ATTN:!!!! Quick correction, in the video I talk about how we connect to a "View" so it doesn't re-query the entire table. I misspoke! Guess I was a little tired when I hit record 😅🤣! A view does re-query the entire dataset. Use a scheduled query instead to not re-query the entire table every time Looker Studio reloads. Let me know in the comments if you want a video on scheduled queries 👇👇👇
Great video, but from practical point of view I hope you can comment on date range, simply because Looker advantage is to have freedom using salect date filters and compare date filters. Your view has only 1 weeks date range.
Hey Albert, thanks for checking out the video. I think you're comment here is a great idea for another video on BigQuery and Looker Studio. Something like, "How to get a dynamic date range in BigQuery and Looker Studio". Thoughts on that?
Yes Albert! Any help on how to make more dynamic DataSets in BigQuery will be a great help. I created a lousy little report in Looker Studio yesterday, and within an hour I hit the quota 😞. I'm not too keen on spending money on 3rd party products to get my data out of BQ if I don't have to 🙂
Yes, ive been searching for somrthing like for a very long time. it would help make my visualisations more dynamic. I've been trying to find a way to get the graph to give total sales vs previous periods sales months wise view with datewise data points. By default, the graph shows current months sales vs previous months sales (I have the default value of date range controller set to yesterday). When the user selects lets say a date range from 18 jan 2023 to march 25 2023, i want the graph to show from beginning of jan to end of march. I hope ive made my scenario clear, but ive been looking for something like that for so long and havent found it yet. @@Vision-Labs
Appreciate the time & effort that went into creating this video, but it's impractical unless you want a 1 table dashboard with only 1 pre-defined time range.
Hey 👋 Yes, this was more of an example as opposed to a workflow. Definitely need to have a date column to partition and filter by in looker studio for a long term solution. -JJ
I want to asked, in regards to best practice, which is better : 1.Save the sql script into view, and connecting to looker studio, or 2.Save the sql script into table(with scheduling.) and connecting to looker studio ?
There are a lot of pros and cons to views. Basically a view is a “saved query” that runs when called upon A table is that saved outcome. Hope that helps! -JJ Ps if you are interested we talk about this inside of The Lab
Thanks for the video! Very interesting tricks and useful tips, thanks for this content! Its really rare. So many questions! :) What key do you prefer to use for blending data, is Session ID? For example Country by days, page location, source etc. or better to make one table with all of this parameters? P.S. And the biggest questions, and where no answers about Quotas with using Queries when you had 3M+ events per day and where you can check it.
Thanks for the comment :) Session ID is usually what we will blend data on. Then we will usually create tables for each type of scope that we're looking at. For example, we'll create 1 table that is all the session stuff. Session source, medium, campaign, purchases, pageviews, adds to cart, purchase value, etc. But everything is aggregated at the session id level. We'll create another table that is at the page level and have all the individual page stuff. Basically you have to create multiple tables and have multiple data sources for your reports. Typically our "standard" reports have 6-10 data sources. All of them are Different BigQuery tables. As for the cost of querying in BigQuery. There is a query that you can use directly inside of BigQuery to determine the cost of this. We'll likely do a video for this very soon because not many people know about it. Let me know if that helps. - Stockton
@@Vision-Labs Thanks a lot for the answer, it helps to understand how to work and manage data inside BigQuery. New meta is hard to configure, all new features, there are too many things that need to be settled up correctly :)
ATTN:!!!!
Quick correction, in the video I talk about how we connect to a "View" so it doesn't re-query the entire table. I misspoke! Guess I was a little tired when I hit record 😅🤣!
A view does re-query the entire dataset.
Use a scheduled query instead to not re-query the entire table every time Looker Studio reloads.
Let me know in the comments if you want a video on scheduled queries 👇👇👇
Scheduled Queries please 😎
Scheduled queries would be great bit also a video on queries to optimise looker studio quering speed, notoriously slow looker is 😅
Great Lesson! Thank you! Yes, I agree with Albert below. A video on how to optimize would be GREATLY appreciated.
Check out this new video: ua-cam.com/video/g1qcu3fJCCg/v-deo.html see if you like it :)
@@Vision-Labs Like? I LOVE IT! THANK YOU!
Great video, but from practical point of view I hope you can comment on date range, simply because Looker advantage is to have freedom using salect date filters and compare date filters. Your view has only 1 weeks date range.
Hey Albert, thanks for checking out the video. I think you're comment here is a great idea for another video on BigQuery and Looker Studio. Something like, "How to get a dynamic date range in BigQuery and Looker Studio".
Thoughts on that?
@@Vision-Labs yes, agreed, also important to explain in detail the benefits why you built that table innBigQuery and not queried from Looker.
Yes Albert! Any help on how to make more dynamic DataSets in BigQuery will be a great help. I created a lousy little report in Looker Studio yesterday, and within an hour I hit the quota 😞. I'm not too keen on spending money on 3rd party products to get my data out of BQ if I don't have to 🙂
Yes, ive been searching for somrthing like for a very long time. it would help make my visualisations more dynamic. I've been trying to find a way to get the graph to give total sales vs previous periods sales months wise view with datewise data points. By default, the graph shows current months sales vs previous months sales (I have the default value of date range controller set to yesterday). When the user selects lets say a date range from 18 jan 2023 to march 25 2023, i want the graph to show from beginning of jan to end of march. I hope ive made my scenario clear, but ive been looking for something like that for so long and havent found it yet. @@Vision-Labs
Appreciate the time & effort that went into creating this video, but it's impractical unless you want a 1 table dashboard with only 1 pre-defined time range.
Hey 👋
Yes, this was more of an example as opposed to a workflow.
Definitely need to have a date column to partition and filter by in looker studio for a long term solution.
-JJ
I want to asked, in regards to best practice, which is better :
1.Save the sql script into view, and connecting to looker studio, or
2.Save the sql script into table(with scheduling.) and connecting to looker studio
?
It depends on your use case.
For the minimum cost. Usually a table.
For more flexibility a view
Great videos! To get the unter day table I need to set bq in ga4 as streaming? Will it show me the current data? Without any delay
Yes, You will need both Streaming & Daily turned on!
- JJ
It would be really helpful if you explained succinctly the point of View and why people use it. Not finding it that clear.
There are a lot of pros and cons to views.
Basically a view is a “saved query” that runs when called upon
A table is that saved outcome.
Hope that helps!
-JJ
Ps if you are interested we talk about this inside of The Lab
Thanks for the video! Very interesting tricks and useful tips, thanks for this content! Its really rare. So many questions! :)
What key do you prefer to use for blending data, is Session ID?
For example Country by days, page location, source etc. or better to make one table with all of this parameters?
P.S. And the biggest questions, and where no answers about Quotas with using Queries when you had 3M+ events per day and where you can check it.
Thanks for the comment :)
Session ID is usually what we will blend data on.
Then we will usually create tables for each type of scope that we're looking at.
For example, we'll create 1 table that is all the session stuff.
Session source, medium, campaign, purchases, pageviews, adds to cart, purchase value, etc. But everything is aggregated at the session id level.
We'll create another table that is at the page level and have all the individual page stuff.
Basically you have to create multiple tables and have multiple data sources for your reports.
Typically our "standard" reports have 6-10 data sources. All of them are Different BigQuery tables.
As for the cost of querying in BigQuery. There is a query that you can use directly inside of BigQuery to determine the cost of this. We'll likely do a video for this very soon because not many people know about it.
Let me know if that helps.
- Stockton
@@Vision-Labs Thanks a lot for the answer, it helps to understand how to work and manage data inside BigQuery. New meta is hard to configure, all new features, there are too many things that need to be settled up correctly :)