Video |

Tableau | Data prep

Column reordering enhancements for Prep - New in Tableau 2022.2

Tableau Prep finally preserves the column order you set inside the product, but the sorting only sticks when you tell your output to create the table.

Tim Ngwena •1 July 2022 •1 min read •8 min watch •5.6K

Part ofWhat's new in Tableau 2022.2 Tableau Prep Builder tutorial Jan-2020

Watch on YouTube

Tableau Prep 2022.2 now preserves the column ordering you set in the product when writing output, a feature requested in the community three years prior with 1,400 votes.
You can reorder columns either by dragging them in the data grid view or in the field summary view.
To make a changed column order overwrite an existing file's sorting, set the output to 'Create table' rather than 'Replace data', as 'Replace data' keeps existing columns and only refills the data.
Newly added outputs persist your current sorting, but pointing at an existing file can pull through that file's previously saved column order.
The feature supports Excel, CSV and Hyper files, plus Google BigQuery, Redshift, Snowflake and published data sources, but not SQL Server or Oracle.

0:00In previous versions of Tableau prep, prep

0:02did not observe the column ordering that

0:04you set up inside of the product.

0:07In 22.2, it does. Let's get stuck in.

0:10Here I've connected to a data source inside

0:12of Snowflake and essentially I've connected

0:15a sample data source.

0:16I've already got a data flow going with the

0:18summary table up.

0:20What you've typically done in the past to

0:22sort columns in this view is just to drag

0:25one of them

0:26and just wait for this cursor to sort of

0:27show you an area that you can place it in

0:30between a specific point and you drop it in

0:31.

0:32Now the problem in the past has been that

0:34when you then go and output this file to a

0:38particular file type, CSV, Excel, or to a

0:40database,

0:41that column ordering has not been observed.

0:44In 22.2, it is observed.

0:46That's something that's been asked for in

0:47the community feedback.

0:49In fact, if I go to the community forums, I

0:52'll just bring it up in another window here

0:54and put it in front of the screen,

0:55this should hopefully load this up.

0:58This was recommended three years ago by

1:00Jonathan Drummy, a Tableau visionary in the

1:03community.

1:04And essentially this has been voted 1.4

1:07thousand points.

1:09It's absolutely crazy. It's such an obvious

1:11thing as well.

1:12You know, just to just preserve the sorting

1:14that I give my data in the data structure.

1:17So now that this is actually coming out, it

1:19's great to see the Tableau listening.

1:21The cool thing about this is that you can

1:23do it in two ways.

1:25So let's just go back to this clean step

1:27number one over here.

1:29And if I go to the field summary view, you

1:32can also do it here.

1:34So if I just go and drag my phone list up,

1:39let's bring nation key to the very top and

1:41I'll put phone second.

1:43What we're going to do is we're going to

1:44output this to a CSV file, which will then

1:46open in Excel.

1:47So let's go ahead and do that.

1:50Let's actually just put it out to an Excel

1:51file straight away so I don't have to open

1:53it into anything.

1:54We'll just call this file.

1:58Create a new worksheet called file.

2:02Replace the entire data set every time we

2:04write this because we're going to open it

2:06more than once.

2:07And in terms of where to put this file, I'm

2:09going to put it inside of my recordings

2:12folder where I'm recording all the videos

2:13for this.

2:14So we'll put it in there.

2:16So we're going to go in the output folder

2:18and what we want to see is that nation key

2:20and phone key come up at the very top.

2:22You can see that's the order that I've just

2:24given it here.

2:25You can see it's also the order that I've

2:27got there.

2:28So we're going to see if that actually

2:29happens.

2:30Let's go ahead, hit run the flow.

2:32It will go out to the Excel file.

2:34While that is running, I'll just go ahead

2:36and get that folder open.

2:38So we will see that file turn up hopefully

2:40anytime soon.

2:42And yeah, we'll have a look at the actual

2:43file.

2:44In fact, this is a very large data set, so

2:46this might have been a very, very bad idea

2:48to do this, especially over Snowflake.

2:50So we'll wait and just let it finish.

2:53Maybe we'll hit the Excel row limit and we

2:54'll see what happens.

2:56So let's wait for this to finish and then

2:59we'll open it up.

3:02OK, it's finished.

3:03I did actually pause and restart this.

3:05I'll explain why in a second.

3:07It's gone to my default location,

3:09unfortunately.

3:10There it is in my data sources folder.

3:12So that's exactly where we'll go.

3:14We'll open up the file explorer.

3:15We go to my documents folder.

3:17We'll go to my Tableau repository, prep

3:22repository, data sources, output.xlx.

3:26I can even see from here that the sorting

3:28has been maintained because I can see the C

3:31Nation column there.

3:33And what I did is I filtered it down to

3:35just one row.

3:36And you can see that the sorting of the

3:38columns has indeed been observed.

3:41The two that I set there have come back.

3:44Now what I'm going to do is close this.

3:46We're going to change the sorting and we're

3:47going to run it again.

3:48It will hopefully overwrite that file.

3:50So let's go back in here.

3:52I think I actually changed this.

3:53Did I change this in the subsequent step?

3:55I think I did.

3:57I didn't change it in the subsequent step.

4:00I just changed it in here.

4:02So what we can do is go ahead and maybe put

4:04the address field at the top again.

4:07And we'll go to the output.

4:09You'll see that the address is now at the

4:10top.

4:11Hit Run Flow.

4:13It will overwrite the file.

4:15That's been done.

4:18Now there's something really interesting

4:19going on here because you can see that I've

4:21actually set address to be at the very top.

4:23But my output is still showing the Nation

4:25key, the old setup.

4:27So I think what might be going on here that

4:29it's caching something about the file

4:31because if I delete this and add it back in

4:37,

4:37I do see the updated sorting.

4:40So whenever I add a new output, the sorting

4:42actually persists.

4:43And if I just go ahead and output that to

4:46exactly the same Excel file,

4:49in this particular case, I think I called

4:50it a file.

4:51I wanted to literally overwrite the

4:53previous one, replace the data.

4:55Everything should be called output, output,

4:57output.

4:58Oh, interestingly, when we select that file

5:04, it pulls that previous sorting that we saw

5:10.

5:11It's not until I actually, if I give it a

5:13different file name, let's say file number

5:16two,

5:17create a new actually called file number

5:19two, then the new sorting persists.

5:23But if I name it after the original one,

5:28then it actually grabs the sorting that's

5:30already in that file.

5:32It seems it doesn't seem to grab it and

5:34then rearrange it.

5:36It doesn't sort of push what I've actually

5:39got here in the data.

5:41I've set it here to replace the data.

5:45Let's set create table and that might

5:46actually be the better thing to do.

5:49Okay, so that's, is that a knowledge gap

5:52for me? Probably is.

5:54If you look over here, the reason that wasn

5:55't actually working is because the setting

5:58here was not set to create a table every

6:01single time.

6:03I mistakenly assumed that replacing the

6:06data would do the same thing,

6:09but actually replace the data only creates

6:12the table if it doesn't exist.

6:14If it already exists, it doesn't create the

6:16table, it keeps the columns as they are and

6:19fills the data in,

6:20essentially drops the table and puts the

6:22data back in.

6:24Creating the table completely deletes

6:26everything and then puts it back in.

6:28So that's why the sorting works a little

6:30bit better.

6:31So I was going to say the sorting doesn't

6:32stick when you change the output,

6:34but you just need to make sure that you got

6:35this to create table for it to overwrite

6:37the existing sorting that was previously

6:40selected.

6:41Okay, so that's a really nice quality of

6:42life improvement.

6:44Now, the only file types this works with, I

6:46'm going to check my notes here, is Excel,

6:48CSV, Hyper.

6:50And then in terms of databases, Google, Big

6:52Query, Redshift, Snowflake, and then of

6:54course published data sources on Tableau

6:56Cloud or Tableau Server.

6:58So it's a new feature, but of course it

7:00doesn't work with absolutely everything.

7:03Microsoft SQL Server is not in that list,

7:04Oracle is not in that list.

7:06So hopefully it comes to more and more

7:07databases over time as they start to roll

7:09this out.

7:10But nonetheless, this is quite nice to have

7:12this happen, that it persists the column

7:15sorting for you inside of the product.

7:19Thanks for watching and I'll catch you in

7:21the next 2020.2 video.

7:23Bye.

7:24[silence]

7:30[ Silence ]

Rollup

Tableau Prep 2022.2 finally preserves the column order you set inside the product when it writes your output, instead of silently reverting to its own default order.

Tim connects to a sample data source and builds a simple flow with an output step. He reorders columns and writes to Excel to see whether the new sorting actually lands in the file.

The Breakdown

Why this matters 0:00

In older versions, however you dragged columns around inside Prep, that ordering was ignored when you wrote to a file or database. This was a long-requested community fix (three years old, 1,400 votes) and now the order you set is respected in the output.
Two ways to reorder columns 1:21

You can drag columns into place either in the data grid view or in the field summary view — whichever you use, the resulting order is what gets carried through to the output.
Write the output and check it 1:41

Set up your output step, run the flow, and open the resulting file to confirm your chosen column order came through. On a first write to a new output, the sorting persists exactly as set.
Changing the order later doesn't always overwrite 4:00

If you change the column order again and rerun into the same existing output, you may find the file still shows the old order — Prep seems to pull through the sorting already saved in that file rather than applying your new one.
Deleting and re-adding the output forces the new order 4:21

As a workaround, removing the output step and adding it back in does pick up your latest sorting — useful if you're stuck seeing stale column order.
The real cause: create table vs replace data 5:41

The setting that actually controls this is whether your output is set to 'Create table' or 'Replace data'. Replace data only creates the table if it doesn't already exist; if it exists, it keeps the existing columns and just refills the data, so your new order never overwrites it. Create table drops and recreates the table each time, which is why it reliably applies your new sorting.
Where this feature does and doesn't work 6:41

Column order preservation works for Excel, CSV and Hyper files, and for Google BigQuery, Redshift, Snowflake and published data sources on Tableau Cloud or Server. It doesn't yet work with Microsoft SQL Server or Oracle.

Worth Knowing

If you're overwriting an existing output file and the column order doesn't seem to update, check the create table vs replace data setting before assuming the feature is broken.
Pointing an output at a file that already exists can pull through that file's previously saved column order rather than your current sorting.
Not all connectors support this yet — SQL Server and Oracle are notable exclusions.

Use It When

Reach for this whenever downstream consumers of your Prep output (an Excel report, a BigQuery table, a published data source) need columns in a specific, predictable order rather than whatever order Prep happened to write them in.

How this Rollup was made provenance & method

A Rollup is drafted by AI from the video's transcript, then reviewed and edited by Tim. Everything used to produce this one is listed below — the model, the exact prompt, and the source video — so the process is transparent and reproducible.

Transcription: On-device — NVIDIA Parakeet v3 for recent videos, OpenAI Whisper large-v3 for earlier ones. The transcript never leaves the machine or gets published.
Drafting: Claude Sonnet 5 in the cloud, from that transcript.
Prompt: The exact Rollup prompt (v2) — the full system prompt, unedited.
Source video: Watch on YouTube
Drafted: 5 July 2026 at 09:38
Reviewed & edited: 5 July 2026 at 09:41 · by Tim Ngwena

Model + prompt + video is everything you'd need to recreate a Rollup like this yourself. The one thing we don't share is the transcript.

Rights. The video and its transcript are the property of TN Media Ltd. Unauthorised use or download is prohibited. © TN Media Ltd.