3 key considerations for embedded analytics

  • 28 March 2022
  • 0 replies

Userlevel 5

This content, written by Shohei Narron, was initially posted in Looker Blog on May 24, 2016. The content is subject to limited support.

You’ve built a great product that resonates with the market. Your customers are enthusiastic, and word of mouth is overwhelmingly positive.

The in-house picture is cloudier. Your end users can’t run their own queries, which means they have to wait for someone else to generate the information they need. Because your solution requires moving data out of your database, analysis is complicated and time-consuming, placing your data analysts in an uncomfortable position: just by doing their jobs, they’re creating a bottleneck. Your engineers would love to help, but wouldn’t focusing on your core offering be a better use of their time?

You’re now considering adding value to your customers by giving back the data you’ve collected through a new analytics offering, and you wonder whether to build the tool yourselves or buy an existing tool.

Enter Powered By Looker, an analytics solution that allows you to provide data access to your customers in precisely the way they want, from a single data point to a fully white-labeled version of Looker. Our customers led us to this solution—they realized how valuable it would be not only to analyze data internally, but also to provide insights to their own customers through Looker. As a result, we extended Looker’s capabilities beyond the walls of their organizations.

Since launching Powered By Looker, we’ve received excellent feedback and suggestions. We just added the ability to customize the look and feel of embedded dashboards, and we plan to implement many other features, such as allowing end users to save their own reports.
With so many vendors offering solutions that incorporate amazing data science, visualization, marketing analytics, and more, differentiating among them can be challenging. Why do our customers love our embedded analytics solution more than others in the oversaturated BI market? Looker is an ideal choice because it helps you:

1) Stop losing engineering focus

Your engineers should be focused on making your product better, not on scrambling to create a new BI product outside of your core competency. In addition to the initial development of a new product, there will be customer support needs and ongoing maintenance, requiring an increasing amount of effort and time as more customers sign up. In companies that try to build an analytics tool in-house, we’ve seen that a) the project draws engineering focus away from core offerings and b) the tool inevitably gets stale very quickly once engineering resources move elsewhere.

Providing you with the best BI experience is Looker’s core focus—our product (and your experience) gets better with every release. All you have to do is to model the data and choose what and how you’d like to expose this data, a process made even easier if you’re already using Looker internally.

2) Stop moving your data

Unlike most BI tools, Looker works entirely in your database. You don’t have to move your data out of the relational format it’s in, nor will you have to give your data up to Looker. This is great, especially if your team is using an MPP database (such as Amazon Redshift, Google BigQuery, or HP Vertica), which is blazing fast and easy to scale with increasing data volume. All questions are answered using your database, so you can take advantage of the investment your team has made in your infrastructure. And because source data doesn’t move out of your network, you enjoy an added security benefit.

3) Stop querying for others

In the traditional analytics process, a person who has a question asks an analyst for answers. The analyst has to identify the data, think about how to measure the results, write a query (and probably find some mistakes), then finally get back to the requester. Then comes the follow-up question, which is often just the first of many.

Instead of this inefficient process, Looker enables end users to query information on their own—without any understanding of programming or data—by translating their questions into SQL via LookML. LookML is a business logic layer that abstracts SQL queries, making it easier to define and maintain attributes and metrics. This not only enables end users to get the insight they need (and more), but also guarantees consistent and reliable calculations across users.

Here are a few examples in practice

Campus Logic

Campus Logic is a customer that purchased Looker in late February of 2016 and went live with its product about 2 weeks later. Through Looker, CampusLogic powers their CampusMetrics service, a platform for student financial aid.

Urban Airship

Urban Airship embeds Looker to enable their customers to uncover user-level insights about their mobile app audience. Urban Airship Insight combines performance dashboards, funnel reporting, cohort analysis, customer segmentation and ad-hoc exploration in Looker to help customers understand the "why" behind ROI.

Google Cloud

Google worked with Looker to embed analytics that convey the power of BigQuery. Try it yourself!

What happens next

You’ve launched a terrific new product, but your work’s not done. You need an analytics solution that respects your engineering team’s time and talent, keeps data in your database, and empowers end users to create their own queries. You need Looker.

If you want more information about embedding an analytics solution in your web page or application, please We’d love to help!

0 replies

Be the first to reply!