Fresh Features: automated data discovery goes prime time

Fresh Features: automated data discovery goes prime time

Here’s part 2 of ‘Fresh Features’ of Yellowfin 9. And this week, we’re looking at Signals.

(If you missed part 1 on the awesome new charts, you can check out that blog post here.)

Signals is Yellowfin’s unique and powerful automated data discovery product. It will automatically scan your dimensional data and find any significant changes in your data. Then, it automatically sends you an alert, complete with analysis and correlations to help you take swift action to nip issues in the bud or build on successes. We released it as part of Yellowfin 8 but we’ve been working to make Signals even slicker and more helpful to you in this version.

Beavering away quietly in the background, the Signals team has been adding a lot of refinements to our Signals offering such as the addition of step-change signals, the ability to export signals to CSV, add a signal to multiple stories, the ability to disable correlations, and other changes. To top off all their good work is the Signals widget. So let’s jump straight into the detail.

Signals Admin

You can now manage and monitor Signals from the Signals area. The biggest change is the ability to create and edit Signal jobs directly from here without needing to edit the View. This simplifies the workflow and even allows you to run ad hoc Signal jobs if you are looking for something specific.

The Impact

You now get better workflows with no need to edit a View to create or update a Signals job.

Signal Filters 

The ability to filter a Signals job is a huge improvement to Signals but being able to run them from the Signals management panel is even better. Using our dataset of liquor sales, we used the Signals filter to home in on only signals of whiskey sales (we found a fascinating spike in Johnny Walker sales). It was here that we realised just how powerful a signals filter can be for running ad hoc Signals jobs. If someone has a specific question like, “What happened with our Johnny Walker brand this year?” a user could just run an ad hoc Signals job and filter for Johnny Walker and - boom! - a set of signals appears that you can start to analyse.

Filter signals analytics

The Impact

You get tailored signals to high business impact problems - make signals more relevant to our clients. You can use signals filters for ad hoc discovery - use Signals to purposefully find changes in a specific dimension of their data - speeding up data discovery. 

Yellowfin analytics dashboard signals widget

Signals on Dashboards 

OK, so, why would you want to put Signals on a dashboard? Well, one of the challenges with Signals is helping your end users understand the value of Signals. 

To help them join the dots, we decided to create a Signals widget that you can filter to only show signals relevant to that dashboard’s data. What you get is a list of Signals that can be filtered by view, type, and job. This way, when using a dashboard, the business user will not have to use filters to do manual discovery. Rather, the outliers they need to know about will be presented to them right then and there on their dashboard next to the relevant contextual data.

The Impact 

Having signals visible on your dashboard shows the value of automated data discovery - signals on the dashboard highlights just how much a dashboard does not tell you.

Signals is one of the biggest differentiators between us and other analytics platforms. You won’t find automated data discovery in the same way elsewhere. Signals adds huge value to your analytics investment. Data discovery is hard work but it doesn’t have to be if you automate it. All of these enhancements to Signals in version 9 - simpler signals management, filtering, and signals on a dashboard - help you gain even more value from your analytics more easily.

Try out Yellowfin

Take a look for yourself and try Yellowfin Signals, dashboards, storytelling, and more.