T4 lecture: Publishing and portfolio¶

Learn what happens to your work and your data when you publish to Tableau Public, decide what is safe to publish, keep your workbook versions in order, and turn one dashboard into a short story that a stranger can follow: question, evidence, limitation, action.

In Session 15 you built your capstone: one decision, evidence from at least three domains, and a validation pack. In Session 16 you learned to brief an executive: recommendation first, so-what titles, numbers with context and honest uncertainty. Today you learn how to show that work in Tableau, and how to share it without sharing anything you should not. In the T4 lab you build a portfolio piece from your capstone: one dashboard and a three-point story, saved on your own computer.

What you will be able to explain¶

What Tableau Public makes public when you save or publish to it: the views, the workbook and the data.
Which data is safe to publish, and which data you must never publish.
What hiding a viz and turning off downloads do, and what they do not do.
How to keep versions of a .twbx file with dated file names and a change log.
How a Tableau story and its story points guide a reader through an argument.
How to write a portfolio piece in four parts: question, evidence, limitation, action.

Part 1: What Tableau Public makes public¶

So far you have saved every workbook on your own computer as a .twbx file. Nobody else could see it unless you sent it to them. Tableau Public is different: it is a free website for sharing visualisations with everyone.

Tableau's own Tableau Public FAQ says it plainly: workbooks and data published to your Tableau Public profile are not private and are freely accessible to anyone. The viz can be seen by anyone on the internet.

When you publish, Tableau sends three things, not one:

What is published	What a stranger can do with it
The views: your dashboards, sheets and story	Open them in a browser, click, filter, read every label and tooltip, take screenshots
The workbook (`.twbx`)	Download it and open it in their own copy of Tableau, unless you turn downloads off
The data inside the workbook	Download it with the workbook, unless you turn downloads off. Publishing creates an extract of your data, which can hold every row and field of the data source, not only what your views show

Saving in the web editor is publishing

If you work in the Tableau Public web editor in a browser, saving puts the workbook and its data on your Tableau Public profile. There is no "save privately" in the web editor. In Tableau Desktop Public Edition you can choose: File → Save As keeps the work on your computer; saving to Tableau Public publishes it.

Turning off downloads¶

The FAQ says you can turn off the option to download the workbook and its data in the viz's settings on the Tableau Public website, and set the default for new vizzes in your profile settings.

Turning downloads off helps, but it does not make the data private:

Every number shown in a view, a label or a tooltip can still be read, copied and screenshotted.
Filters and actions let a viewer step through the data, one selection at a time.
Anything already downloaded before you changed the setting stays downloaded.

Hiding a viz¶

You can hide a viz on your profile. The FAQ is clear about what that means: visitors to your profile page cannot see the hidden viz, but anyone who knows (or stumbles across) its URL can still see it.

There is no private option on Tableau Public

Tableau's FAQ describes no private or unlisted way to publish. Hiding only removes the viz from your profile list; the viz and its data stay on the public website. Tableau's own advice is to use Tableau Public only for data that is suitable for a public audience. If data must stay private, keep the workbook on your own computer and never publish it.

Sources checked on 19 October 2026: Tableau Public FAQ and Save workbooks to Tableau Public. Tableau changes its website; read the FAQ again before you publish.

Part 2: What is safe to publish¶

Use one rule: publish only data that you would be happy to print on a poster in the street, and that you have the right to share.

Data	Publish?	Why
Data from your workplace: sales, customers, suppliers, budgets, staff	Never	It belongs to your employer. Even totals can reveal business secrets, and you almost never have permission
Customer or employee data, even with names removed	Never	People can often be recognised from a few fields (department, age band, start date). Removing names is not enough
A dataset you found online with no licence or unclear terms	No	Being downloadable is not the same as being allowed to republish
Superstore, Tableau's sample file	Not in this course	Tableau invites learners to download it and practise, but the course has not found a clear permission to republish the file. Keep Superstore work on your own computer
The course's synthetic cases: marketing-v1, finance-v1, operations-v1, hr-v1	Only once the course confirms its reuse terms	Generated by the course for fictional organisations; no real company or person. Ask your facilitator first. If they agree, label the data as synthetic and link to its source
Openly licensed public data whose licence allows sharing	Yes, with care	Follow the licence: name the source and keep any required notice

Synthetic does not mean anything goes

A dashboard on synthetic data can still mislead. Say clearly on the page that the data is synthetic and describes a fictional organisation, give its units (currency units, never a real currency) and its period, and never present it as a real company's results.

The privacy checklist¶

Before anyone in this course publishes, they complete this checklist. Every line must be yes.

Every data source in the workbook is synthetic course data or openly licensed public data. (Check Data → Edit Data Source for each one; a forgotten second data source is published too.)
No data source holds workplace, customer or employee data, or anything copied from them.
Superstore is not in the workbook, not even in an unused sheet.
The page says the data is synthetic (or names the public source and its licence), with units and period.
Nothing on the page names a real person, a real company or your employer.
I have decided whether to allow downloads, and I know that the views stay public either way.
I understand that hiding the viz will not make it private, and that a downloaded copy cannot be recalled.

Part 3: Keeping versions of your workbook¶

Tableau Desktop Public Edition does not keep a history of your local file. If you overwrite capstone.twbx with a broken version, the old one is gone. So keep versions yourself.

Dated file names¶

Each time you reach a milestone, choose File → Save As and give the file a new name:

T4_portfolio_YourName_2026-11-20_v01.twbx
T4_portfolio_YourName_2026-11-22_v02.twbx
T4_portfolio_YourName_2026-11-24_v03_submitted.twbx

Write the date as year-month-day, so the files sort in date order.
Add a version number so two saves on one day stay apart.
Mark the version you submitted, and never overwrite it.

.twbx, not .twb

A packaged workbook (.twbx) holds the data with the views. A plain .twb file only points to the data on your computer: send it to someone else and the views break. Keep saving as Tableau Packaged Workbook (*.twbx).

A change log¶

A file name says when; a change log says what and why. Keep it as a short table in your evidence log:

Version	Date	What changed	Why	Checked
v01	2026-11-20	First dashboard: tiles, trend, scope note	Capstone outline	Tiles reconcile with the Python register
v02	2026-11-22	Changed the headline; added the limitation story point	Peer feedback: headline was a topic	Titles true for every selection

The Checked column matters most. It records what you verified after the change, as in the validation pack of Session 15.

Part 4: Stories and story points¶

A dashboard shows everything at once and lets the reader explore. A story takes the reader through an argument one step at a time. In Tableau, a story is a separate tab made of story points; each point shows one sheet or dashboard with a short caption.

	Dashboard	Story
Reader	Looks at everything, chooses where to look	Follows your order, point by point
Best for	Monitoring and exploring	Explaining one decision
Text	Headline, titles, scope note	A caption for each point, plus the titles

Building a story¶

Choose Story → New Story. A new tab opens with the Story pane on the left.
Double-click the story title at the top and write your headline.
Drag a sheet or dashboard from the Story pane onto the empty area. This is your first story point.
Click the caption box above the view and write the point's message in one sentence.
Click Blank (under New story point in the Story pane) to add the next point, and drag the next view in. Click Duplicate instead to keep the same view and change a filter or highlight.
If you change a filter or a selection in a point, click Update above the caption to keep the change in that point.

Three points are enough for a portfolio piece

A good short story follows the question → evidence → limitation and action order. More points do not make a stronger argument; they make a longer one.

Part 5: Writing a portfolio piece¶

A portfolio piece is read by someone who has never seen your course: a hiring manager, a reviewer, a classmate next year. They give it a minute. In that minute they must understand four things.

Part	Question it answers	Example (operations case, Session 16)
Question	What decision did this analysis support, and for whom?	Should the operations manager review the carriers before next year's peak season?
Evidence	What does the data show, with numbers in context?	December promises were kept far less often than in January to October, and one carrier accounts for most of the misses. (Add the rates, with their denominators.)
Limitation	What can this data not tell us?	The data shows which promises were missed, not why. It is synthetic data for a fictional retailer.
Action	What should happen next, who owns it, and how do we know it worked?	Review the carrier's peak capacity before November; track the December on-time rate.

This is the pyramid from Session 16, Part 1 in a shorter form. The rules you already know still apply:

So-what titles on every view, true for every selection (Session 16, Part 2; T3, Part 6).
Numbers with context: denominator, period, comparison, unit (Session 16, Part 3).
No false joins. Your capstone datasets are separate fictional organisations. Never relate or blend two of them in Tableau, and never explain one with another (Session 15, Part 5). Put them side by side, each on its own data source.
No causal claims the data cannot support.

A portfolio piece is not the whole capstone

Pick one finding that carries your recommendation, and tell it well. Your full capstone stays in your notebook and your brief. The portfolio piece links to them in words: "Full analysis and checks: capstone notebook, Session 15."

Quiz¶

Answer these at the start of the T4 lab. Your facilitator will go through the answers.

You publish a dashboard to Tableau Public and turn off downloads. Name two ways a stranger can still learn what is in your data.
A friend says: "I hid the viz on my profile, so it is private now." What is wrong with this, according to Tableau's FAQ?
Your manager asks you to build a quick dashboard of last quarter's team sales and "just put it on Tableau Public so I can see it on my phone". What do you say, and what do you do instead?
Why may a dashboard on the course's synthetic marketing data be published only once your facilitator confirms the course's reuse terms, while one on Superstore is not, in this course? What must the published page say about the synthetic data?
You have files named final.twbx, final2.twbx and final_REAL.twbx. Name two problems, and rename them following the lecture.
Write the four parts of a portfolio piece for this finding: "Display and Video campaigns returned less attributed revenue than they cost in 2025 (synthetic marketing-v1)." One sentence per part.