The typical analytics ecosystem is rigid and based on two silos: business intelligence and data science. With Looker and Tellius, organizations around the world are breaking these silos and enabling more business and data analysts to make an impact using data.
Written by: Radu Miclaus,
at Tellius
In This Post
Why analysts spend so much time on KPIs
Analysts spend hundreds of hours every year tracking their key business performance metrics and identifying what causes those metrics to change. Look at job postings across every industry, and you will find demand for skills like “identify insights”, “find key drivers” and “interpret trends”. Analysts are expected to investigate the “why” behind key metrics systematically and determine how to address them — while simultaneously answering ad-hoc questions from business users.
Today, the familiar Looker Explore interface gives users the power to flexibly find insights in their data by slicing and dicing based upon all of the dimensions available to them. This can lead to hundreds of “slices” based upon the mountains of data organizations expose to Looker today. How can business users and analysts overcome this data deluge?
Tellius brings AI to Looker
The answer is artificial intelligence.
We built Tellius to provide both business users and analysts with the right piece of information at the right time by simplifying the analytics process and making it accessible to all.
Augmenting the Looker experience with AI can open up an entire new world of insights that take a fraction of the usual time to find and empower anyone to make predictions. Tellius uses techniques such as NLP, random forests, decision trees, neural networks, and more to quickly surface all possible factors impacting KPIs based upon all of your business data. These techniques dramatically compress time spent analyzing data, freeing up analysts and business users alike to start taking action.
These technologies accelerate time-to-insight for more users, overcoming the burden of hundreds of columns, billions of data points, and noise throughout it all. From here, the next logical step is making predictions to take business action. Tellius brings AutoML to Looker, empowering anyone to train models and start taking action in minutes.
Tellius + Looker in action
Take this Looker view for example, with a sample dataset based upon an eCommerce use case:
We can see some basic metrics that are key to this organization’s business. Check out the Average Order Sale Price (AOSP). To understand exactly which factors drove recent changes to this KPI, an analyst would need to go through the potentially time- and skill-intensive process of understanding how the metric was created in LookML, recreating it with ad-hoc analysis, and manually analyzing the different factors at play.
How to use the Tellius-Looker integration
With the Tellius-Looker integration, it takes only minutes to understand why this metric is changing. Simply connect Tellius to Looker:
Your LookML KPIs and their supporting data are now accessible in Tellius.
Let’s use Tellius to start asking some questions about our AOSP metric:
Instead of manually slicing and dicing, a user can apply AI to our data to identify what is driving change.
This automatically generated insight uses machine learning to scan across all of your data in minutes, finding key drivers of change. We can see that teas and juices, bread, and the brand Bermina have been the largest drivers behind the AOSP decrease.
Predictive insights with Tellius
To make a prediction about this metric would have required significant data movement and data science skills in the past. Tellius allows users to build predictive models based on their Looker data using an intuitive AutoML engine powered by Apache Spark. Simply choose your metric to predict, and Tellius will train the best model for the job:
Unifying analytics based on LookML
Providing AI-powered insights and predictive ML based on your LookML curated data dramatically lowers analysis time, removes analytics bottlenecks and extends access to more users. With Looker and Tellius, you can use AI to empower more users.
Integrating Tellius with Looker allows more users to take advantage of AI, compressing analysis time and enabling predictive analytics. In just minutes, AI combs across all of your columns, finding correlations, cohorts and anomalies that today can take hours of manual data work.
The partnership between Tellius and Looker allow organizations to:
- Extend data access to a broader population of users with natural language driven self service
- Find hidden insights that are difficult to uncover given contemporary methods
- Save significant time by automating complex analysis
- Enable fast machine learning predictive modeling
To see Tellius and Looker in action, request a demo with the Tellius team today.
