September 4, 2019
How You Should Actually Be Testing and Evaluating Google’s Responsive Search Ads
One of Think With Google’s most powerful tools is the Customer Journey to Online Purchase. In a previous post, I explained how to use Google Analytics’ attribution modeling to evaluate existing efforts and determine where to shift budget from one digital channel to the next.
This begs the question: how do we know which channels to test if we don’t have any previous data? This tool is a terrific starting point. Google collected purchase data from over 40,000 ecommerce accounts and evaluated the assist-click vs. last-click ratios of different online channels. They isolated this data by industry and size (based on the number of ecommerce transactions). Google used this data to show which channels most often assist in driving an online transaction compared to which channels most frequently close the transaction.
The Customer Journey to Online Purchase tool displays this information in an easy to read spectrum:
Direct site visitors drive the largest proportion of last interactions for most industries, but we can see large variances between industries in the positioning of other channels. Let’s compare a medium-sized home & garden advertiser (pictured above) vs. a medium-sized beauty & fitness advertiser (pictured below):
A comparison of these two models shows that Social is most likely to provide early assist interactions for a beauty & fitness company, while Generic (i.e. non-brand) Paid Search fills this niche for home & garden. So if we’re looking to grow top-line digital revenue for the home & garden company, we should consider testing a larger non-brand search budget to target additional top-of-funnel keywords. For a beauty & fitness company, testing an increased Facebook advertising budget could be a good way to grow total revenue.
We can also see that Brand Paid Search and Email are flipped as far as their proximity to the right side of the spectrum. This suggests that for a beauty & fitness consumer, a brand searcher is bargain hunting. We know that email is a terrific way to highlight sales and it would appear that channel is closing the deal more frequently than brand search. This would inform my ad copy and messaging. Brand paid search copy should highlight pricing when it is competitive. Regardless, we should at least test a landing page that has a prominent email list signup so that we can close the deal that way. In addition, email language should create urgency when we are offering specials and sales.
On the home & garden side, email language should be more geared at nurturing consumers through the funnel. Also, all channels with a larger assist-ratio should be evaluated in part on their ability to increase branded search volume.
Although this is macro level data, it can be extremely useful for brainstorming new tests as part of your digital marketing strategy. Some data is always better than no data, and Google’s Customer Journey to Online Purchase tool can provide insights for where you should start.