There are two basic approaches to Optimisation:
1. Reduce marketing spend while maintaining the same returns
2. Increase returns while maintaining marketing spend
Approach #1
Approach #1 involves cutting investments on channels which are ineffective. Channels with low quality (e.g., high bounce rate, low time spent) traffic contribution, high CPA and low ROMI are good indicators of inefficiency. Cut them out as soon as you can!
Approach #2
Approach #2 can be achieved in 2 methods as well:
Method A:
Reallocating budgets to effective traffic sources
Method B:
Reducing drop-offs (improving conversion rates via better targeting/content/messaging)
Method A involves shifting your marketing spend/efforts to more effective channels that show attributions of high quality (e.g., low bounce, longer time spent) traffic contribution, low CPA and high ROMI from lower performing channels. The advantage of this method is its ease of activation and quick turnaround time, thus it’s a common practice while campaigns are still running.
Method B involves more analysis work from identifying the cause of drop-offs at each stage of customer journey and reducing them through testing. These tests involve experimenting with various elements (such as content offer, layout, messaging, etc.) or combinations of them to determine what works best for the business. Although this optimisation method is slower and requires more effort, it is a more ideal solution for better conversion rates and offers you much more learnings towards the behaviour of your target audience.
For example, a simple A/B test can be used to determine the type of hero image used in banners that generates more clicks. Jane’s hypothesis – “My target audience responds better to banners that display lifestyle images.”
Illustration – A/B Testing
Version A
Version B
The findings from the above test will help Jane make informed and data supported decisions when it comes to deciding on images for her banner ads.
Presently, there are many A/B testing tools available in the market – Google Content Experiments, Optimizely, Unbounce, etc. Most of these tools require minimum effort to set up and offer comprehensive functions and reports to help you sample your audience and identify winners for your test.
As You Advance in Measuring Marketing Effectiveness
Measuring marketing effectiveness is an on-going process and may get complicated, especially when a business progresses to manage multiple marketing channels. By then, a customer may interact with a number of channels before making a purchase, therefore making it increasingly difficult for a business to single out an effective channel or understand how each marketing channel contributes to a sale. At this point, one can consider stepping into Attribution Modelling which is the science of determining the value of each customer touch point leading to a conversion.
Attribution modelling, however, is very complex and is different for every individual company due to brand values, channel mix and offer levels. This is especially so when a business has multi-channels across offline & online space. You will either need to be very statistical savvy to build your own attribution model or have large budgets to hire professionals. If your business focuses only on digital efforts, you may consider using Google Analytics’s free attribution model reports: https://support.google.com/analytics/answer/1662518?hl=en..
Find out more about more advance offline & online attribution modelling here: https://blog.kissmetrics.com/cohort-and-multi-touch-attribution/