As researchers and analysts, we interact with a wide array of market research pros in the health and wellness space. In recent months, we’ve noticed a hesitancy to adopt algorithmic-based research techniques. Many traditional life sciences market research pros have confessed a fear that big data and machine learning will take over their jobs.

That fear is unfounded. Here are two reasons why:

Machine learning can enhance our understanding. The key is to strike a balance between the strengths on each side — human and machine — to achieve a depth of understanding not possible with either approach alone.

See also: Wellpepper wins Alexa Diabetes Challenge

Algorithms are great at analyzing data and predicting human behavior based on stimuli and can process data over thousands of variables to detect human responses. By using data to identify existing behavioral patterns, they can predict future outcomes. Such predictions offer brands an opportunity to not only anticipate and meet consumer needs, but also to produce strategic PR.

Social listening platforms are a good example of how these algorithms are used by market researchers today.

Researchers and marketers can gather data on sentiments expressed about certain drug classes or brands and use it to target hyper-specific patient groups, fill urgent information gaps, and inform future outreach campaigns.

But as any social-listening expert will tell you, the human lens is essential to distill actionable and disruptive insights from that data, regardless of the platform’s technical abilities.

See also: Diabetes marketers turn to big data, gaming to improve outcomes

Machine learning uncovers the “what.” Qualitative research explains the “why.” This requires us to step into customers’ shoes to understand their experience and what shapes their behaviors. Once the data identifies existing behavioral patterns (or predicts future behaviors), it’s time to unleash qualitative methodologies to get to underlying barriers and motivators. Qualitative insights enhance understanding of the data and can change interpretation of observed or predicted behavior.

For example, a classification algorithm may predict a group of physicians has a high likelihood of adopting a certain medication. Deeper qualitative dimensionalization may confirm this predicted behavior indicates true brand preference for that group — or for a subset of this group, it may show this behavior is circumstantial and driven by other outside influences (such as the brand preference of patients or the influence of payers).

This qualitative understanding shifts the target of strategies for influencing customers toward the desired behavior.

Kevin Troyanos is SVP of marketing analytics at Saatchi & Saatchi Wellness.

Earlene Worrall is VP and healthcare practice lead at Insync.