Auto-Insights (Beta)

What's Auto-Insight?

Auto-insight is like having a smart and hardworking analyst nearby who can look at your data and tell you things you might not have noticed before. It can automatically find patterns, trends, and anomalies in your data and present them in an easy-to-understand format, without you having to do the analysis yourself.

For example, if you have a lot of sales data, auto-insight can help you identify which products are selling the most, which customers are buying the most, and which sales channels are most effective. It can also alert you to unusual sales patterns, such as a sudden increase or decrease in sales for a particular product, that you might have missed otherwise.

This is immensely useful in saving you time and effort by automatically analyzing your data and presenting you with meaningful insights that you can act on swiftly.

PresightIQ Auto-Insight

Where

For every anomalous data point, PresightIQ will automatically dive into a lot of the combinations of dimensions and dimension values of that particular metric. The purpose of this is to find out interesting patterns or behaviors of those smaller segments, then surge these up to you.

Example:

When you see that Sales of March 2023 is anomalous because it dropped 25% vs the month before, it is much more revealing and actionable if you know right then and there what exactly are causing this drop.

PresightIQ might help you save a lot of time and manual digging here by telling you immediately that we discovered Sales of Product X in Region Y for User Group Z of March 2023 behaving weirdly and might worth looking at, with a high change it might have been the culprit for the overall drop.

PresightIQ's Auto-Insights currently only runs as deep as 3 levels of breakdown (it can diagnose to the granularity of, say, Sales by Product X in Region Y for User Group Z) - anything more granular than that, please await our upcoming updates.

The 3 Types of Insight Recommendations

1️⃣Insight Type 1 - Based on Contribution2️⃣Insight Type 2 - Based on Strangeness3️⃣Insight Type 3 - Based on Strangeness (but at a more granular level)

How do you take this and go

While not claiming to be the exact diagnosis of the anomaly, this helps you to narrow down your search for the exact reasons behind your performance. Combine this with your understanding of the business, and the ubiquitous capabilities around collaboration on Presight, we recommend you drill the recommendations further with your teams (add the segment to doc, or chat with Metric Owner)

Why

(coming soon)

Action

(coming soon)

Last updated