Frequently asked questions (FAQ)

These are the frequently asked questions we hear, as well as the answers. If you have other questions, please do not hesitate to contact us!

Sigma is drastically different with a unique architecture and focus on different use cases.  Sigma focuses on self-service, ad-hoc data exploration and analysis at scale (but we do also offer dashboards and visualizations). Traditional BI tends to focus on dashboards and reports on limited data sets that answer limited questions.

Sigma can uniquely do data exploration and analysis at Cloud Data Warehouse (CDW) scale because of our points of differentiation from traditional BI, including:

  • We are a SaaS solution purpose-built for the CDW and we push queries to the CDW for execution against live data to the lowest level of detail.
  • We have an intuitive, spreadsheet-like UI and collaborative data canvas anyone can use.
  • We do not require data to be heavily modeled to support unanticipated questions. We enable end users to freely explore and analyze data.
  • We use the power of the CDW for near-unlimited scale and speed. Also, where possible, we use the local browser to instantly perform certain operations and queries.

Other BI tools tend to be the opposite. They struggle with ad-hoc analysis and require data to be highly modeled to explore or analyze it. They have interfaces that are usable for data exploration only by highly-technical users. They often have on-premise components that limit scale and speed, rely on stale and risky data extracts, and require significant maintenance. Lastly, with many of these tools, for deeper analysis the end user ultimately has to export data out of the tool and convert it into spreadsheets where they then encounter spreadsheet limitations such as stale data and limited scale.

Reporting/dashboards primarily focus on showing the historical  “what has happened” and are great for showing KPIs and high-level trends for different areas of the business.

Exploration/Analytics are more focused on drilling down to the lowest possible level of detail, a record or row, and iterating on the data to learn the “why something happened.” The result of analytics are ad-hoc, new insights to understand and improve business performance.  This insight typically is obtained by performing actions like joining data, running formulas and calculations, performing segment/cohort analysis, building pivot tables, doing projections, and performing scenario or ”what if” modeling.

It depends and is based on each customer’s situation, use cases, and requirements. We can complement traditional BI tools where they are used for dashboarding and reporting. Sigma addresses gaps they have around data exploration & analytics, self-service, ease-of-use, scale/speed and team collaboration. But for some organizations, especially ones building a modern data stack from scratch, Sigma can be their single BI & analytics product as we also offer robust reporting and visualizations.

From Day 1 Sigma has been 100% fully-managed SaaS and accessed via a simple URL. We are not on-premise and we have no plans to be so, as we are built to take advantage of the scale and efficiency that only the cloud enables.

No. Sigma converts user actions in the UI into optimized, machine-generated SQL which is pushed down to the CDW for execution against data in the CDW. So table data never leaves the CDW and it is queried in place.

Benefits of this include:

  • Queries are always run against live/accurate data in the CDW (not stale data extracts).
  • Queries run fast as they use the near-unlimited scale/speed of the CDW.
  • Minimal security/governance risk because no data ends up as local desktop extracts/spreadsheets where it could eventually lead to inappropriate exposure or loss  of sensitive data.

Essentially none. Since we push optimized SQL queries to the CDW, we use the near unlimited scale and speed of the CDW and the underlying cloud provider platform. If sized correctly, the CDW has essentially no limits on compute or storage and we benefit from that. We have run performant queries against hundreds of billions of rows of data…yes hundreds of billions. We can also handle wide tables with hundreds of columns. Also, where possible, we use the local browser to instantly perform certain operations and queries.

Conversely, Excel has a hard limit of approximately 1.1M rows and we typically see on-premise BI tools struggling at 100M+ rows of data.

Sigma, like a traditional spreadsheet:

  • Lets business users view and explore data in a table-like format, perform calculations with formulas and spreadsheet-like functions, and turn analysis into charts and visualizations.
  • Brings together data, analysis, visualizations, text and images in a single document with multiple tabs
  • Does not require knowledge of SQL or any coding to use

Sigma differs from spreadsheets because it is a direct connection, or “window,” to all your live data in your Cloud Data Warehouse (CDW) and uses its unlimited scale. So Sigma does not have the scale limits of spreadsheets and can quickly join and analyze up to hundreds of billions of rows of data. Since Sigma does not extract data from your CDW, you always query accurate and live data, unlike most spreadsheets which work off stale data. Lastly, unlike on-premise spreadsheets, Sigma is in the cloud so:

  • Security and roles-based access control are easy to enforce from a single connection point.
  • Team collaboration and reuse of analysis is easy.

Yes. Sigma offers a wide range of customizable visualizations, charts and dashboards that can visualize the results of data analysis in Sigma and be shared or exposed to others to drive further insight.

Chart types include: line, pie, donut, stacked bar, area, scatterplot, map, funnel, single value, dual axis, gauge, and trellis.  Pivot and data tables, and custom text and images can also be added to dashboards. All these elements can be added to a Sigma dashboard with a high-level of customization, including multiple fonts, colors, dashboard elements, controls, and levels of interactivity to choose from. Interactivity includes the ability for an end viewer to click on a dashboard visualization to then view and explore the underlying, detailed data to answer more questions.

Sigma’s cloud architecture enables the easy, controlled sharing and reuse of live data objects, whether workbooks or datasets. This lets teams work off the same, live data to unlock the collective power and insight of the team for faster, better decision making. Some highlights:

  • Data collaboration canvas where tables, charts, and text can be unified and commented on to tell a live data story that can easily be collaborated on with others
  • Easy, one-click ability to share any data object with other users.
  • Workspaces and Folders make it easy for teams or individuals to find relevant data and analyses.
  • Workbook comments facilitate group collaboration on analysis and visualizations.
  • Granular role-based access control ensures data and analyses are shared appropriately.

Yes. Sigma offers embedded analytics where visualizations and dashboards created via the Sigma UI can be securely embedded into other internal, external, or custom applications to put relevant data in front of users or customers or partners in a manner that is easiest for them to view it. Like other Sigma dashboards, these embedded dashboards offer customizable themes and layouts, are interactive and dynamic, and are powered by live data in the cloud data warehouse.

Unlike most other ABI tools, with Sigma embedded visualizations can optionally be clicked into by the end viewer to get to the detailed, underlying data to do more data exploration and analysis.

The three main options for embedding are application embedding (where end viewers only see their data and do not have to log into Sigma), private embedding, and public embedding.

It could be within minutes. This is because Sigma is SaaS, can connect to your cloud data warehouse (CDW) tables in seconds, and does not require data modeling to query the CDW (although data can be modeled/curated with Sigma). So within minutes we often see new customers running queries on live data in their CDW to get to new insights that generate value.

That said, the precise answer is “it depends”, but typically the business value comes within hours or days. The answer is dependent on variables such as: how well defined are your analytical use case requirements (datasets, queries, dashboards, etc), how much of the data needed for these is already in the CDW, and how “clean” this data is. Regardless of any dependencies, Sigma and our Customer Success team will be there to help.

Sigma has robust Administrative capabilities to limit who can do what with specific data. Capabilities include:

  • Data permissions to limit who can access CDW connections, databases and tables, and also Sigma objects like datasets and workbooks. Permissions can be applied for individual users or teams, and each user is assigned an account type that limits what they can do in Sigma. Besides several built-in account types, Admins can also create custom account types with specific features and functionality enabled for certain groups of users.On the most restrictive end of the spectrum, some user groups could be “view-only” and against specific visualizations. There is zero chance of them inappropriately accessing data or editing objects. On the other end of the spectrum, architects on the Business Intelligence team could have the ability to do anything in Sigma, including connecting it to CDWs, building datasets, and editing workbooks.
  • Workspaces can be set up that are private and require an Admin or Team Admin to invite in members. A Workspace can have relevant workbooks and datasets in it. They help ensure users see data on a “need to know” basis.
  • Multiple authentication options including passwords or integrations with an Identity Provider (IdP) so user management (authentication, authorization, de-provisioning, etc) is done centrally in the IdP. IdP integration options include SAML or OAuth, and optionally SCIM to extend them for real-time metadata syncing between the IdP and Sigma.

Sigma offers flexible, intuitive data modeling, or data curation, options. On one end, advanced users with appropriate permissions can directly query raw table data in the Cloud Data Warehouse (CDW) with no data modeling required at all. Or, on the other end, data can be curated, joined, and enriched into user-friendly “datasets”. These datasets save Business Intelligence teams time by making it easy to pre-model and curate data for non-technical users in a way they understand, so business users have guided access to modeled data which drives faster, better analytics. Datasets are powered by live data so they are always accurate, can be easily reused for use in multiple workbooks, and optionally can be materialized in the CDW for better performance and so other tools can leverage them.

Yes. Via the Sigma interface you can access the “SQL Runner” interface which lets you write and execute “manual” SQL directly from Sigma. As you execute queries, the results show up on the bottom half of the SQL Runner interface. This SQL is also pushed to the CDW for execution.

Yes. For any CDW Sigma supports, we can easily extract JSON from a CDW table column. The extracted data can then be analyzed in a Sigma workbook.

Yes. Sigma can address some data validation use cases to accelerate your CDW deployment. Because Sigma is essentially a window into your CDW, we can easily give you visibility to all the data in your CDW, even if limited, to understand what data is in it, what tables and columns the data is in, and the structure of the data. This visibility can help you determine if the data is in the desired location with the expected format, quality, and level of accuracy.

Sigma can help you identify possible data quality issues in the CDW in a variety of ways including:

  • Filtering data to identify “nulls.”
  • “Count distinct” queries, sorts, or grouping of data to look for repetitive data.
  • Creating visualizations, including scatter plot charts, to visually surface data outliers and anomalies.

Based on this visibility, you can take appropriate steps as needed to address any data validation or quality issues. Sigma can potentially help here as well, as Sigma can take raw data in your CDW and curate, join, and enrich it as needed into “clean,” user-friendly datasets so all your users have guided access to quality data they can immediately explore and analyze. Net result is Sigma can help accelerate getting data in your CDW and in the appropriate format so users can get to insight and value faster.

Possibly. It depends on the use case/activity and if this aligns to Sigma capabilities. A lot of data science work is relatively straight-forward data preparation and analysis we can help with. For example, Sigma could help if your data science team wants to:

  • Explore data in the Cloud Data Warehouse via SQL for ad-hoc data exploration/analysis.
  • Curate/refine CDW data and then export it out of the CDW so a data science product can analyze it with Python, R, Spark, etc.
  • Write the results of data science analysis back into the CDW, visualize this in dashboards, and make it available for exploration and analytics by business users.
  • Free up time for higher-value tasks by having less technical users self-serve for basic data analytics that the data science team otherwise would have to do in a BI product.

Possibly. It depends on the use cases/activities and if they align to Sigma capabilities. For example, Sigma could help if your data engineering team wants to:

  • Be able to quickly explore data in the Cloud Data Warehouse (CDW) down to the lowest level of detail without writing SQL queries to understand what data is in the CDW,  and in what structure. This could be for data validation purposes to check and validate the accuracy, clarity, and detail of data.
  • Model (join, reduce, enrich, etc.) data in the CDW and then optionally materialize this modeled data back to a CDW table so any application can query it for better performance.
  • Free up time for higher-value tasks by having less technical users self-serve for  basic data engineering work the data engineering team normally would do, such as joining data or uploading new data into the CDW via a CSV upload.