Hire senior and proven Tableau Developers

Among the leaders in this transformative field is Tableau, a software acquired by Salesforce in 2019 for $15.7 billion, which was also recognized for the 13th year as a Leader in the Gartner® Magic Quadrant™.

About Tableau

At its core, Tableau is a powerful and intuitive data visualization tool that allows users to connect to various data sources, such as Excel sheets, databases, and even websites; explore and analyze data; and create interactive and shareable dashboards.

In a period when tools to create interactive and accessible visualizations were barely mentioned, Tableau was created in 2003 by three partners who met at Stanford University. Together, they conceived Tableau's mission: to make data more accessible and understandable to everyone, regardless of their technical background.

And it was with this philosophy that Tableau became one of the main dashboard creation tools in the world. It is no wonder that today it is widely used by technical and non-technical people from all types of industries, used by over 63,000 organizations worldwide and holds roughly 17% of the Business Intelligence tool market.

And this recognition is deserved, since Tableau is widely democratic, fast, highly customizable, and very powerful in complex analyses and calculations. For those who have experience with Tableau's competitors, you may be shocked at how quickly a chart materializes in Tableau. Tableau uses VizQL technology, which translates drag-and-drop actions into data queries, enabling users to explore data visually and in real-time without writing a single line of code.

Oh, and anyone who thinks Tableau "only" helps you build charts and dashboards is mistaken. Over the course of its 22 years of existence, other tools focused on the data universe have been built to facilitate the entire chain that begins with raw data, through its collection, processing, aggregation, and translation into insights until it becomes an important asset for decision-making.

Tableau's product suite includes:

• Tableau Desktop: The primary authoring and publishing tool where developers create interactive dashboards, reports, and stories.

• Tableau Server and Tableau Cloud (formerly Tableau Online): These platforms allow for the sharing and collaboration of dashboards across an organization. The key difference is that Server is hosted on-premises, while Cloud is a fully hosted SaaS solution.

• Tableau Prep Builder: A tool designed for data preparation, allowing users to clean, shape, and combine their data before analysis.

• Tableau Public: A free version of Tableau Desktop where users can create and share visualizations publicly, fostering a vibrant community of data enthusiasts.

Industries and applications

As we have seen, Tableau's mission is truly put into practice, and it is no wonder that it is recognized as a market leader. This means that the variety of industries and application possibilities is endless, helping any type of department in a company to make the best data-driven decisions. That's why Tableau is able to deliver the most varied examples of industries and cases (see Tableau for Industries). Here are some examples:

• Retail and eCommerce: Retailers leverage Tableau to analyze customer behavior, track sales performance by region and product, optimize inventory management, and personalize marketing campaigns. Imagine a marketing manager being able to visualize the customer journey from ad click to purchase, identifying drop-off points and optimizing the conversion funnel in real-time.

Want to learn how to create a Sankey chart for customer behavior flow? See this article

• Finance and banking: In the financial sector, Tableau is used for risk management, fraud detection, customer profitability analysis, and tracking key performance indicators (KPIs) like loan origination and investment performance. A financial analyst can create a dashboard to monitor market trends and portfolio performance, drilling down into individual securities with a single click.

Sometimes Tableau (or any similar tool) won't have a specific type of viz built in. Fortunately, Tableau is highly customizable, allowing us to create new visualizations without having to download anything. See here how to create Percentage Gauges in Tableau.

• Healthcare: Healthcare providers use Tableau to analyze patient data to improve outcomes, manage hospital resources effectively, track disease outbreaks, and streamline operations. A hospital administrator can monitor patient admission rates, bed occupancy, and average length of stay to optimize staffing and resource allocation.

Tableau supports real-time streaming data, which is indispensable for industries like Healthcare. See this Hospital ER example dashboard.

• Manufacturing: Manufacturers rely on Tableau for supply chain optimization, production monitoring, quality control, and predictive maintenance. A plant manager can visualize production line efficiency, identify bottlenecks, and predict when machinery might require maintenance, minimizing downtime and saving costs.

Tableau is part of the day-to-day operations of the world's most important companies. See how it helps Tesla with its turbo-charges.

• Marketing and sales: Marketing and sales teams use Tableau to track campaign performance, analyze customer segmentation, measure return on investment (ROI), and forecast sales trends. A sales director can have a comprehensive view of the sales pipeline, individual team member performance, and progress towards targets.

For those who have read Storytelling with Data (if you don't know what I'm talking about, this is an important book about creating data visualizations) you know how important it is to choose the best type of visualization for each type of information. Here are two tips for creating better pie charts or how to create better stacked bar charts.

Skills for Tableau Developers

While Tableau is designed to be user-friendly, creating truly impactful and efficient dashboards requires a specific set of skills. These can be categorized into "must-haves" and "nice-to-haves."

Strong understanding of Data Visualization best practices

Probably the most important skill on the list, this is a non-negotiable skill. BI tools like Tableau have, on average, more than 20 types of data visualization methods built in, not counting the extensions created by the community that can be downloaded.

So it is clear that, faced with a world full of possibilities for displaying data, the developer responsible for building the dashboard must know when and where to use each type of visualization and method. I strongly recommend reading the book Storytelling with Data, as it, as stated by the authors, "teaches you the fundamentals of data visualization and how to communicate effectively with data".

Understanding of Tableau’s order of operations

This crucial concept defines the sequence that Tableau follows when applying filters, calculations, and aggregations. Misunderstanding this order often leads to incorrect results, especially when using context filters, Top N filters, or LOD expressions. Using the proper sequence is a must for accurate and reliable dashboards.

Data connection and preparation

It is important to understand that data can come from anywhere, not just from that usual Excel spreadsheet. Furthermore, the best dashboards are probably those that connect the most varied data sources in a single cohesive view, bringing together various data sources that are then transformed into important insights. Therefore, it is very important to understand this wide variety of data sources, how to connect to them, and get the best out of them. Sources include SQL databases, text files, JSON files, websites, APIs, and even PDFs. See the list of Tableau's supported connectors with more than 50 data sources.

Understanding of Data Warehousing and ETL concepts

In addition to understanding where the data comes from, it is very important to know how it is stored, what types and structures are supported, and how to create efficient relationships between them.

• Why are there multiple tables in the database and not one big table with everything?
• What is the granularity of a table? What is the cardinality of a relationship between tables?
• What consequences are caused by the relationship between different data sources, such as Cartesian products?
• What is the difference between a fact table and a dimension table?

This knowledge is crucial to ensure that the accuracy of the data displayed on a dashboard is reliable.

Problem-solving and analytical skills

Before you build any dashboard in Tableau, you need to consider the motivation behind that motivation. In most cases, if not all, it is to solve a business problem.

Even if you didn't have Tableau available, the problem would still be the same, and you would still strive, even if it took more time, to find the truth in the data and extract solutions.

So, this skill is independent of Tableau but goes beyond that for any developer who seeks to analyze data: understand business requirements and translate them into analytical questions; raise hypotheses and find patterns; use data to find information that, combined with your business knowledge, will become valuable insights.

Excellent communication and storytelling skills

Tools like Tableau have come to democratize access to data and the intelligence behind it. It is up to us, those responsible for creating dashboards, to promote this democratization by creating dashboards that are easy to interpret, tell a transparent and objective story, and are accessible to any stakeholder in any position in the company.

As they say, a good dashboard does not raise the question "How do I download this data into my Excel?"

Nice-to-have skills

• SQL proficiency: SQL is the main language for manipulating data in databases, and having this skill in your repertoire will greatly facilitate the interaction with tables, especially if you are pulling data from tables with a large number of rows. In these cases, using SQL is very useful for limiting or aggregating the table, rather than bringing it to the Tableau extract.

• Experience with other BI tools (e.g., Power BI, Looker): The Gartner® Magic Quadrant 2024 BI ranks at least 7 BI tools in addition to Tableau as leaders, meaning that Tableau is indeed good, but it is not alone. In a world of diverse economies, it is no surprise that companies choose other tools as their primary internal BI tool, so it is always welcome when the developer already knows how to use another tool. This reduces the learning curve and allows for good knowledge transfer between one tool and another.

• UI/UX design principles: You can't judge a book by its cover, but you can judge a dashboard by its first page. It's very important to remember that building a dashboard also involves building an interface, which isn't far from design work. I'm not saying that you need to have a degree in design to create dashboards in Tableau, but understand that this is a critical point and, especially for directors and high-level positions, it may be the most essential item. So, try to learn how to create a beautiful and practical dashboard; understand what color theory is and how to apply it; reuse your own company's colors, if possible; learn about user experience principles and use this to your advantage.

• Basic statistical knowledge: When we try to understand data behaviors, we get closer and closer to the basic concepts of statistics. These will help us gain more clarity about the data we are dealing with, so we can understand concepts such as correlation, distribution, and statistical significance. Such knowledge can lead to deeper and more accurate insights.

• Scripting skills (Python or R): Someone who already has experience with programming language logic will always be one step ahead, for two main reasons: the first is that Tableau has its own language for creating measures, VizQL, so if you already have some knowledge of programming logic, learning how to create new measures will not be a challenge.

The second reason is that, when working with more complex data sources, we will eventually encounter problems that require more complicated data manipulation, such as using loops, or when we need to perform complex web scraping. In such cases, languages ​​like Python can help us solve our problem in just a few lines.

Interview questions and example answers

If you are looking to hire for a Tableau developer position, here are some common interview questions that you should have the answers to at your disposal.

1. What is the difference between a dimension and a measure in Tableau?

In Tableau, dimensions and measures are the two fundamental types of data fields. Dimensions are qualitative, categorical data that you use to slice and dice your data. You can think of them as typically discrete fields, like the 'who, what, where' of your data, i.e., customer name, product category, or region. On the other hand, measures are quantitative, numerical data that you can use to perform mathematical operations. They represent the 'how much' or 'how many,' such as sales, profit, or quantity, usually classified as continuous fields.

2. Explain the difference between a live connection and an extract in Tableau.

The choice between a live connection and an extract depends on the use case and performance requirements. A live connection queries the underlying data source directly, meaning you always have real-time data, which is crucial for applications that require up-to-the-minute information.

However, the performance of the dashboard is dependent on the speed and workload of the data source.

An extract, on the other hand, is a snapshot of the data that is imported into Tableau's high-performance in-memory data engine, and it generally results in much faster dashboard performance as the queries are not hitting the original database. Extracts are ideal for large datasets or when the underlying data source is slow, as you can also schedule regular refreshes of the extract to keep the data up-to-date.

3. What are Level of Detail (LOD) expressions, and why are they useful?

Level of Detail, or LOD, expressions are a powerful feature in Tableau that allow you to compute aggregations at a different level of detail than the one in your view. Three types of LOD expressions exist: FIXED, INCLUDE, and EXCLUDE.

For instance, you could use a FIXED LOD to calculate the total sales for each customer, regardless of the other dimensions in the view. This is incredibly useful for creating complex calculations like customer cohort or market basket analyses without manipulating the underlying data structure. They provide a level of control over calculations that was previously difficult to achieve. LOD measures can be compared to window functions in SQL.

4. What is a context filter?

Unlike regular filters, you can think of a context filter as an independent filter because it's the first filter happening when we create a sheet in Tableau (see Tableau's Order of Operations). When you create a context filter, Tableau creates a temporary table for the context. All other filters, including Top N filters or FIXED LOD calculations, will then process based on the data that has passed through the context filter. So, if you set 'Region' as a context filter and then apply a 'Top 10 Products' filter, Tableau will first filter the data down to only that region and then find the top 10 products within that context.

5. Can you explain the use of a parameter in Tableau?

A parameter is a dynamic placeholder value that a user can change, and it can be integrated into calculations, filters, and reference lines. Unlike filters that only narrow down data, parameters allow for more complex 'what-if' analysis and user-driven inputs.

For that, it is possible to create input elements allowing the user to write their own values and have the dashboard affected by it. Thus, it is a good way to keep values needed for the analysis that do not live in the regular data sources, like a suspended list with personalized values.

6. What is the difference between discrete and continuous fields in Tableau?

Discrete fields (blue pills in Tableau) often represent categories, which are one of a limited, and usually fixed, number of possible values. Think of it as something you would not run any math operation on. Regions, Product Category, or Gender are examples of discrete fields. Continuous fields (green pills) represent a continuous range of values.

They are usually numeric and are directly affected by the business operation or by its environment. Think of them as an unbroken whole, like 'Sales' or 'Temperature'. These are the values you usually put on the X-axis of a line chart.

7. What is a "story" in Tableau?

In Tableau, we can create single visualizations, known as Sheets, we can also create dashboards, called Workbooks, and there's a third way to represent data that is called Story. A story is a sequence of visualizations that work together to convey information. A story is usually a collection of sheets, arranged in a sequence to explain something in a logical sequence. You can create stories to tell a data narrative, provide context, demonstrate how decisions relate to outcomes, or simply make a compelling case.

8. What are "actions" in Tableau?

Actions are powerful resources in Tableau that can make any dashboard more interactive and automated. They allow a user's action on one part of the dashboard to trigger a change in another part. With actions, you can differentiate your dashboard from static reports to make them look more like dynamic websites. The three main types are Filter, Highlight, and URL actions. For example, I can set up a Filter action so that when a user clicks on a state in a map view, other charts on the dashboard are automatically filtered to show data for only that selected state.

9. What is Tableau's Performance Recorder?

Tableau's Performance Recorder is a not-so-well-known but very powerful tool for analyzing and treating dashboard performance. It ​​helps the user get a detailed breakdown of events and identify the exact pain points, whether they're in query execution, rendering, or geocoding. Based on its analysis, the user can see which operations are taking more time and work on reducing them.

10. How do relationships differ from traditional joins on the data source page?

Joins combine tables by creating a new, fixed data table before analysis begins. This can sometimes cause data duplication or loss of information if the tables are at different levels of granularity.

Relationships, on the other hand, do not create a single new table. Instead, they maintain the separation of the tables and create context-aware connections, fetching data from each table only when it's needed for a specific visualization.

The primary advantage of relationships is their flexibility and intelligence; they automatically manage the level of detail, preventing common data duplication issues and making working with complex, multi-table data sources easier and more intuitive.