New Relic Insights: More Analytics, Visualization, and the Power of Custom Attributes

Today at FutureStack14, New Relic Insights releases new ways to analyze and visualize your data. We focused this release—the first since Insights became generally available in July—on delivering key features requested by our customers who wanted more ways to slice, dice, and visualize their data.

Top among those requests were features like funnel and cohort analysis, but that’s only part of what we delivered. This product release includes four key features:

  • Funnels to see how consumers are progressing towards your business goals in your application.
  • Cohort analysis for analyzing engagement within your application over time.
  • New math functions that let you create composite metrics from multiple attributes. That is, you can see the number of actions per unique registered user that are changing over time.
  • Bucketing for grouping numerical values over a range, to help you better summarize data.

Funnels

Funnels let you understand how customers are moving through your application via a series of predefined steps towards a specific goal. For example, a SaaS product company might want to see how many users have completed all of these steps within a session:

  1. Visiting the storefront homepage
  2. Registering by submitting an email address
  3. Comparing features on the feature-comparison page
  4. Upgrading to a paid account

Insights 1

Note: You can easily relabel these steps for your funnel visualization in NRQL using “AS” functionality, so it’s clear what each step in the funnel represents.

Adding custom attributes gives you more flexibility when building Funnels. If you add a custom attribute that identifies a user or an account you can now track how they move through your funnel over longer periods of time by replacing funnel(session with funnel(user and specifying a larger time frame in your SINCE clause.

Cohort analysis

Cohort analysis lets you group any attribute you have stored as a timestamp in your FACET clause. For example, if you had added “account_created” as a custom attribute you can now group your users into cohorts based on the week, month, quarter, and year that they created their account. You can then quickly see who your most engaged customers are at any given time. Are they the most recent signups? Did the customers you acquired from your marketing campaign last year stay engaged over time or did they just want the free trial you offered?

Supported cohorts include

  • monthOf– returns the month of a timestamp (e.g. July 2014)
  • dateOf– returns the date of a timestamp (e.g. July 1, 2014)
  • weekdayOf– returns the week day of a timestamp (e.g. Monday)
  • hourOf– returns the hour of a timestamp (e.g. 13:00)
  • weekOf– returns the date of the week beginning on Monday (September 29, 2014)
  • yearOf– returns the year of a timestamp (e.g. 2014)
  • dayOfMonthOf– returns the day of a month (1-31)
  • quarterOf– returns the quarter of a year (e.g. Q1 2014)
  • Insights 2

Cohorts can also take advantage of our new bar chart visualization available for any query that uses a FACET:

Insights 3

Math functions

New Relic Insights now supports basic math functions including addition, subtraction, division, and multiplication. These functions let you see things like average PageViews per session faceted by country. The new math functions also let you to calculate many key marketing and sales metrics, such as conversion rates, subscription rates, drop off rates, and more.

Insights 4

For example, if you’ve added custom attributes, like “totalCartSize” to your PageView or Transaction event types, you can see how the ROI on your marketing spend in Argentina compares to your spend in Brazil:

Insights 5

You can also break out your Front End Duration from your total duration to pinpoint that spike on 9/27:

Insights 6

Bucketing

Speaking of duration, let’s say you wanted to know how slow-performing pages were impacting your customers’ total purchase amount. A new function called Bucketing lets you break out page load duration into equal-sized buckets and look at the average total revenue for each:

Insights 7

In this example, once our page load times get over 4 seconds we see an impact on revenue. The FACET Buckets() clause can be applied to any numerical value in your data.

Gauges

And lastly, our new Gauges let you see how you are tracking towards certain goals that you specify when creating a widget. In this instance, we set an hourly goal for our e-commerce site of $1,000 an hour, but we’re currently running at only 38.9% of that goal. The problem could be that too many pages are taking more than 4 seconds to load…

Insights 8

Using these features with Custom Attributes

After talking with our customers about the power and simplicity of adding Custom Attributes, we unearthed a trend in the kinds of data they were interested in tracking in New Relic Insights, and how they could combine external data with the software data already generated in their other New Relic products.

New Relic products’ default data sets, combined with three new custom attributes—People, Things, and Money—take your analytics to another level. Adding a few of these new attributes can help the new features around funnels, cohorts, and math, get meaningful real-time data directly from your software.

Types of “People, Things, and Money” custom attributes that might be relevant to your business:

1. People

1. Users

2. Companies

3. Accounts

2. Things

1. Product ID

2. Subscription level

3. Features and/or site content

3. Money

1. Product cost or cart value

2. ARR or subscription cost

3. Advertising and marketing revenue

We believe these new features can make New Relic Insights even more useful, and we hope you enjoy using them. Please feel free to provide feedback on the new features in our community forum.

For more details on these new features, see the New Relic Insights documentation site. Or watch this short video:

Jim Kutz is a technical product manager for New Relic. Before joining New Relic, Jim was a senior product manager for Nike Digital's Consumer Insights and Big Data Platforms, director of business intelligence and analytics for The Motley Fool, and a data analyst manager at Capital One. View posts by .

Interested in writing for New Relic Blog? Send us a pitch!