By Remy on Friday, March 22, 2013 -
3 weeks ago, our newest and largest spending AdWords account for a gambling product was performing particularly well, so we decided to push for more conversions by testing new keywords.
We chose the new keywords based on search query reports and also by looking for any keywords we were missing from other accounts in the same vertical.
In one of the accounts (same client and product but different brand) we found a seemingly innocuous keyword; a competitors brand term in exact match that had picked up 2 conversions at a decent CPA - so we thought we may as well try this keyword in our new account. So we added the keyword in all four match types (we always try every match type as most search marketers will know – the same keyword will perform very differently in a different match type and some broad and broad modified keywords can perform extremely well).
So why write a blog post about it?
Well the exact, phrase and broad modified match types generated limited traffic and clicks but the broad match keyword picked up over 500,000 impressions and more than 2,500 clicks in less than a month. To put this in context; 500,000 impressions represents 3 times the number of our Top Keywords campaign and 2,500 clicks represents half of all our non-brand search traffic.
Now some of you may start having bad memories of PPC newbie mistakes – you added the wrong keyword in broad match or forgot to check the campaign settings before activating and saw your CPA skyrocket.
Then everyone laughed at you…“Go and boil your bottoms, you sons of silly persons!”
Well, it didn’t turn out like that.
Firstly, here at Jack we do our job properly, so all new ad groups are labelled and automated rules are set up to send notifications to us by email for all ad groups labelled as “new” that spend more than a set amount per day. After receiving an email for this particular ad group we investigated it. It was spending a lot. It had picked up 72 conversions in the first week, a staggering 33% of all non-brand conversions at a CPA below our target for non-brand campaigns (once again a very challenging CPA target!).
We had found the search marketers Holy Grail keyword!
So what happened?
This keyword despite being a competitor’s brand term (always important to bid on competitors when you are advertising a gaming product due to the poor customer brand loyalty in the gambling vertical) also happens, like many competitors brands, to be a very generic sounding term. Google totally ignores the second word which forms the brand name and was serving this keyword for an incredible amount of other competitor brands for which we already had dedicated ad groups.
Was it phagocytising the rest of our campaign and Top Keywords campaign? Thus taking conversions away from them?
It wasn’t. So we ran a couple of reports and found that our other campaigns & ad groups had kept the same traffic, average number of conversions and spend week on week, so these extra incremental conversions were the result of new traffic not traffic taken away from other ad groups.
Then why did it perform so well?
Our competitor’s campaign is performing very well and has one of the best CPAs for our Non-Brand generic campaigns, but due to the fact that we are bidding on other marketers brands our quality score is obviously very low at around 2. Such a low quality score means, firstly, that you will have to bid higher to compete in the auction and, secondly, that some your ads are rarely shown due to your low quality score.
This keyword despite being a branded term had a far higher quality score because it sounded generic. This meant that for a much lower CPC we were able to appear higher in the rankings and that when other lower quality keywords were not appearing this one was! The end result - we were not impeding on impressions from of other ad groups – it was brand-new traffic!
In red: the increase in clicks and conversions following the launch of the new ad group.
Normally we would try to direct the right traffic to the right ad group but in this case following best practice would have meant that we wouldn’t have generated all these new conversions. So we left it as it is – and are monitoring it closely as you would expect for a keywords that spends more than £4,000 per week. We even tested deleting all the ad group negatives for this ad group, which resulted in even more conversions, although CPA did increase slightly.
We always try to follow best practice guides and tell our clients to follow stringent principles, which are key for the success of any search strategy. However, at the end of the day we need to focus on what’s working, even if it is messy and against rule book. Sometimes it’s best to stick to what works even if this goes against best practice.
Thanks to this keyword we are now spending an extra £4,000 per week and have increased our clients non-brand conversions by 30% at a below target CPA.
In the age of old fight between common sense and best practice – here common sense is winning.
By Remy on Monday, March 18, 2013
By Andy on Wednesday, March 06, 2013
By Andy on Friday, February 22, 2013
By Sergi on Friday, February 15, 2013