This is an important step that requires some planning and consideration ahead of your transition in to automated bidding strategies. Due to automation being a relatively new concept for many native digital marketers, it can be easy to find yourself analysing the smart bidding performance incorrectly.
One of the major pitfalls advertisers may enter is analysing their smart bidding performance too soon. By too soon, we’re talking before the algorithm has had time to learn and work towards your goal.
Google say there are two main factors that you should wait for before analysing your smart bidding campaigns, which are:
This is the time the Google Ads algorithm needs to gather data on the account and learn how to optimise bids based on this. The more data the account has the quicker this learning period can be; however, it is typically four weeks.
Depending on the industry some clicks don’t convert instantly so, this needs to be taken into consideration. For example, if your clicks don’t typically convert before 30 days, it makes sense to wait till those 30 days before analysing whether smart bidding has performed in line with expectations.
Whilst it’s important to consider all metrics when analysing the performance of an account, another hole advertisers may find themselves in is putting too much focus on the wrong metrics. The focus with smart bidding is typically:
Therefore, Google advise these should be your main KPIs rather than CPCs, impressions and CTR. When setting these KPIs, be realistic; although the Google Ads algorithm has advanced over the last year, it uses real data to provide your results.
If you have a nervousness around making the big switch in bidding, try setting up an A/B experiment where you can split your campaign traffic and spend and limit your chances that something might go wrong
This experiment feature allows you to test and measure your new bidding strategy side by side against your manual bidding campaign for a direct comparison of performance.
You can decide the percentage split that is best for you to see the true impact; we typically do a 50/50 split for an equal weighting.
When carrying out any PPC experiment, avoid making any changes to the campaigns. This is so that it is a fair and controlled test. If you do need to make changes, ensure they are replicated on both the control and the experiment to get an accurate measurement of success.
Now we’ve come to the end of our smart bidding series we hope you to feel confident and comfortable with testing PPC automation
Whilst automation gives more optimisation control to Google, the strategy behind a PPC account should not be neglected. Still unsure? Get in touch to find out how we can help you reach your business goals through PPC.