4 lessons learnt from the launch of Signals

4 lessons learnt from the launch of Signals

As a key part of Yellowfin release 8, we recently introduced Signals. Signals delivers automated alerts to users about critical changes in their business. We’ve been using the product internally and have learned four things about how Signals can be used within your organization.  

1. Signals don’t replace dashboards

Initially, we thought Signals’ discovery process would completely replace dashboards, but that has not been the case. Signals augments the dashboard experience but doesn’t replace it. While Signals can tell you what's changed, there's an understandable imperative for people to view their business as a whole even if nothing changes. This is what dashboards do. They give you the ability to monitor your business, see long-term trends, and get a general understanding of what's happening in your organization. That use case is critical to business users, but it’s different to the use case of discovery.

2. Signals are great at discovering the unexpected

Where Signals substantially outperforms dashboards is in finding the unexpected. When we were building and testing Signals we used our own internal marketing and sales datasets. We thought we knew and understood these datasets well, but when we applied Signals to them we discovered things that were quite unexpected. We learned a lot about our business and how it was running as a result of testing Signals. As a business user, I can confidently say the product made me think about my business differently.
Signals is the starting point - automated analytics

3. Signals are a starting point for analysis not the end point

When we were building Signals, we also thought that it would provide answers so that people could take action but we’ve learned that's not the case. Signals are the start of the analytical process not the end point. They provide a starting point for engaging with data analysts so that you can better understand your business. Signals provide a trigger for action. The product helps you identify specific things that you want to know more about. It makes you think long and hard about your business to determine whether or not you need to do more analysis. This means that Signals becomes something you have to manage, like an inbox. You have to be able to dismiss or assign people to do things around the Signals you are seeing.  

4. The role of the data analyst is even more critical

The final lesson we learnt by using Signals is that it doesn’t automate the data analyst’s job. In fact, the role of the data analyst is more critical with Signals than it was in the past. This is because Signals helps identify issues that data analysts can then analyze and explain. The art of analysis doesn't go away but the role of the data analyst does change. In the past, data analysts have been employed to build dashboards and reports without any context or triggers. Now they need to do analysis that is highly targeted and specific. They need to investigate what happened and why it happened which is really powerful.
Analysts are more crucial than ever
By using Signals internally, we’ve learnt a lot about what Signals are and what they aren’t. This has shown us how effective the product is in helping organizations identify unexpected trends. Whilst there’s no doubt the product is going to change the practice of analytics, I think ultimately we’re likely to see more reinvention than replacement of the data analyst role. Signals provide a better starting point, but there’s still huge value in investigation and building a compelling narrative for change.

The future of analytics departments: your analytics team needs to restructure for automation

If you want to make sure you are building a team for tomorrow, download this paper on the future of analytics departments.