Overall Satisfaction with Sigma Computing
We use Sigma across the entire company to conduct all our business intelligence research. We have lots of dashboards to track company and team level metrics of all sorts, as well as ad hoc sheets to answer specific questions. Pretty much all our understanding of our business besides event-based analytics (for which we use Mixpanel) comes from Sigma. We migrated from Looker.
- Segmenting data.
- Building dashboards.
- Incredibly powerful slicing and dicing of input data with its "levels" system.
- Building visualisations.
- Ability to create ad hoc data sets to share between multiple sheets allowing you to quickly update shared information/criteria.
- Complicated queries can be slow.
- More customization of visualisations, especially the number metrics.
- Ability to collaborate on a shared sheet.
- Ability to more easily create pivot columns within a sheet.
- Better understanding of customer segments.
- Ability to understand performance of our business.
- Creating audit-level views of financial information about our platform.
- Easily create and share dashboards for full team and company monitoring.
Allowing non- or less-technical partners to self-serve questions saves our Product and Data team immense amounts of time, and/or allows us to answer questions that might have been dropped had we not had such a powerful tool. The insights this provides is invaluable in helping us prioritize new features and understand the impact of past ones, as well as dig in to understanding issues that arise with customers.
I haven't done all that much with the templated dashboards myself to be honest, building mostly pretty standard/static ones, but the ability to have variables carry through from a dashboard to the underlying sheet lets us do a lot of nice things with date ranges and the like. That said, even if we're not using the most powerful features, the dashboards we have built, which are many, look really nice and understandable, and are easy to build. The fact that I've never bothered with their templates I think speaks to how easy it is to just build them myself.
This is probably the biggest single gap to me, is that a sheet, while anybody can view it or duplicate it, can't be modified by another user, which means that anything I have embedded in a dashboard, for example, can only be modified by me. I understand this is a very hard problem and realistically I probably would be the only one modifying my own dashboards and analytics most of the time, but occasionally we'll build reports for the finance team for audit purposes and me letting the finance team modify the fields I've provided them easily themselves would be handy. This is something I probably won't do that often, realistically, but definitely want.
On the flip side, though, the ability to make a sheet viewable by everyone, including those outside my organization, is really powerful and something we've only begun to really leverage but believe will really be powerful for us as a platform - creating custom, unmodifiable, live reports for our clients is a huge opportunity for us.
On the flip side, though, the ability to make a sheet viewable by everyone, including those outside my organization, is really powerful and something we've only begun to really leverage but believe will really be powerful for us as a platform - creating custom, unmodifiable, live reports for our clients is a huge opportunity for us.
I feel like I'm beating the same drum here, but Sigma's access to all of our data is incredible and I cannot recommend it enough. The usability for non- and less-technical people to build their own sheets is awesome, allowing us to better understand our customers and our business without bogging down a team of data scientists with every random ad hoc query and a bunch of process. Having used a number of other BI tools, Sigma is by far my favorite.
Sigma really was the powerful Swiss army knife for parsing and slicing all our data. A bit of the dark horse in our evaluation, against larger players (Tableau, Amazon) and the incumbent Looker, Sigma was at first almost overlooked but then quickly became my favorite by far, for the sheer power of its level-based sheet building and its reusable, custom data sets. While not the fastest of the set - mostly for Sigma's letting you build incredibly complex queries without you necessarily realizing it - it was by far the most powerful, at least for the average non-data-scientist user. Ability to create your own join tables, and store those joined tables as data sets to use across multiple sheets, was absolutely a game changer for those of us who don't/won't/can't write SQL and build our own tables in our warehouse. (Some of the performance issues we have seem to come from it liking Snowflake better than Redshift, for what that's worth. I don't know why our organization is attached to Redshift over Snowflake, but I'm not on the data team, either.)