The magic of finding the unknown unknowns in customer feedback

Large volumes of customer feedback make it overwhelming and challenging to separate signal from noise. Being able to find essential insights, aka your proverbial needle in the haystack, is when product teams can make magic happen.

Imagine if you could surface insights you didn't know to look for. This is what we like to call “finding the unknown unknowns.” or Anomalies in Enterpret. In these moments, companies can use feedback to elevate customer experiences and stand out from their competition by finding what their users care about most when it matters most.

Finding the big thing before it actually becomes a problem

Anomalies are significant deviations from standard patterns in your feedback data, manifesting as unexpected spikes or dips in volume, sentiment, or content.

In the context of feedback and product development, Anomalies are not just statistical outliers but critical indicators of underlying issues or successes in your product or service. Paying close attention to these anomalies can uncover valuable insights that would otherwise remain hidden.

How to Uncover Anomalies with the Z-Score

The heart of accurate anomaly detection lies in the Z-Score. The Z-Score is a statistical measure that indicates how far a data point is from the mean using standard deviations. (More about the Z-Score from Khan Academy.)

A Z-Score helps determine whether a feedback point is typical or unusual, providing a consistent basis for identifying noteworthy trends or patterns in user feedback.

Visualizing Feedback Data: Techniques

Visualization plays a crucial role in harnessing the power of feedback data. At Enterpret, we offer two essential types of graphs to aid in this endeavor:

• Bar Charts: These provide a clear, side-by-side comparison of different data categories, helping you quickly grasp the aggregate of various feedback aspects.
• Trends + Anomalies Chart: This specialized chart highlights significant deviations from standard feedback patterns, enabling you to spot anomalies easily.

Use Cases of Anomalies

Understanding anomalies can serve many purposes, from identifying product issues and ongoing monitoring of trends, problems, or the impact of new features to assessing marketing effectiveness and improving customer support. These insights allow businesses to make informed decisions, enhance user experience, and drive product success. Use cases include:

1. Identifying product issues: Spot sudden harmful feedback spikes to address potential problems.
2. Monitoring feature impact: Gauge user reception of new features or updates through feedback changes.
3. Assessing marketing effectiveness: Detect successful campaigns via increased positive feedback or tracked keyword mentions.
4. Detecting opportunities: Discover areas of excellence by identifying sudden increases in positive feedback or reasons.
5. Improving customer support: Evaluate support quality and identify improvement areas.

Using Enterpret for Anomaly Detection

Enterpret sets the default Z-Score for anomalies at 1.5, which can be customized to suit your analysis needs better. Adjusting the Z-Score threshold allows for more precise filtering of feedback data, enhancing the detection of subtle changes or reducing noise. Learn more about Enterpret Anomalies here.

Use Enterpret to Set Up Anomaly & Visualization Reports

Your team can stay updated on unusual patterns or deviations in your feedback data by subscribing to anomaly reports. These reports are automatically generated and delivered, ensuring you never miss crucial insights. Setting up these reports is straightforward:

1. Create and save your Quantify Query.
2. Navigate to the desired dashboard and add your Anomalous Quantify chart.
3. Subscribe to the dashboard to receive regular anomaly reports.

How Enterpret has helped product teams identify anomalies

Here are some examples of how Enterpret has helped teams with anomaly detection

• A retail and real estate investment platform identified a critical bug with Enterpret Anomalies that was making it impossible for its customers to complete a critical action on the platform. Enterpret helped them find the root cause, an integration partner update, and eliminate the bug the same day. Without Enterpret, the team estimated it would have taken weeks to identify, let alone fix, the bug, which would have caused a sizeable dent in revenue.
• A popular consumer app launched a new product that allows live calls with celebrities instead of pre-recorded messages. The app launched and saw a significant spike in refund requests/no-shows, and it wondered why. Using Enterpret, the app company noticed that customers needed clarification on the call's time zone. They could ship a fast-follow immediately and saw their metrics improve immediately.

Want help with Anomaly Detection? Enterpret can help!

By leveraging the insights gained from feedback data anomalies, businesses can proactively address issues, capitalize on opportunities for improvement, and better understand their users' needs and preferences. With the right tools and approaches, anomalies can transform from mere data points into powerful catalysts for growth and innovation.