How to Increase ROAS by Analysing Path to Conversion Data

You might have heard that the return on advertising spend (ROAS) from search advertising is typically the highest in all marketing channels. Search advertising accounts for 61 percent of all online marketing investment in the U.S., according to a Forrester Research report.This is becoming a trend in Asia Pacific as well, where marketers are investing more money on search advertising.

For the marketers who I work with, while some campaign objectives are still driving awareness or traffic to clients’ websites, we can see more advertisers now utilise paid search to drive online sales, especially in travel, banking/finance, and education industries.

As an agency, our goal is to help our clients maximise their ROAS. What are some of the effective ways to achieve the highest return on investment? Our experience tells us that one thing that works well is to analyse the “path to conversion data”.

Before introducing the “path to conversion” notion to my clients, many are still applying “the last click attribution model” to their paid search campaigns. They might be using a bid management tool, but the tool only takes into account the “last click” that contributes to the final sale.

As today’s campaigns grow larger and more complex, skewed data can have severe repercussions when it comes to making strategic decisions. Coremetrics research has found that customers interact with online brands 5.5 times before a conversion takes place. By relying solely on the “last-click attribution model”, marketers’ performance data quickly becomes skewed towards the keywords that are at the end of the conversion funnel. This is one reason why some marketers might find an expensive keyword drives traffic to their websites but does not result in an acceptable number of high-margin conversions.

I work with clients in the travel industry. The obvious search marketing goal is to acquire online bookings. For a hotel client that we work with, we discovered that a typical conversion takes three days to complete, and users search at least three keywords before a conversion takes place.

Let me give you an example. A user might be planning a trip to Melbourne for vacation. He goes to Google, searches for the keyword phrase “Melbourne city hotel deals”, and clicks on my client’s ad, but he doesn’t book any rooms. The next day, the same user searches again. This time he types “places to stay in Melbourne city”. Once again he clicks the ad but doesn’t convert. On day three, he goes to Google again and types “Melbourne hotel deals” and from our path to conversion data, he made a transaction by booking a hotel room for five nights.

Looking at the above example, if we solely rely on the last-click attribution model, we would have given full credit for the sale to the last keyword “Melbourne hotel deals”, even though it should share credit with the keywords “Melbourne city hotel deals” and “places to stay in Melbourne city”.

It is therefore important to leverage tools that monitor the entire path from the first click to the conversion. Some bid management tools have the ability to do that and they provide you with detailed reporting on the search terms that participated in the path. They can also automatically distribute conversion credit weights for all clicks leading to conversions based on your marketing goals.

This attribution modeling opens up the benefit of being able to view all your customers’ interactions and to attribute proper ROI to each stage in the funnel. This will ultimately increase the overall ROAS from search advertising.