Senior Account Manager Elliot talks to us about embracing broad-match key words
So, you want to use pure broad? Or, at least, Google wants you to.
Google Ads’ ever-evolving recommendations section recently started to place an emphasis (and pretty big opti-scores) on replacing broad match modifier keywords within campaigns using smart bidding with *GASP* pure broad match keywords.
Now, rewind to as recently as 6 months ago and the thought of using pure broad match gave most PPCers the heebie-jeebies. Pure broad has always prioritised reach over relevance, with a search query not really having to match up to your keyword for your ad to be triggered.
For example, the keyword ‘black shoes’ could be triggered by the search term ‘black dress’. That’s ok if you sell both, but if you’re a shoe shop then you’re in trouble, especially as the looser reigns of pure broad means that it attracts a lot more traffic and therefore a lot more spend. That’s a recipe for a decimated budget and poor ROI right there.
However, in an age of smart bidding where the focus is more on query intent, we’re now being advised that the coast is clear – pure broad is safe to use, just as long as you couple it up with a smart bidding strategy.
Of course, there was, and still is, a lot of skepticism around this claim. Yes, smart bidding has been revolutionary for PPC – I don’t have a single account that doesn’t utilise it – but using pure broad would still take a lot of testing and a lot of convincing. Lucky for us Google stepped in and forced our hand by not only dangling a carrot in the form of big opti-scores for adopting pure broad, but also announcing the end of broad match modifier (RIP BMM).
After a few days of contemplation, and seeing pure broad recommendations in EVERY account, my curiosity got the better of me. I set up some pure broad keyword experiments in a couple of campaigns running on maximise conversions in a lead-gen account. I informed the rest of the team I’d pressed ahead, to which I received this image in response:
That’s me, Pure Broad Braveheart
Anyway, the juicy stuff, how did it perform? It’s been 6 weeks (at time of writing) since I set up my first pure broad experiment, splitting traffic 50/50 between pure broad and modified broad match. Here are the results:
Not bad, hey? As expected pure broad has more impressions and a lower clickthrough rate on both campaigns as it’s eligible for way more searches. The important metrics, though, are in the conversion columns and pure broad wins nearly every time. Where it fell down on conversion rate in campaign 2 it more than made up for with a higher volume of conversions and lower CPA.
DISCLAIMER: I’m not going to pretend that pure broad with smart bidding works every time. We are running several experiments across different accounts, in different sectors, and using different bid strategies – the results so far are mixed. In some cases the experiments are on par with the above example, in other cases there’s not much in it and in a small handful of cases the experiment really hasn’t taken off at all.
FURTHER DISCLAIMER: None of these experiments have run for as long as the above example so there is plenty of time for the strategies to learn and turn things around. 2 weeks into the initial experiment I was having second thoughts as the pure broad campaigns were nosediving. 4 weeks later they’re by far the better performers.
This is new to all of us and it very much feels against the grain of PPC best practice, but the only way to learn how to work in this way is to get stuck in.
However, for those of you who are still curious or nervous about using pure broad, I’ve listed some tips below on how best to embrace it and become a pure broad ‘Braveheart’.
Think About Your Goal & Speak To Your Client First
How strict is your CPA target? Are you after sales/leads volume or a good ROI/CPA? Feedback from clients suggests that lead quality will dip to begin with when running pure broad, which doesn’t come as a surprise. A few STRs and a number of negatives later and the lead quality returns to normal, but it’s best to speak to your client first to pre-warn them. The same goes for e-commerce – expect a higher CPA/lower ROI to begin with.
We’ve found that maximise conversions has by far been the best strategy to utilise pure broad with thus far, most likely as you are suppressing budget so ending up with more conversions for your spend. More open-ended bid strategies like Target CPA and Target ROAS, where spend tends to be more uncapped, has had some questionable results so take care if using this strategy.
Run An Experiment
Don’t accept the changes in recommendations straight off the bat – not only will this make analysis a little more difficult, it will also DELETE all of your BMM keywords and replace them with pure broad. Instead, set up a 50/50 traffic split experiment. Again, you can do this within the recommendations.
Give It Time
I’ve already touched on this, but it’s important to give as much time as possible to the experiment to get a more accurate set of results. This depends on budget but I would recommend at least 6 weeks so you can fully overcome the 2 week dip that I experienced.
Negatives, Negatives, Negatives!
An important one. Every PPCer’s initial fears about using pure broad is the sheer amount of search term reports you’d need to carry out in order to add negative keywords and reduce wasted clicks. So far I’ve found that you need to be hot on STRs for the first week, checking in daily, but after that the necessity for negatives begins to diminish. I’m still running STRs more often than usual but I’m not overwhelmed by them. Something to note is that there seems to be far more competitor terms appearing in STRs due to them matching intent. Depending on your competitor bidding strategy you might want to keep your eyes peeled for these.
Audit Your Ad Copy
You will get unwanted clicks from using pure broad, so other than negatives, how else can you limit them? Specific ad copy, that’s how. Include key information such as pricing, location, delivery etc to try and stave off irrelevant or unqualified clicks.
Focus On Conversion Metrics
When analysing performance it’s easy to get caught up in click data. Impressions will go through the roof and your CTR will fall when using pure broad. However, the proof is in the pudding, or the conversion columns in this case. Obviously don’t completely ignore click data, but concentrate more on the conversion columns to judge just how well your experiment is going.