5 ms·

Has anyone tried both this and Metabase? I've used Metabase in a few projects and I find it very nice. This seems more powerful, perhaps? Is it worth it for BI

by adlpz 3y ago

Has anyone tried both this and Metabase? I've used Metabase in a few projects and I find it very nice. This seems more powerful, perhaps? Is it worth it for BI on small datasets?

CalRobert 3y ago
Yes, I am at a company using Metabase, but I have a decent amount of experience with Superset (albeit from many years ago). The reason we chose Metabase was that it had table joins, while Superset doesn't (unless it has added them since I used it). It also looks a bit sleeker. But I strongly prefer Superset; I found that with Metabase I had to turn a lot of things off to make it usable (Let me see "the_table" not "The Table"!), I was constantly annoyed at the opacity around models vs "questions", etc. and every time I wanted to change a question Metabase insisted on creating a new one instead. The real issue here was when we wanted to swap out the data source for a lot of questions but there was no clean way to do so without MB just creating new questions. Also, Metabase doesn't have serialization unless you pay them AND you self-host, (if I'm self hosting then what exactly am I paying for?) and that's pretty annoying. https://www.metabase.com/docs/latest/installation-and-operation/serialization https://www.metabase.com/docs/latest/installation-and-operat.... But it does let you join tables. Sometimes that's enough to make MB worth dealing with.
- noduerme 3y ago
  The "model" vs "question" thing is really annoying as there's no real difference from the user's perspective, and it's easy to accidentally convert a model back to a question without noticing when you publish something. You notice when you try to drill into the chart. There's a lot of annoying manual labor in metabase, e.g. I want to filter something into 10 different charts and I need to duplicate it 10 times and change a filter on each one. Still yeah joins are nice. A non-bugged aggregate count/sum as a window function would be nicer.
- itsoktocry 3y ago
  >I was constantly annoyed at the opacity around models vs "questions" Yeah, somewhere along the line Metabase decided to get opinionated on "self-serve". I imagine it works well for some teams and companies, but for the tech-oriented, it's annoying. I prefer my BI tools to be platforms that make for easy charting and cross-filters, while I build and control the models behind the scenes with a tool like dbt.
- adlpz 3y ago
  Thanks! Very detailed answer. I've found the weird "make it easy" mindset a bit annoying with Metabase too. The whole questions, nice table names... I'll give Superset a try in my next project I think.
- rusackas 3y ago
  Superset lets you join tables within the same database. If you want to do cross-DB joins, we have a new (beta) in-memory meta-DB that lets you do this, but we generally see and recommend people using things like Trino for this.
  - Cilvic 3y ago
    Is that new? Last time I checked this was the major downside from superset
  - CalRobert 3y ago
    Nice! When was that added?
skadamat 3y ago
Metabase is a bit more user-friendly to be honest than Superset. Superset has a WAY more liberal license, so it's ideal for people who want to customize Superset and build data apps.
sokols 3y ago
Metabase is great, I use it with a Oracle Database.
code_biologist 3y ago
Reposting from a comment of mine about 60 days ago: I recently ran a little shootout between Superset, Metabase, and Lightdash — all open source with hosted options. All have nontrivial weaknesses but I ended up picking Lightdash. Superset is the best of them at data visualization but I honestly found it almost useless for self-serve BI by business users if you have existing star schema. This issue on how to do joins in Superset (with stalebot making a mess XD) is everything difficult about Superset for BI in a nutshell. https://github.com/apache/superset/issues/8645 https://github.com/apache/superset/issues/8645 Metabase is pretty great and it's definitely the right choice for a startup looking to get low cost BI set up. It still has a very table centric view, but feels built for _BI_ rather than visualization alone. Lightdash has significant warts (YAML, pivoting being done in the frontend, no symmetric aggregates) but the Looker inspiration is obvious and it makes it easy to present _groups of tables_ to business users ready to rock. I liked Looker before Google acquired it. My business users are comfortable with star and snowflake schemas (not that they know those words) and it was easy to drop Lightdash on top of our existing data warehouse.
- oliver101 3y ago
  > YAML, pivoting being done in the frontend, no symmetric aggregates (one of the maintainers of Lightdash) You touched on some of our most interesting problems here! Would be especially interested to hear about what you liked / didn't like about symmetric aggregates in Looker and how you find dev with YAML. If you have an idea of how you'd like these to look in Lightdash, the team would be really open to making that a reality. For pivoting in the backend, this is coming! Issue here: https://github.com/lightdash/lightdash/issues/2907 https://github.com/lightdash/lightdash/issues/2907

by adlpz 3y ago

Has anyone tried both this and Metabase? I've used Metabase in a few projects and I find it very nice. This seems more powerful, perhaps? Is it worth it for BI on small datasets?

CalRobert 3y ago

Yes, I am at a company using Metabase, but I have a decent amount of experience with Superset (albeit from many years ago). The reason we chose Metabase was that it had table joins, while Superset doesn't (unless it has added them since I used it). It also looks a bit sleeker. But I strongly prefer Superset; I found that with Metabase I had to turn a lot of things off to make it usable (Let me see "the_table" not "The Table"!), I was constantly annoyed at the opacity around models vs "questions", etc. and every time I wanted to change a question Metabase insisted on creating a new one instead. The real issue here was when we wanted to swap out the data source for a lot of questions but there was no clean way to do so without MB just creating new questions. Also, Metabase doesn't have serialization unless you pay them AND you self-host, (if I'm self hosting then what exactly am I paying for?) and that's pretty annoying. https://www.metabase.com/docs/latest/installation-and-operation/serialization https://www.metabase.com/docs/latest/installation-and-operat.... But it does let you join tables. Sometimes that's enough to make MB worth dealing with.

noduerme 3y ago
The "model" vs "question" thing is really annoying as there's no real difference from the user's perspective, and it's easy to accidentally convert a model back to a question without noticing when you publish something. You notice when you try to drill into the chart. There's a lot of annoying manual labor in metabase, e.g. I want to filter something into 10 different charts and I need to duplicate it 10 times and change a filter on each one. Still yeah joins are nice. A non-bugged aggregate count/sum as a window function would be nicer.
itsoktocry 3y ago
>I was constantly annoyed at the opacity around models vs "questions" Yeah, somewhere along the line Metabase decided to get opinionated on "self-serve". I imagine it works well for some teams and companies, but for the tech-oriented, it's annoying. I prefer my BI tools to be platforms that make for easy charting and cross-filters, while I build and control the models behind the scenes with a tool like dbt.
adlpz 3y ago
Thanks! Very detailed answer. I've found the weird "make it easy" mindset a bit annoying with Metabase too. The whole questions, nice table names... I'll give Superset a try in my next project I think.
rusackas 3y ago
Superset lets you join tables within the same database. If you want to do cross-DB joins, we have a new (beta) in-memory meta-DB that lets you do this, but we generally see and recommend people using things like Trino for this.
- Cilvic 3y ago
  Is that new? Last time I checked this was the major downside from superset
- CalRobert 3y ago
  Nice! When was that added?

skadamat 3y ago

Metabase is a bit more user-friendly to be honest than Superset. Superset has a WAY more liberal license, so it's ideal for people who want to customize Superset and build data apps.

sokols 3y ago

Metabase is great, I use it with a Oracle Database.

code_biologist 3y ago

Reposting from a comment of mine about 60 days ago: I recently ran a little shootout between Superset, Metabase, and Lightdash — all open source with hosted options. All have nontrivial weaknesses but I ended up picking Lightdash. Superset is the best of them at data visualization but I honestly found it almost useless for self-serve BI by business users if you have existing star schema. This issue on how to do joins in Superset (with stalebot making a mess XD) is everything difficult about Superset for BI in a nutshell. https://github.com/apache/superset/issues/8645 https://github.com/apache/superset/issues/8645 Metabase is pretty great and it's definitely the right choice for a startup looking to get low cost BI set up. It still has a very table centric view, but feels built for _BI_ rather than visualization alone. Lightdash has significant warts (YAML, pivoting being done in the frontend, no symmetric aggregates) but the Looker inspiration is obvious and it makes it easy to present _groups of tables_ to business users ready to rock. I liked Looker before Google acquired it. My business users are comfortable with star and snowflake schemas (not that they know those words) and it was easy to drop Lightdash on top of our existing data warehouse.

oliver101 3y ago
> YAML, pivoting being done in the frontend, no symmetric aggregates (one of the maintainers of Lightdash) You touched on some of our most interesting problems here! Would be especially interested to hear about what you liked / didn't like about symmetric aggregates in Looker and how you find dev with YAML. If you have an idea of how you'd like these to look in Lightdash, the team would be really open to making that a reality. For pivoting in the backend, this is coming! Issue here: https://github.com/lightdash/lightdash/issues/2907 https://github.com/lightdash/lightdash/issues/2907