Webinar Q&A: Key Insights Into Best Practices for Embedding Analytics

From answering the question of build vs. buy, to explaining integration options, our recent webinar offered in-depth insight into best practices for embedding analytics. (If you missed the webinar, you can watch the full recording here.)

While our GoodData expert addressed topics such as the essential requirements of security and compliance, as well as the need for a forward-thinking strategy around scalability, attendees also submitted their own questions for further discussion.

Below are the top questions and answers from the Q&A portion of the webinar.

Can you share the GoodData and Pulse survey results [that include the statistic that “98% of IT executives say they need to shift from data collection to insights consumption”]?

Yes, you can find the survey results here.

What are some real-life benefits of a “headless BI” engine?

Headless BI removes the tight coupling between the BI components and exposes the semantic model as a shared service via APIs and standard interfaces. This guarantees that every single data consumer (e.g., machine, workflow, end user, etc.) is using consistent metrics and insights across the whole organization.

If you would like to know more about this topic, you can watch our “Introduction to headless BI” webinar here.

Do you offer any JS SDKs out of the box?

We offer our own GoodData.UI, a React-based JavaScript library for free. You can find more about GD.UI here.

How is your platform different from, let’s say Cube.js, which is open-source?

Compared to the GoodData platform, Cube.js is just the middle part that is the analytical API. On the other hand, GoodData’s full-fledged platform offers data storage (Agile Data Warehouse Service Service, or ADS), a data exploration tool called Analytical Designer, a responsive dashboards application, and the open source GoodData.UI JavaScript/React library for building custom solutions on top of the GoodData platform. GoodData’s professional services team can help with ETL, Data Distribution, LDM updates, etc.

Can I generate a custom URL to a dashboard for use by our customers?

Dashboard URLs are unique per workspace (i.e., for each of your customers). The URL format is: https://your.gooddata.domain.com/dashboards/#/project//dashboard/. Both the workspace ID and the dashboard is typically some random generated hash, but can be altered/adjusted if needed.

What is the easiest way to integrate GoodData into my web app? Back-end data absorption and visualizations are built; I just want to display a dashboard to a user logged into my web app with the data filtered for that user.

There are several options. You can embed the dashboard using a standard HTML iframe, and there is this setFilterContext command that uses the Window.postMessage() method to filter the dashboard as desired. A better option might be to avoid iframes and integrate the dashboard directly with a React component from GoodData.UI library. You can read about the pros and cons of each approach here.

What is the biggest cost when building analytics in-house?

The actual development: The analytics field changes rapidly, and your users’ expectations are growing all the time, which means you must commit more and more resources to improve your analytical solution to fulfill those demands in the long term.

You were talking about the physical and secure isolation of data. How is this done in the GoodData platform?

In GoodData, each customer or organization has its own dedicated workspace with its own database. This physical isolation makes it impossible for users from one organization to see data that belongs to another organization.

We have our data in PostgreSQL. Do you support this database?

Yes, we support data distribution from your PostgreSQL database. In GoodData.CN, you can deploy GoodData directly next to your database to decrease the latency and support the real-time analytics use case.

If I understand correctly, ADS is GoodData’s own data warehouse that you offer to your customers. Is there some limit on how much data can be uploaded to ADS?

That’s correct. ADS or Agile Data warehouse Service is our proprietary data warehouse solution. The maximum size of a Data Warehouse instance is 384 TB, but this is not a hard limit.

What is your data refresh frequency? Does your platform support real-time analytics?

Our new GoodData.CN platform offers real-time capabilities. More about GoodData.CN here.

Can we use third-party visualization libraries with GoodData?

Yes, you can integrate any third-party visualization library with GoodData via GoodData.UI, our JavaScript library.

By giving the end users a self-service option, doesn’t that mean we would lose control of what they do with the data?

Not necessarily. If you provide your end users with well-defined base metrics, and if your analytical solution ensures that no one can combine data sets that don’t make sense, then you are not losing control. Instead, you are empowering your users to become more data-driven.

What is the biggest benefit of self-service analytics?

Self-service capabilities allow more people, especially non-technical users, to participate in data exploration and discovery, which will help your company to find insights that will drive value and growth.

Can we decide which of our end users have access to this analytical designer self-service tool?

Yes, absolutely. You can give the self-service tool only to specific users.

We need the analytics in the same colors and branding as our product. What options do you offer?

Good question! You can use our theming capabilities — read our recent blog post about this topic. Or, you can use GoodData.UI t, our front-end development library that offers even more flexibility.

What kind of automation options does GoodData offer?

With GoodData, you can automate user provisioning and de-provisioning of your workspaces and users, and you can automatically roll out new versions of your dashboards and data products to all your hundreds or thousands of end users without rewriting their customizations.

So, GoodData pricing is based on these so-called customers, not individual users? How many end users can each “customer” have?

That’s correct. You can have an unlimited number of users per customer, or per workspace as we call it. With this kind of predictable pricing model, you do not need to worry about the costs of adding new end users to use your analytics solution. You can read more about our predictable pricing from this blog post.

How much data can each customer have?

Each customer, or workspace as we call it, comes with 1GB for free. And each workspace can have up to 1TB of data. You can find more about our pricing here.

If you are interested in GoodData.CN, please contact us. Alternatively, sign up for a trial version of GoodData Cloud: https://www.gooddata.com/trial/