This is the fifth post in our o360 Series on 360 Degree Optimisation; using SEO, SEM, UX Optimisation, Content Optimisation, Customer Journey Optimisation and Conversion Rate Optimisation to enhance the user experience to not only garner positive results for our clients, but also foster a positive experience for users that they will want to return to. In this post we highlight some simple but very effective optimisation techniques to help improve the performance of your PPC account.
The day-to-day life of a PPC professional can be heavily weighted towards reporting, which is why it is very important to set aside time for testing and optimisation because ultimately, it is very unlikely that performance will improve unless regular tweaks and improvements are made to the overall PPC structure. Throughout this post I am going to outline 5 simple optimisation techniques to help maximise your PPC efforts.
Probably the simplest of all optimisations but potentially also the most effective. A crucial component of any PPC account set up is the negative keyword list. Open a search query report from the keywords tab (Details -> Auction Insights) and view the data for a time period that is large enough to outline some trends, usually at least two weeks. Next, filter the data by impressions, highest to lowest, and you will see the search terms that are triggering your ads most frequently.
Well, any PPC account will contain many keywords whose match type is either broad or broad match modifier which we know will trigger your ads for a wide range of keywords with only a potentially vague relevance to the actual product or service you are offering.
A good example would be the car brand Mini Cooper, if you were to include the keyword +mini in your campaign structure without an extensive negative keyword list, your ad could potentially appear for searches of ipad mini, ladies mini marathon, mini golf ect. resulting in huge amounts of impressions and a very poor CTR which in turn will damage the campaigns Quality Score.
If you are systematically pulling search query reports on a bi-weekly basis and using it as your reference to exclude any irrelevant or unnecessary terms it can deliver a lot of cost efficiencies to the campaign and improve overall CTR.
This is an optimisation that can have unexpected positive results on an account. I recommend downloading a keyword report, including the Quality Score column and then filtering the data file by separating keywords with a score of 8 and above and 7 and below.
The terms with a Quality Score of 8 and above are likely to be brand terms or some high performing generic terms, in some instances there will be an opportunity to turn an 8 into a 9 here but for the most part these keywords are in good shape.
It is the keywords with a Quality Score of 7 or below that you want to focus on. It is likely that they are not appearing at the top of the results page. The easy option would be to increase bids which would in turn increase your CPC’s for that term but before you do that consider the below scenario.
You are bidding €2 and still you are only in average position 2.6. One option you have is to raise your bid, for the purpose of the example I am going to double my bid to €4 which pushes you up to the more desirable average position of 1.4 BUT as you have doubled your bids you can expect to see your CPC increase significantly.
Another good optimisation to make at a keyword level is to look at actions per keyword. I say actions rather than conversions because I also want to look at keywords that might be assisting conversions.
Download a keyword report that along with regular metrics includes conversions and click assisted conversions and then create a new column adding these two metrics and label it actions. This column now identifies any keyword that has either converted itself or has assisted a conversion.
An example of an assist is if a user has visited a site on a generic term, not converted but later returned on a brand term and converted. The brand term gets the conversion but the generic term also played a role and so is attributed with an assist.
Going back to our spread sheet, calculate your cost per action for each and then sort them low to high. The keywords at the top of your list are going to be the ones that are the top performers in the account, these keywords convert at a very efficient level. Now the idea is to go down the list of keywords and find terms that are spending heavily but are neither assisting nor converting. These keywords are doing nothing for performance and are inflating the overall cost per sale.
Pause these keywords and analyse your results after two weeks, the likely result is that the levels of conversions have remained stable but by stripping out some non-performing keywords you have managed to reduce the overall CPA.
Again, quite an obvious optimisation but also an often overlooked one. The graph below represents actual performance for an account I previously worked on after two weeks of activity.
We can see that while users are just as likely to search any day of the week but are far more likely to convert over the weekend. In this instance it was a no-brainer to allocate the majority of the budget and to bid higher on weekends.
To discover your results by day download a report over a longer period of time, ideally minimum 3 weeks. Segment the report by day of the week and using a pivot table filter your results to look at clicks, conversions and conversion rate by day of the week. The results of this analysis will clearly outline what days of the week you need to focus your PPC budget on.
A final optimisation to look at is to analyse traffic and sales by device. We all know search volume and user experiences are improving on mobile but still for certain products or services or for clients whose website offers a poor user experience, conversion rates on mobile lag behind those on desktop and tablet. Again, consider the below graph, this example is from a client who did not have a responsive site.
Early on we detected that while there was high volume of searches coming from mobile (one in five) only six percent of our overall conversions were being delivered through mobile. The reason was twofold – firstly the clients mobile site was not responsive and secondly the conversion form was quite lengthy and not the type of action you would like to complete on your phone.
As a workaround to this and in an effort to not lose the 20% of potential customers who were searching on mobile we implemented call only ads across mobile (scheduled to be active only during the hours that call centres were open) whereby rather than being directed to the clients mobile site the user would activate a phone call to a customer service representative who could facilitate the sale over the phone.
So there you have it. I find that neglected Adwords accounts will show a steady week on week decline in performance but implement these five simple optimisations and it will be no coincidence when your PPC performance improves considerably.