Confirmation bias has been widely observed (if not well understood) fact of human psychology for many years. These days we frequently hear about it in the context of polarized political opinion and people’s tendency to trust and believe only those facts and messages that align with their existing worldview.
Less frequently discussed is the impact of confirmation bias in marketing. Since the advent of digital advertising, the proliferation of measurement tools and techniques has vastly expanded the number of sources of information from which we can conclude the effectiveness of marketing. This has increased the potential for people’s cognitive biases to play a role in how media performance is assessed and what actions are taken as a result.
Before digital media and measurement, there may have only been a single source of information about campaign performance. Marketers might have had a media mix model, maybe some survey data and for the more sophisticated, some test results to work with. You could either believe the results from your media mix model or not.
Today, all of those measurement strategies still exist, but there are also several different types of attribution models to work with from a variety of vendors. There are site tag-based lift studies and digital live recruit brand surveys. There are location data and app usage data. All of these data sources and measurement techniques have their benefits and drawbacks, but the fact that none are perfect means that our biases will always have something to latch onto. “I see what the data says, but I do not believe it.”
Additionally, data granularity and richness have increased enormously as digital advertising has grown in scale and complexity. This allows marketers to focus on small parts of their campaigns to find data that aligns with their pre-existing expectations, then generalize that data across their entire campaign. “If I look just at this channel in this DMA then I see this, which must be true everywhere.”
Certain kinds of marketers are more susceptible to this than others. Consumer product marketers, for example, who manage brands that sell through a wide variety of channels (both online and offline, owned and third-party) and therefore have multiple points of sale to track, are particularly vulnerable given the challenges involved in comparing data from different points of sale. Likewise, marketers dealing with a wide variety of channels — from traditional to digital, highly targeted to mass reach (in financial services, for instance) — need to compare performance across very different channels with very different measurement solutions.
If you want to be an effective marketer, you need to avoid letting your confirmation biases affect your decision making. Here’s how.
Understand your biases and where they came from.
You cannot fix what you do not recognize, so understanding what your biases are is an essential first step before trying to address them. Most likely they are rooted in experience (“this worked at my old agency/on a campaign I worked on”) and understanding that will help you avoid bias in the future.
There is always a place for subjectivity in marketing decision making, but it is best left for times when there is no data to rely on.
Be critical of results.
This is especially true for channels you are most familiar with, where your comfort level is high and so is the possibility of being less critical than you might be with unfamiliar channels.
Set up accurate systems.
Set up your measurement systems to be as accurate and as comparable as possible. This does not necessarily mean having a single centralized source of truth — aligning methodologies allows for better decision making.
Remain open to changing your mind.
This is easier said than done, so regularly take a step back and listen to your colleagues and agencies. Just because what they have to say is new to you does not mean that it is wrong.
Confirmation bias is real, and it is just as real in marketing as any other area of life and business. Marketers must avoid falling prey to confirmation bias — the success of their campaigns and their companies depend on it.