Data Analytics

Your embedded analytics questions, answered

It’s natural to have a lot of questions when you’re evaluating embedded analytics solutions. In my years working with companies that embed analytics with Looker, I’ve noticed the same questions come up time and time again — and today I want to share the answers (and related resources), organized by theme. Think of this as a reference guide you can come back to in your research.

And in the coming months, we’ll showcase examples of real embedded analytics dashboards our customers have created — so prepare to get inspired by what your peers are up to. But for now, on to the Q&A.

Scalability and performance

How do you scale your Looker instance?

Generally speaking, Looker embedded analytics deployments are clustered with multiple Looker nodes to achieve high availability. Based on usage, you have the option to upsize both vertically (by increasing node size) and horizontally (by adding new nodes). If you choose to have Looker host your deployment, Looker Ops will take care of sizing and monitoring for you.

We currently support AWS and GCP hosting, with Azure to follow this year. You also have the ability to use additional Looker instances as dev or staging environments to achieve your desired application deployment workflow.

What are my main considerations for data warehouse setup (size, volume of queries, separate warehouse, etc.)?

Database and warehouse selection should match both your business and analytical use cases, and may involve more than one database/warehouse. Your evaluation parameters should include ETL, storage, usage volume, and concurrency considerations.

One of the most common problems that we see our embed customers run into is database concurrency limitations, so be sure the database you choose can handle your estimated concurrent query volume.

Data schema design should largely be informed by the types of analysis you plan to do and one of the main goals should be to produce results to analytical queries quickly. You can use the Looker model layer to test out aggregations and transformations before making the decision to formally incorporate those changes into your ETL layer.

Security

How does Looker ensure data security?

From the database connection perspective, an SSH tunnel can be established for a more secure connection. For normal connections, a variety of mechanisms prevent hacking/SQL injection and other security risks.

Once the database connection in Looker has been set up, you can easily implement row level, column level, and table level security at the user or tenant level using user attributes.

Where do I provision different levels of data access for my users?

You have the option of implementing data access controls in Looker, in your database layer, or distributed across both the database and modeling layers.

  • Database layer: You can use Looker user attributes to dynamically control which databases and schemas each embed user can access.
  • Looker modeling layer: You can assign field, row, column, and table permissions at the user or group level.

Looker also offers the ability to set content level access controls, which acts as a second layer to ensure security for your users.

Customization

When a customer or client encounters Looker, does it look like they are using Looker, or are there masking/co-branding options?

Whitelabeling and theming are available to change colors, fonts, and remove Looker mentions throughout embedded Looker content so that it ties in more seamlessly with your application UI. There is an additional cost for whitelabeling. Additionally, you have the option of using the Looker API to extract data and fully customize your application’s look and feel.

Is embedding my only option?

In short, no. There are other ways you can bring data to your customers.We have an option to whitelabel Looker and allow direct login to the app that involves almost no custom application development. Alternatively, you can use the Looker API to generate raw query results and build a fully customized application experience that is powered by Looker on the back end. Embedding is a middle ground that offers a lot of convenience along with customization options. Most of our embedded analytics customers use a mixture of embedding and the API.

We offer an embed SDK and API SDKs to jumpstart your application development. For additional tools that will help you get ahead, check out the extension framework.

Monetization Strategies

What are common monetization strategies customers employ?

We have seen customers successfully monetize embedded analytics through a tiered offering. Here’s an example of what that might look like:

  • All customers get access to a “Basic” analytics portal where they can view reports
  • “Advanced” users can also schedule reports to be sent to emails or other destinations, download reports, and set up alerts that notify them automatically when a metric reaches a certain threshold
  • “Premium” users then gain access to custom reports, as well as the ability to build their own reports

We’ve also seen companies succeed with a single tier, add-on analytics product offering.

Resources and skills needed to deploy embedded analytics with Looker

What team members and/or skills are needed to deploy an embedded Looker application?

Many companies have have successfully deployed Looker with a 2-3 person team while others have orchestrated massive deployments with 20-30 person teams. The key factors that make this possible are:

  1. A database with cleanly organized data that is well suited for analytical queries
  2. A software development team to build your embed application
  3. A SQL developer who knows data well and can build out the Looker model and content
  4. A data product owner with a vision and access to customers

We’re here to help. Our Professional Services team and partner network can act as supplemental resources in any of these areas.

How should we educate internal team members and embed users on

Looker has numerous self-service education resources like docs and trainings available for use. Send them directly to your internal users, or let them serve as a starting point for you to build your own help center with your branding. You can also work with our Professional Services team to set up customized training courses and content to enable your internal teams.


Ready to delve into more details?

Check out the resources linked below to satisfy your technical curiosities.

Embedding

Extension Framework

API

Data and Content Security