Spectacles part of ecosystem around LookerML

How often have you seen broken visualizations? Imagine: you need to extract calculated data quickly but all you can see is the “No data” sign. If you’re using Looker for visualizations, take a look at Spectacles.

Looker is a powerful data visualization instrument like Tableau or Power BI. The main difference is that it has its own coding language. On the one hand, Looker enables business users to drag-and-drop the fields and then auto-generates the SQL for them.

However, in case there are any changes in the database, Looker is totally unaware of them and subsequently can’t change metadata for you.

Those are the possible errors which are reasons for breaks:

• The table doesn’t exist;
• The column doesn’t exist;
• The data type is wrong.

What is the value of the Spectacles product?

The product is particular since its development is rooted in attempts to solve a real pain. Insecurity in terms of the access to real-time clear data about your business is unacceptable for companies growing in a fast-paced environment.

The problem solved by this product is managing the code and database changes proactively. When code changes are made by several people, the content in data visualization tools breaks easily. As a result, stakeholders will see wrong graphs instead of data.

Spectacles makes sure that all SQLs in Looker work with your database, which means:

• Every table is correct
• Every JOIN is correct
• Every column is correct

Moreover, Spectacles allows data engineers and analysts to conduct data tests.

The content in Looker is break proof even in case of development changes.

When database error occurs, the alert message doesn’t merely warn the user, but also guides through fixing before merging code to Looker.

The above is based on examining the clients’ expectations, who await the data quality to be more consistent, which means:

• Less bugs
• Less errors
• Less downtimes
• Developing more quickly

Feedback obtained demonstrates that when reviewing the code becomes easier, deployment changes do so as well. The above leads to increasing data confidence across the organization.

Where does the idea of this product come from?

The idea emerged while 2 data engineers were chatting at a dbt meet-up. The present co-founders initially built an open source tool. In July 2020, they launched monetization and customer acquisition.

What is the product all about?

Spectacles is the first commercial product built on the top of Looker tool.

In terms of the customer base, Spectacles is to be used by data analysts and engineers more than business people. They basically should know SQL and the special coding language called LookerML. Knowledge of the latter makes using Spectacles pretty straightforward.

It is worth mentioning that Spectacles is better to be used in small teams — it saves time and makes it possible to coordinate changes in a robust way.

“We are the first proper commercial offering who build the dedicated tool on top of the Looker ecosystem. Based on feedback from our users, Spectacles helps to serve the organisation better and increase data confidence.” © Dylan Baker, Co-founder of Spectacles.

In case of having a big complex deployment multiple environments, Spectacles brings the greatest value by validating the content available for end users.

What’s next?

First, an integration with dbt is to be conducted. Looker seats on top of the code developed by dbt. In case changes are made in dbt and not updated, there is a danger of break in Looker at the visualization stage. Data quality is always a pain for data engineers, therefore tests around the data are essential. Specific cases to mention here are primary key validation, data recency, data freshness.

The next challenge is that when a certain tool is used by several people, the order is not necessarily there. Thereafter it is essential to delete unused pieces of data and manage the rest in order to increase its functionality.

Summary

The future of Spectacles as a product is everything but boring due to the challenges anticipated. Creating a tool means caring not only about product development but also infrastructure management. In small teams roles are vague, therefore developers acquire new skills essential for handling sales and marketing challenges. At the same time, perfect knowledge of the base, Looker API, is still a must-have, lack of it might lead to many edge cases.

Without any doubt, the product created by developers focused on putting themselves into the users’ shoes is worth attention.