Transform’s Metric Catalog is a user interface that sits on top of your metric definitions. It allows you to view and query all your defined metrics across various dimensions. Our Metrics Catalog also helps users build a knowledge base around a given metric.
With Transform’s Metric Catalog, data teams and business teams can interface with their metrics and ask/answer questions about the metric’s behavior, definition, and other attributes. Additionally, users can see anomalies in the data and gather context with the Annotations feature.
The focus of this article is to showcase some new additions to our Annotations feature, allowing users to explore annotations directly on the graph for an affected metric.
Annotations in the metrics lifecycle
Annotations are one of the most critical parts of a metric life cycle. If you use data for day-to-day decision-making, you’ve likely been in a situation where a spike or dip in data is present, but no one knows why. Executives are up in arms, data teams are frantic, and there’s no single source of truth to obtain information on why a metric behaved the way it did.
It turns out, after hours of investigation, one of the veteran analysts already knew why the anomaly occurred. She had documented the anomaly in an old PowerPoint presentation, but there was no place to gather this information and share it with the rest of the organization. Transform is aiming to change this paradigm so that anyone can answer a question about an important anomaly in the data directly in the tool.
Introducing Annotations on Graph in Transform: Understand the full context of your data
In Transform, we support annotating metrics on our metric page. You can select your dates as any metrics and dimensions that you’d like to annotate. Annotations can be prioritized from low, medium, high, and critical based on the impact. In the picture below, you can see the table of annotations and the impact on the right-hand side.
Before, Annotations were accessible to both data producers and data consumers in a tabular format as demonstrated in the image above, but there was no way for you to see these annotations in the context of the metrics graphs. Since annotations are so tightly connected to the data time series, the tabular format does not paint the full picture of how the annotation relates to the data.
Now we’re excited to announce Annotations on Graph. This feature puts your annotations directly on the metrics chart in addition to the tabular format—helping you tell your data stories even more effectively to your internal stakeholders.
This feature provides a bubble, color-coded by the impact of the annotation, above the given date range that clicks into each annotation. Additionally, you can toggle displaying and hiding annotations as you wish. If multiple annotations are present for a given day, you can drill into each independently, and they are organized by annotation impact from critical to low.
The images below show a metric called Messages Sent with three annotations colored by impact. You can dig into each annotation by clicking the top bubble.
How to add Annotations in Transform
In this video, you’ll learn how to use Annotations to:
- Describe expected outcomes: For example, if there is an outage in Toronto which dropped “messages sent” in Canada, you can add an outcome to your annotation that describes the impact, like “there will be less messages sent in the application.”
- Show how other metrics are affected: If you have correlating metrics that would be affected by the outage in Toronto, you can add other metrics to this annotation for context.
- Describe the impact: You can mark a metric as low, medium, or high impact to arrange and prioritize your metrics.
- Use your metrics graph to explore annotations: Click on the annotation in the graph to see the context that you or someone else added.
We can’t wait to see how organizations use this feature to quickly find and track anomalies in their most important business metrics. Explore other Transform product updates on our blog. Want to try Transform for your own metrics? Schedule a product deep dive.