# Is it ok to join marts tables ?

**URL:** <https://discourse.getdbt.com/t/is-it-ok-to-join-marts-tables/7345>\
**Category:** In-Depth Discussions\
**Tags:** modeling\
**Created:** [March 16, 2023, 1:09pm UTC](https://discourse.getdbt.com/t/is-it-ok-to-join-marts-tables/7345 "2023-03-16T13:09:09Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![greenglass](https://avatars.discourse-cdn.com/v4/letter/g/d2c977/32.png) [@greenglass](https://discourse.getdbt.com/u/greenglass)\
**Post date:** [March 16, 2023, 1:09pm UTC](https://discourse.getdbt.com/t/is-it-ok-to-join-marts-tables/7345/1 "2023-03-16T13:09:10Z")

</div>

Let’s suppose I have on my marts layer the following tables:

- dim\_customers
- dim\_accounts
- fct\_transactions

I want to create a report that joins all these tables. Therefore my query will contain the 3 tables.

BUT I do not want to put the query in my visualisation tool. Instead I would like to create a table/view that joins these tables from the marts folder and my visualisation tool would simply do a select \* from “joined\_table”.

Should I create that “joined\_table” in the marts layer by joining the dim and fct tables ?

How would you approach this scenario and what is your current approach ?

Thanks in advance

---

<div class="post-metadata">

**Author:** ![a\_slack\_user](https://sea2.discourse-cdn.com/flex020/user_avatar/discourse.getdbt.com/a_slack_user/32/1629_2.png) [@a\_slack\_user](https://discourse.getdbt.com/u/a_slack_user)\
**Post date:** [March 16, 2023, 2:33pm UTC](https://discourse.getdbt.com/t/is-it-ok-to-join-marts-tables/7345/2 "2023-03-16T14:33:45Z")

</div>

We do lots of this on my team. We have a separate directory called /datasets/ which is downstream of /marts/ where we put our joined and pre-aggregated datasets for our BI users to benefit from. This works well with our BI tool, Sigma, which has a presentation-layer concept of “Datasets” where we do the friendly naming, set column formats, add descriptions, etc. We materialize our datasets as tables in dbt. 95% of the time our BI tool is querying a flat table for a given viz/chart. We find it to be extremely fast, and our non-data expert end users get a lot of value. Happy joining!

Note: `@Joe Bloom` originally [posted this reply in Slack](https://getdbt.slack.com/archives/CBSQTAPLG/p1678977217763549?thread_ts=1678972173.803999&cid=CBSQTAPLG). It might not have transferred perfectly.