then some. I especially noticed this in my brand campaigns as CTRs, CPAs, CPCs, and more started to trend in the wrong direction. In other words, I had some broad match keywords for my brand name in place that had historically performed fairly well starting to match with general terms within the vertical, competitors, and more. Below is a general example of the types of queries branded broad keywords were triggering.
While these are all fine to bid on, we typically don’t want this traffic overlapping with our brand traffic. Otherwise, it will get super messy and conflict with how we bid and budget for brand campaigns. This can become problematic by spiking impressions and creating the façade that your branded searches are growing exponentially, it can drive up spend, it gives a false sense of where branded impression share is, and the list goes on.
In this post, I’ll discuss how I discovered what these pesky keywords were doing and the steps I took to remedy the situation. Then, I’ll compare the before and after and show you how I lowered brand CPA by 90% all while still increasing my conversion volume.
The Experiment
I first started noticing these unwanted trends in my brand campaign late summer/early fall of last year. Cost per conversion was still fairly competitive and search volume appeared to be growing, however, it was the CTR that first alerted me. The campaign was trending in the realm of 2% and I was really starting to doubt my ability to write ad copy. However, after looking into my search query reports, I quickly realized that we were getting a ton of general traffic that sometimes wasn’t even related to the brand and that the CTRs were horrible.
It was around this time that I made changes to the campaign and quickly noticed improvements. However, for the sake of these comparisons, I wanted to focus on the busy time of the year so traffic, demand, etc. were all peaking, which happened to be spring in the case of this account. Below is a breakdown of how the brand campaign faired in the spring of this year vs the spring of last year.
At first glance, you’ll notice that clicks and impressions appear to be tanking which in most cases would cause immediate panic, especially if it was concerning your brand campaign. Upon closer examination, you’ll see that all other aspects of this campaign improved immensely year over year. The campaign spent less, was more engaging, produced more conversions, and beyond, all with lower CPAs and higher conversion rates. Additionally, last year it appeared as though we owned less than half of our branded impression share whereas this year, we owned 76%.
So, what was the main change? We simply paused broad match keywords. Granted, there were more levers being pulled per usual day to day management, but the clear driver was our keywords. Again, when looking at search query reports, we noticed that the majority of queries this year all contained some element of our brand, whereas last year we saw general terms, competitor terms, and more. This essentially meant we were bidding the same on top-of-funnel terms as we were on brand terms which drove up spend significantly. This also naturally increased CPC, which was $0.60 last year and only $0.19 this year.
Additional Steps
That said, most of the traffic that was coming into our brand campaign last year was still relevant traffic (and traffic we wanted to show for), it just wasn’t organized in the right way. The functionality of match types had outgrown the account and we needed to adjust. So, once we paused those keywords, we also created a new general campaign that targeted the non-brand queries that seemed to resonate well in the brand campaign last year and left those that didn’t out. Here is how the new general campaign performed this year.
Keep in mind we also had other higher-funnel Search campaigns. This data is just a snapshot of those general keywords that we hadn’t built out yet. One of the main data points that stood out with this campaign was the CTR improvement. More specifically, this campaign had a better CTR (3%) this year than the brand campaign last year (2%). This further speaks to the importance of quality account structure so you can ensure you’re serving the most relevant/informative message possible to potential customers.
Conclusion
If I could sum up the key win with this experiment, it comes down to one word: budget.
Last year, the brand campaign was spending an incredible amount which meant we had less to spend in some other areas of the account. However, with this new setup we saved a ton of budget. Even when you combine this year’s brand campaign with this year’s general campaign and compare it to last year’s brand campaign, we still spent 82% less.
Further, we still saw positive performance trends this year versus last. It not only improved the overall account health but also freed up a ton of budget to re-allocate elsewhere in the account. As an example, we were able to spend an additional $35k in Shopping campaigns that directly contributed to 286% more conversions, not to mention the additional budget we could put toward other channels.
In summary, these broad keywords were negatively impacting the account and telling a story about the brand campaign that just wasn’t true. If you’re currently using broad match keywords in your brand campaign, double check that they aren’t doing more harm than good. I’d love to hear about what other folks are seeing in their accounts, so definitely reach out if you have any insights to share!
