Quick Answer: how an analyst gets a working Tableau dashboard in two weeks
Give it ten working days at about an hour a day, one focused video and one build task each, ending with a published dashboard built on your own data. You do not need a 21-hour course or a certification. You need clean data, a few calculations, one layout with working filters, and a safe place to publish.
That is the whole plan. Days 1 to 5 build the foundations: orientation, connecting tables, the blue and green pill logic, calculated fields and table calculations. Days 6 to 10 build the dashboard itself: LOD expressions, parameters, KPIs and layout, filter actions, then publishing it somewhere your stakeholders can open.
Two rules make it work. Every day ends with something built on your own data, not the tutorial's sample file. And every video below was picked because its title names exactly the job of that day, so you watch it once, close it, and build.
Why Tableau is not Excel with nicer charts
In Excel you build a chart from cells you have already arranged: a pivot table, a range, a column of totals. In Tableau you never arrange the numbers first. You drag fields onto shelves, and Tableau decides how to group and summarise the rows. Learn that shift on Day 1.
This is why strong Excel people often struggle in week one. You go looking for the place to type a total and there is none. Instead, every field in your data is sorted into one of two kinds when you connect. Dimensions are the things you slice by: region, product, customer, month. Measures are the things you count or add up: revenue, units, cost, tickets.
The second surprise is aggregation by default. Drop a measure like revenue into a view and Tableau does not show you each row. It shows SUM(Revenue), added up for whatever dimensions are already in the view. Add region and you get revenue per region. Add month and you get revenue per region per month. Same field, different answer, because the view changed.
The third surprise is colour. Each field you drag becomes a pill, and pills are either blue or green. Blue means discrete, which Tableau draws as separate headers or labels. Green means continuous, which Tableau draws as an axis. This is not the same thing as dimension vs measure, and mixing the two ideas up is behind a lot of week-one confusion. Day 3 is spent on exactly this.
One thing does not carry over from Excel at all: fixing a number by hand. If a total looks wrong in Tableau, the fix belongs in the data source or a calculated field, never typed over the chart. That feels slower the first time. By the second review it is why the dashboard updates itself.
If your Excel is shaky on pivot tables or lookups, spend an evening on our one-week Excel plan for data analysis first. Everything below assumes you are comfortable in a spreadsheet.
Before Day 1: pick one question and one dataset
Before you open a single video, write down the one question your dashboard must answer at the review, such as "which regions are behind target this quarter", and export the data that answers it. Without your own question and your own data the plan turns into tutorial-watching, and tutorials do not ship dashboards.
Export the data as a plain Excel or CSV file into one folder. Two or three tables is ideal: a main table of transactions or tickets or leads, plus one or two lookup tables such as a region list or a target sheet. Do not tidy them by hand, because the point is a dashboard that refreshes when next month's export lands in the same folder.
Also save whatever report your stakeholders read today, the Excel sheet or slide that already answers part of your question. That is your answer key. By Day 5 your Tableau totals have to match it, and any difference is either a join problem or an aggregation problem until proven otherwise.
Finally, get the software. Per Tableau's download page, checked October 2026, Tableau Desktop Free Edition needs no trial, expiration or credit card. Ask IT on Day 1 whether your company already gives analysts a licensed version, because licence requests can take longer than the build.
Week 1: build the foundations on your own data
Week 1 is five days of foundations with little to show a stakeholder: orientation, connecting your tables, blue and green pills, calculated fields and table calculations. It is the week people want to skip, and it is the reason Week 2 takes hours rather than days. Each day still ends with a real sheet built on your own data.
Day 1 (Monday): get oriented with one end-to-end video
Watch Alex The Analyst's Learn Tableau in Under 2 hours | Dashboards, Relationships, Visualizations + Mini Project (Alex The Analyst, 1:30:10). It is enough because today's job is orientation: where the data source page, the sheets and the dashboards live, how to connect a file, and what a finished dashboard looks like. Every deeper piece gets its own day.
At an hour and a half it is the longest video in the plan, so it is fine to spread it over Day 1 and the start of Day 2. Follow along with the video's sample data for the first half if that helps, then switch to your own export for the second half.
On your own data today: connect your main table, drag one measure and one dimension into a sheet, and check the grand total against your answer key. If it does not match, find out why now, while the workbook has one table in it.
Day 2 (Tuesday): connect your tables with relationships, joins or blends
Watch Tableau's Relationships vs. Joins vs. Blends in Tableau (Tableau, 20:45). It is enough because the official channel explains all three ways to combine tables side by side, and choosing the wrong one is the most common reason a beginner's totals come out doubled or missing.
If it does not click the first time, sqlbelle's Tableau Blend vs Join vs Relationship (Data Model) Explained - Part 1 (sqlbelle, 16:31) covers the same ground from a different angle. One of the two will make sense; you do not need both.
On your own data today: add your lookup table to the data source and relate it to your main table on the shared field, such as region code or product ID. Make sure that field has the same data type on both sides. Then rebuild yesterday's total and confirm it has not changed. If it grew, your tables are combining in a way you did not intend.
Day 3 (Wednesday): learn why some pills are blue and some are green
Watch OneNumber's Discrete vs. Continuous in Tableau (Blue or Green?) (OneNumber, 10:39). It is enough because this one idea explains why your chart sometimes draws headers and sometimes draws an axis, and why dates behave in two different ways depending on which you choose.
If you want a two-minute primer first, Data with Baraa's Tableau Data Types vs Dimension & Measure vs Discrete & Continuous (Data with Baraa, 2:00) lays out the three ideas that beginners blur together, so the longer video lands faster.
On your own data today: build the same chart twice, once with your date field as a blue pill and once as a green pill. Note which one gives you separate bars per month and which gives you a continuous line over time. Pick the one your stakeholders would read faster and keep it as the first sheet of your dashboard.
Day 4 (Thursday): write calculated fields that do not break
Watch Devyx's Tableau calculated fields: understanding row-level and aggregate calculations (Devyx, 14:57). It is enough because the title names the exact distinction behind the most common calculated field error: mixing a calculation that runs on each row with one that runs on the total.
This is a small channel, and it earns its place here because the video does one job and does it well. Fifteen minutes on this today saves hours of staring at error messages next week.
On your own data today: write three calculated fields. One row-level field, such as a margin per order or a flag for orders over a size you care about. One aggregate field, such as total revenue divided by total orders for an average order value. And one that compares your actual to the target from your lookup table. Check each against your answer key.
Day 5 (Friday): add percent of total and running totals
Watch Tableau's How to Calculate a Percent of Total for Measure Values in Tableau (Tableau, 2:06), then Jellyman Education's Tableau - Quick Table Calculations Running Total, Difference and Moving Average (Jellyman Education, 9:03). Together they are about eleven minutes, and enough because quick table calculations are a right-click away once you know they exist.
Table calculations work on the numbers already in your view, not on the underlying rows. That is why they are fast to add, and also why they change when you move a field around. Both videos show the menu; your job is to notice which direction each calculation runs.
On your own data today: add a percent of total to your region or product breakdown, a running total to your monthly trend, and a difference from the previous month. Then compare the numbers against your answer key one final time. End the week with a workbook whose totals you trust.
Week 2: build the dashboard and hand it over
Week 2 turns the foundations into the dashboard your stakeholders will open: LOD expressions for the numbers that ignore the view, a parameter for what-if choices, KPI tiles and a clean layout, filter actions that link the charts, then publishing. Each day ends with a part of the dashboard they will actually see.
Day 6 (Monday): use LOD expressions for numbers that ignore the view
Watch sqlbelle's Tableau LOD (Level of Detail) Expressions Explained - FIXED, INCLUDE, EXCLUDE - Complete Tutorial (sqlbelle, 32:22). It is enough because the title covers all three types in one sitting, and the half hour pays off: LOD expressions let a number stay put while the rest of the view changes.
The typical work case is a comparison. You want each region's revenue next to the company total, or each customer's first order date regardless of which month is being shown. Without an LOD, Tableau recalculates everything for whatever is in the view, which is exactly what you do not want here.
On your own data today: write one FIXED expression, such as total revenue per customer or the company-wide total. Use it to show each region's share of the company, or to group customers by how much they have spent in total. Check one number by hand in Excel.
Day 7 (Tuesday): let stakeholders choose with a parameter
Watch OneNumber's Tableau Parameters for Beginners (OneNumber, 19:17). It is enough because the title says beginners, and a beginner's parameter is all a first dashboard needs: one control that lets a viewer pick something, such as which measure to show or what target to compare against.
Parameters are the difference between a dashboard people read and one they use. A single dropdown that switches between revenue, units and margin often replaces three separate charts, and makes the review conversation about the business rather than about which chart is which.
On your own data today: build one parameter that lets the viewer choose between two or three measures, and a calculated field that returns whichever measure is selected. Put that field in your main chart and show the parameter control next to it. Try every option and confirm the chart and its title make sense each time.
Day 8 (Wednesday): build the KPI row and lay out the dashboard
Watch Andy Kriebel's Create 10 ACTIONABLE & IMPACTFUL KPIs in Tableau (Andy Kriebel, 15:28), then Data with Baraa's Mastering Tableau Dashboards: Layout Containers Explained (Data with Baraa, 26:34). Together they cover the two halves of today: the numbers people read first, and the boxes that hold everything in place.
Pick three or four KPIs, not ten. The video shows many ways to build them; your review needs only the few numbers your opening question asks about. Layout containers are the part most people skip and then regret, because without them the dashboard falls apart the moment someone opens it on a different screen.
On your own data today: build three or four KPI tiles that answer your opening question, such as total revenue, percent of target and change from last month. Create the dashboard, put a horizontal container across the top for the KPIs and a vertical one below for your charts, and set a fixed dashboard size that fits a laptop screen.
Day 9 (Thursday): link the charts with filter actions
Watch OneNumber's Tableau Dashboard Filter Actions Explained (OneNumber, 23:39). It is enough because filter actions are what turn four separate charts into one dashboard: click a region on the map or bar chart, and every other chart filters to that region.
This is the moment a dashboard starts to feel finished. It is also the moment stakeholders find out whether they trust it, because clicking around exposes every chart whose totals do not agree with the others.
On your own data today: add one filter action so clicking a bar in your region or product chart filters the trend chart and the KPI tiles. Test what happens when you click again to clear it. Then show the dashboard to one colleague for five minutes, without explaining it first, and write down the first thing they ask.
Day 10 (Friday): publish it and share the link
Watch OneNumber's How to Publish a Tableau Dashboard (OneNumber, 10:12). It is enough because today is a process, not a concept: tidy the workbook, decide what the published version is allowed to show, publish, and share a link your stakeholders can open.
Where you publish depends entirely on what your company already uses, and that is a question for IT, not for a tutorial. The section below covers what to ask. Whatever the answer, never publish company data to Tableau Public.
On your own data today: hide the sheets you used for testing, give the dashboard a clear title that states your opening question, and add a one-line note on where the data comes from and when it was last updated. Publish it to the place IT named, open the link in a private browser window to check what a stakeholder will see, then send it ahead of the review.
What to skip in these two weeks
Skip anything that does not get your one dashboard working before the review: the 21-hour full course, Tableau Public for work data, advanced chart types, certification prep and heavy formatting polish. Each is real and useful in some other month. None of them makes your numbers match the answer key, which is the only test that matters right now.
The 21-hour course. Data with Baraa's Tableau Ultimate Full Course (21 Hours) for Beginners - From Zero to HERO (Data with Baraa, 20:47:05) is thorough, and over twenty hours long. Do not watch it end to end in two weeks. Treat it as reference: when one screen confuses you, find that chapter, watch it, close it.
Tableau Public for anything from work. Per Tableau's Tableau Public page, checked October 2026, it is a free platform to create and publicly share visualizations online, and published visualizations are available for anyone to see. That is great for a personal portfolio on public data. It is the wrong place for your company's numbers, full stop.
Advanced chart types. Sankey diagrams, radial charts and custom maps look impressive in other people's portfolios. Your review needs bars, lines and a few KPI numbers that answer one question. If you want a quick burst of ideas instead, Andy Kriebel's Make an AWESOME Tableau Dashboard in Only 10 Minutes (Andy Kriebel, 11:31) is optional and short.
Certification prep. A Tableau certification is a broad test of the whole tool. Studying for it now trades your deadline for a credential. Come back to it once the review is behind you, if your role asks for one.
SQL, for now. If your data lives in a database, someone will eventually suggest connecting Tableau to it directly instead of exporting files. That is a good month-two project, once SQL is on your list. It is not a two-week one.
Where to publish and what to ask IT about licences
Building is the easy part: per Tableau's download page, checked October 2026, Tableau Desktop Free Edition needs no trial, expiration or credit card. Sharing with colleagues is where licences come in. Ask IT where your company publishes dashboards and which licence you have before Day 10, not on it, because the answer can take days.
For context, per Tableau's Tableau Cloud pricing page, checked October 2026, Creator is listed at $75 per user per month, Explorer at $42 and Viewer at $15, all billed annually. The same page says all Tableau products require an annual contract. That is why this is a question for IT, not something to put on a personal card.
Five questions cover most of it. Does the company already have Tableau licences, and which kind should I get? Where do our dashboards live, so I publish to the right place? Which data sources am I allowed to connect to, and is a file export fine for now? Who else needs access to see the dashboard, and do they already have it? Is there a rule about what data can leave the team?
If IT tells you the company does not use Tableau for sharing at all, that is useful too. Ask what they do use. Some teams that build in one tool share in another, and our two-week Power BI plan for finance teams follows the same ten-day shape if you end up switching.
Proof at the end: the dashboard you ship, plus the certificate
The proof that these ten days worked is not a quiz score. It is a published Tableau dashboard that answers your opening question, matches the numbers your stakeholders already trust, filters when they click, and lives somewhere they can open without asking you for a file.
Put three things in front of your stakeholders at the review. The dashboard link, with a note saying the totals match the old report. The KPI row, with each number tied to the question you were asked. And one sentence on how it updates: next month's export lands in the folder, the dashboard refreshes, nobody rebuilds anything.
LearnPath builds a path of YouTube videos for a topic and generates a quiz from each video's transcript, so you find out whether the watching stuck. Free opens step 1 of up to two paths; every later step is Pro, and the completion certificate is Pro-only, so free accounts do not get one. The certificate shows you finished a plan. The dashboard shows you can do the job.
If this review goes well and you want to turn it into a career move, our guide on how to become a data analyst with no experience shows where a Tableau dashboard fits in a portfolio.
Build a work-ready path on Tableau before your next review.
Frequently Asked Questions
Six questions that come up when an analyst is told "we use Tableau here" and handed a review date. The short answers below point back to the day in the plan that covers each one in practice, so you can jump straight there when the question comes up in the middle of a build.
Can you learn Tableau in two weeks for work?
Enough to ship one working dashboard, yes. Ten working days at about an hour each covers connecting your data, the blue and green pill logic, calculated fields, table calculations, LOD expressions, parameters, KPIs, layout, filter actions and publishing. It will not make you an expert, but it gets a dashboard you understand in front of your stakeholders.
What is the difference between blue and green pills in Tableau?
Blue pills are discrete: Tableau treats them as separate labels and draws headers, one per value. Green pills are continuous: Tableau treats them as a range and draws an axis. The colour is not the same as dimension vs measure, which is why a date can turn up either blue or green.
Should I use relationships or joins in Tableau?
Start with relationships. They keep each table at its own level of detail and let Tableau decide how to combine them for each chart, which avoids the double-counting a careless join causes. Use a join when you truly need one flat table. Day 2 of the plan covers both, plus blends.
Can I publish a work dashboard to Tableau Public?
No, not with company data. Per Tableau's Tableau Public page, checked October 2026, published visualizations are available for anyone to see. That is fine for a portfolio built on public data and wrong for anything from work. Ask IT where your company publishes dashboards before you press publish on Day 10.
Should I watch a full Tableau course before my review?
No. A 21-hour course is a reference, not a plan, when the review is two weeks away. Watch one short video a day aimed at that day's job, then spend the rest of the hour building on your own data. Open the long course only when one specific screen has you stuck.
Does LearnPath give a certificate for finishing a Tableau learning path?
Yes, on the Pro plan; free accounts do not get one. LearnPath builds a path of YouTube videos for a topic and generates a quiz from each video's transcript. Free opens step 1 of up to two paths, and every later step is Pro. The real proof is still the dashboard you ship.
