Item: Sigma Computing
Rating: 9
Author: Verified User

Use Cases and Deployment Scope

We're in the process of rolling Sigma out to our entire company as our "Analytics-endorsed" BI tool, replacing Domo. We're looking to facilitate easy self-service for business users to answer 80% of their own questions. For explorer and creator users, they have a ton of flexibility in slicing, grouping, and aggregating the data as they see fit.

Pros and Cons

Easily connects to Snowflake.
dbt Integration to show column level definitions and dataset metadata (freshness, size, rows, etc.)
Very quickly create workbooks and content.
Unlimited viewer model, so everyone in the company can have read-only access.

Viewer level license is quite limited. These users can't download data or even add filters on datasets. Something to keep in mind.
Directly querying the underlying data warehouse will lead to increased usage. Not a big deal on something like Redshift, but your Snowflake consumption will increase, potentially by a lot.

Return on Investment

Users have expressed significantly easier experience creating their own content, compared to our previous BI tool.

Self-Service Analytics

Since we're still in the midst of our rollout, this is a bit of "yet to be seen". We intend on rolling viewer access out to the whole company, so every person will have read-only access to certain company wide content, such as dashboards showing our level 1 metrics. For power users, they will have all of the available data marts from Snowflake to query.

Ad Hoc Analytics

Since creating Sigma content is quite quick and easy, our analysts can create Sigma content to answer ad-hoc questions, if the user thinks they'll need the data more than once. We can easily revert back to earlier versions of dashboards to see what the state of the data would have been in a previous iteration.

Alternatives Considered

Tableau Cloud and Looker

With Looker, to be effective, a substantial amount of coding & modeling needs to happen in LookML. Being another language to learn, users have to context switch again from at a minimum either SQL or Python into LookML. The concept of being able to source control, code review, and deploy your models is a plus though.

Tableau is the gold standard for data visualization, no question. Power users will be able to create dazzling content that Sigma won't necessarily be able to easily match. However, since development usually happens via an extract, helping other users troubleshoot is an arduous process. Trying to re-do or un-do all the transformations and calculations that cause a certain number is very difficult.

With Sigma, all the queries happen directly against Snowflake and you can see the query logs. The data modeling happens right in a tabular, spreadsheet-like manner, so within only a few minutes, substantial transformations can happen, with visualizations just a few more clicks away.

Key Insights

Do you think Sigma Computing delivers good value for the price?

Yes

Are you happy with Sigma Computing's feature set?

Yes

Did Sigma Computing live up to sales and marketing promises?

Yes

Did implementation of Sigma Computing go as expected?

Yes

Would you buy Sigma Computing again?

Yes

Other Software Used

Astro by Astronomer, dbt, Snowflake, GitHub, PostHog

Likelihood to Recommend

Sigma is well suited for the 80% of BI & reporting needs your company will have. If you already have a data engineering function, are leveraging a managaged extract & load tool (e.g. Fivetran,Stitch), and are transforming data in your warehouse (I.e. ELT via dbt), Sigma does the consumption piece very well. However, if you're a small analytics team with no BI or Data Engineering skills, it doesn't cover those needs, such as something like Domo would. Our stack of Fivetran/Snowflake/dbt/Airflow plus Sigma works very well for us.

Getting data into your users' hands quickly and easily

Overall Satisfaction with Sigma Computing