The Case for Behavioral Metrics

It’s often said that digital is the most measurable medium. With all of the data available to online marketers, it can be tricky to know which metrics work best for your campaign. With newer behavioral metrics, like lift in related searches, becoming more readily available its important to understand how these metrics work, and what their value is to the marketer.

Within the world of brand marketing, there’s a strong tradition of using surveys to collect brand metrics like awareness, favorability and purchase intent. Although well established, survey solutions can generate pain points for advertisers, publishers and users. Surveys can be arduous for consumers and for publishers with abundant campaigns, fulfillment also becomes a challenge. When it comes to measuring an advertising campaign’s impact on sales, most advertisers are reliant on database matches or mapping to panels with credit card information. These methods are accurate but require extensive scale and budgets to execute.

Today it’s important to look at how behavioral metrics — like trademark searches, site visits and page views — correlate with more traditional brand and sales metrics as this data is collected passively without interfering with the consumer-site experience. Additionally, collecting behavioral data can be done internally by publishers and are more cost-efficient than brand surveys or sales-impact studies.

In a recent analysis by Yahoo and ComScore of over 100 marketing campaigns online, Yahoo found that changes in online behavioral metrics were often strongly correlated with changes in brand and purchase metrics. This is new evidence to support the idea that for some categories, behavioral metrics can be predictive of traditional metrics as well as near term purchases, but that the specific relationships vary by category.

Based on these results, behavioral shifts appear to be predictive of changes in favorability within the CPG categories of health and beauty. This is especially true of a consumer’s likelihood to recommend a product and brand awareness. Behavioral shifts also signify changes in awareness, purchase intent and likelihood to recommend in the computer and technology sector. For retail and CPG food and grocery campaigns, behavioral shifts signify changes in purchase intent. Within the quick-service restaurant sphere, behavioral metrics are strongly predictive of both brand and purchase impact.

I’m not recommending marketers abandon tried and true methods for evaluating campaign success, but using behavioral metrics can serve as reinforcements for traditional studies or be considered useful proxies in a few situations. First, when surveys are not ideal or possible and, second, in order to optimize campaigns while in the field. Advertisers should collect both survey and behavioral metrics to paint a more holistic picture of a campaign’s impact. This is especially true in categories like retail and CPG food and grocery categories where behaviors are indicative of success in driving immediate sales, and brand metrics are more indicative of changing longer-term perceptions.

Collecting both sets of metrics for campaigns in all categories will allow advertisers to understand how brand, behavioral and purchase metrics are related for their specific campaigns and objectives (which may differ vs. their respective categories). As our understanding of the relationships between metrics continues to get more sophisticated, marketers will have more options for measuring success against traditional goals.