I was trying to not write this post, but I feel like I have to. If you have read any of my other posts you already know that I feel passionately about making sure the playing field is as level as possible for businesses using AdWords, particularly those who are not professionally managed. So with that in mind, I have to dissect the latest piece touting the magical powers of broad match terms as conversion machines. Let’s dive right in…
From the article, the methodology used is described thusly (note – all sections highlighted in orange are from the original article):
Why is this problematic?
Using the definition of a “long-tail query” as something that appeared a maximum of 10 times in a 28 day period sets up a bad data point right from the start. Anyone who has been doing PPC for a while knows that you need more than a couple of conversions to determine the ultimate viability of a term. A couple of conversions does not automatically a great keyword make – it takes more than that.
Study only includes advertisers spending more than 20% on broad match. By definition they are going to have some conversions from broad match, but are they the most efficient conversions? There does not seem to be analysis of cost of these conversions specifically or relatively when viewed against the traffic from irrelevant and/or non-converting long tail queries.
This one is huge – stats come from advertisers that are tracking conversions in AdWords. This alone implies a certain level of sophistication of these advertisers. Adwords themselves rolled out a huge product, SmartGoals, aimed at all those poor souls who can’t or won’t implement conversion tracking (read my take on Smart Goals here). By only tracking data from those who do use conversion tracking, the results are already skewed, in my opinion. Also, what was defined as a conversion in these cases? Because I am sure we have all seen conversion tracking set up to track things that are not actually what pros would call a conversion. This matters quite a bit too.
The article goes on to extoll the awesomeness of broad match:
Twice as many as what? How much in total do they spend on long tail queries that are clicked on that do not convert? How does this stack up against using broad match modified or other match types designed to capture more established long tail queries? Also, do long tail queries drive more conversions in general for these brands or if you use broad match do you get more long tail query conversions than accounts that do not, because those are two VERY DIFFERENT THINGS.
Except that you can’t know that if you had other match types actually set up to capture these types of terms that they would not perform at least as well, if not better. My guess is that they would perform better with other match types because you would also likely have better targeted ad copy, perhaps a different landing page version, etc. You’d probably also pay less for those clicks as your quality score would likely be better too.
Also, when you have terms that could match for a more specific term or a broad match term, there is no guarantee that the broad terms won’t get traffic that you would prefer go to broad match modified or phrase terms. We have all seen queries trigger ads from campaigns or ad groups that are not what we wanted or designed, even when you think you have things set up to not do this.
Gotcha, broad match is not about performance. So if you want tons of impressions and some associated clicks and a few conversions that can serendipitously come from them, knock yourself out with broad match? Is that the message? And what about broad match modified or phrase match? With the every match type now including “close variants” you might be surprised what you’re already capturing without a lot or any broad match.
If you care about performance and actual targeting, well….
I cannot disagree more strongly about broad match allowing you to be efficient in campaign management. Broad match requires THE MOST CLOSE MANAGEMENT OF ANY MATCH TYPE. It is in no way a “set it and forget it” way of running quality AdWords campaigns. And again, why no talk about broad match modified which is a much better option for non-professionally managed accounts in almost every instance.
No matter which match types you use, actively and very regularly viewing Search Term Reports is vitally important. If you are using broad match it is critical that this be done very, very regularly and that negative keywords are also added regularly. Again, pros know this, but I doubt businesses who manage their own accounts really get this the way they should.
I am not a huge fan of automated bidding for most accounts. I also think suggesting automated bidding for non-pros is a slippery slope. To use it properly you need to be able to have a logic about your bidding strategy that requires a deeper understanding of how all this stuff works. I’m not sure that a feature that has a 14 page PDF explaining how it works is very entry level or non-PPC pro user friendly… And, I have been in countless client meetings where when the question is asked about what their current cost per acquisition (CPA) is, you can hear the crickets as everyone looks around the room hoping someone has an answer. I’m not sure how many nonprofessionally managed accounts can even really wrap their heads around the bidding?
Retargeting those who came to your site via a broad matched query that has never converted is something that you could try, but I would exercise caution here too, especially for nonprofessionals. I’ve seen too many clients get enamored with big impression numbers and lose focus on getting more people to their sites that convert the first time. Not every long tail query is worth chasing. You have to think holistically and measure effort, cost and results to decide where to focus your PPC resources.
And I would also note, you have to watch your other campaigns and ad groups more closely too to make sure that your broad match terms are not stealing traffic from other more targeted places in your account and manage accordingly.
I come from the world of SEO way back when, so I am intimately familiar with long tail queries and terms. Although in the SEO world, long tail has a slightly different definition referring more the the length or specificity of a query rather than the frequency of its use. There can be some real gems that fall into either long tail category that can turn into strong content and conversion streams. But, there are also an awful lot of long tail queries that are best left alone. It takes skill and experience to be able to differentiate between the two. I’m not saying to never use broad match. I think it can have its place in accounts, but it can also be a resource hog – both for actual dollars and management time.
My fear in articles like this one is not how it will be received or internalized by PPC professionals, but rather what happens when businesses and organizations who are not using professional management for their PPC efforts read this and do everything that is recommended because “the AdWords guy said to do this and we will get more conversions”. That is the part that creates difficulty for me as a professional.
What do you think? Love to hear your thoughts on the topic. Sound off in the comments or hit me up on Twitter (@NeptuneMoon).