You start your morning by checking email on your phone, one of which is from your favorite place to shop for shoes. Then, while scrolling through Instagram on your bus ride to work, you notice the same pair of shoes being promoted in your feed. Over lunch, you open up Facebook on your laptop and the shoes that initially caught your eye are there again. This time, you click on the ad, purchase the shoes online and decide to pick them up later that evening.
As consumers, we enjoy the convenience of connected experiences across devices and locations. As marketers, we are challenged with organizing our media channel teams in a way that aligns with a consumer’s cross-channel journey. As businesses, we aspire to understand which parts of the consumer experience are generating a positive return on investment.
As we assess our media investments throughout the consumer journey, one of the biggest challenges is choosing the most effective measurement solution. A frequently used solution, attribution, estimates a granular amount of credit to give a digital touch point on a consumer’s path to conversion. Within this, there are different kinds of attribution models you can use: first touch, last click, U-shaped, algorithmic and more.
At Ovative/group, we look beyond attribution and measure causal impact to understand if an individual touchpoint could make or break the purchase as a consumer works their way through the buying process. This idea is commonly referred to as incrementality. Here’s an example of where incrementality might come into play …
A brand is deciding whether or not to spend money on branded keyword paid search terms.
The question they would want to be able to answer is: If we didn’t have paid placement for our branded keywords, would consumers still click through to our website and purchase?
Your first thought might be, “But my attribution technology is telling me the return on ad spend for brand search is really high. Why would we consider turning that spend off?” While attribution is a vital tool, it is not – like many other forms of measurement – going to help you answer the question, “If that one touchpoint was taken away, would that person still have converted?”
In order to answer this question, Ovative often employs an unified marketing analytics (UMA) model that incorporates the results from controlled incrementality testing. Controlled testing means changing a single variable, under isolated circumstances, to determine whether that factor truly impacted sales and on what level.
For example, in a specific geographic area, you could turn off your paid search branded keywords on mobile, but leave them on for desktop. The results of this change are then compared against a test group with a similar market in which everything has remained consistent.
Whether the impact is negative or positive, incrementality testing allows you to determine the influence of what a single variable, or set of variables, did on someone’s path to purchase.
Based on our incrementality testing at Ovative, we have experienced branded paid search revenue to be between 0% to 65%+ incremental. The results that we have observed across different client profiles vary based on numerous factors including things like competition, device and seasonality. In some cases, we’ve seen opportunities to save, or re-invest, up to $3MM from branded keyword incrementality testing alone. In other cases we have observed some channels being undervalued by up to 30% even when using an algorithmic attribution model.
By incorporating incrementality testing, marketers can confidently take action on their attributed metrics while validating results and adjusting benchmarks through controlled tests.
There are always going to be gaps in data that prevent us from calculating a perfect return on the marketing investments we make. Whether it’s because of emerging touchpoints, siloed data, or security policies, understanding your customers’ entire path to purchase – and the impact of your marketing on it – should be an ongoing exercise for brands.