Tableau BlogModify

What are Tableau Cloud Custom Domains?

Custom Domains (per site) is a new feature in Tableau Cloud that allows site administrators to configure a subdomain, like analytics.example.com, enabling them to easily embed and access published content. With a custom domain, they can efficiently route stakeholders to Tableau Cloud content, offering an organization-specific URL subdomain that maintains their brand identity instead of having to remember and solely depend on the complicated and non user-friendly pod url's of today. 

Key Benefits for Tableau Customers

Seamlessly Embed Your Analytics
Custom Domains particularly resolves instances where third-party cookies are blocked by default in some browsers for embedding customers. Tableau Cloud Custom Domain feature resolves these 3rd party cookie issues by allowing site administrators to configure a subdomain of their own, and thus the embedded content comes from the customer owner domain (aka 1st party). 

Strengthen Your Brand Identity and Security
Custom domain reinforces this by presenting your Tableau content under your company's own URL. This provides a more professional and trustworthy experience for your employees, partners, and customers. vs. embedding. Even if browsers are not blocking by default, this enables them to embed and access published content without exposing Tableau site details resolving Safari, Chrome and other browsers and addressing their 3rd party cookies' privacy and security concerns.

Simplify Access for Everyone
With a custom domain, you can provide a simple, memorable web address for your stakeholders to access their dashboards and reports. This makes it easier to share links and direct traffic to your Tableau Cloud site.

How to get started with Custom Domains

For this Tableau Cloud feature Site Admins can create and configure Custom Domains in site settings. There is no additional cost to Custom Domains feature. Customers/Site Admins can utilize and set up Custom Domains on any of their sites by going to the individual sites settings for each site.

Note: To add a custom domain to Tableau Cloud, you must meet the following user and TLS certificate requirements. Full info and steps for setup are in the online docs, the summary below is just to get you started.

Requirements: 

• Administrator access to your Tableau Cloud site.
• Administrator privileges to the DNS configuration of the parent domain.
• While not required, it is a good idea to acquire a valid TLS certificate chain file and private key from a trusted authority (for example, Verisign, Thawte, Comodo, GoDaddy) ahead of starting the setup. It is also a good idea to give a heads up to the IT/DNS team as they will need to update the records once the custom domain dns info is provided during setup. 

To setup Custom domain (per site), sign in to the Tableau Cloud site as an administrator, go to settings and follow the steps under the Custom Domain section.

If you’re ready to experience Tableau Cloud yourself, start a free trial.

The migration journey from Tableau Server to Tableau Cloud can be a daunting experience. Without a reliable process, the migration can be a time-consuming and error-prone endeavor, potentially leading to a multitude of issues. To help solve this challenge, Tableau offers a suite of tools designed to facilitate content movement, both between Server environments and Server to Cloud in the form of the Migration SDK, the Cloud Migration App, and the Content Migration Tool. 

While they share the common goal of simplifying content transfer, they differ significantly in their design, target audience, capabilities, and intended use cases. Understanding these distinctions is crucial for selecting the right tool for your content migration and management.

Overview of Tableau Tooling 

Here's a breakdown to help clarify which tool might be right for you when considering a move from Tableau Server to Tableau Cloud:

• Tableau Migration SDK: This is a Software Development Kit (SDK), a toolbox for developers to build their own custom migration applications. It is the recommended tool for the technical movement of users and content during a migration from Tableau Server to Tableau Cloud. The Migration SDK is an API-driven engine with a networking layer and uses hooks like Filters, Mappings, and Transformers. It is not a turnkey solution or a pre-built application and requires customization and development effort. It's designed for junior level developers and above who are familiar with Tableau and have experience in Python or .NET. It's intended for a one-time migration event from Server to Cloud, not recurring content promotion.
• Cloud Migration App: This is an open-source desktop application built on the Tableau Migration SDK. It is designed to allow administrators to copy content and users from Tableau Server to Tableau Cloud with ease, specifically simplifying the process for administrators with smaller deployments of around 100 workbooks. A major limitation of this app is that it only has basic features, like user mapping, but lacks functionality such as filtering, renaming, and authentication types.
• Tableau Content Migration Tool (CMT): This Windows-only desktop app with console runner provides a user interface and is designed for copying or migrating content between Tableau Server sites or between Tableau Cloud sites. It is explicitly not recommended for migrations when moving from Tableau Server to Tableau Cloud.
   

Comparison 

Choosing the Right Tool for the Job

The choice of migration tool depends heavily on the specific scenario, the size and complexity of the Tableau environment, the technical expertise available, and the specific content types that need to be migrated. In general, the scenarios below should help guide the right choice when deciding on what to use.

Server to Cloud Migration:

• Customers with small deployments (around 100 workbooks), limited development resources, and core content types (users, groups, projects, data sources, workbooks, permissions, custom views) may find the Cloud Migration App to be the most suitable, straightforward option.
• Customers with larger or more complex Server deployments, significant customization needs, specific content types not supported by the App (e.g. Prep Flows or Virtual Connections in the future), or who require integrating migration into existing workflows should consider using the Migration SDK and building a custom application. This is also the intended path for Professional Services and Migration Partners serving large customers.
• Using the Tableau Content Migration Tool for Server to Cloud migration is not recommended.

Server to Server or Cloud to Cloud Migration:

• The Tableau Content Migration Tool (CMT) is the primary and recommended tool for migrations and content management tasks between Tableau Server sites and between Tableau Cloud sites. Its broad capabilities for content promotion, environment migration, site consolidation, and maintenance make it well-suited for these scenarios.
• The Migration SDK is also able to support Cloud to Cloud migrations with code customization and limitations with embedded credentials and subscriptions, with plans to expand on this functionality in the future.
• The Cloud Migration App is not able to support either Server to Server or Cloud to Cloud migrations.

Ultimately, the three tools provided by Tableau to help with content management and migration are distinct offerings with different use cases and supported scenarios. In general, the rule of thumb is that all Server to Cloud migrations should be done with the Migration SDK or the Migration App, while content migration across Server to Server or Cloud to Cloud should be done with the Content Migration Tool. At the end of the day, organizations should carefully evaluate their specific requirements, technical resources, and migration path to determine the most appropriate tool for their specific needs.

Learn more about best practices for migrating to Tableau Cloud from Tableau Server with the Trailhead Badge for Cloud Migration.

In today's digital landscape, businesses increasingly rely on cloud solutions, often involving sensitive information, including payment card data. This achievement means that Tableau Cloud not only empowers organizations in the financial services industry to leverage their data effectively but also enhances our existing robust security and compliance built into Tableau Cloud.

Tableau Cloud has achieved Payment Card Industry Data Security Standard (PCI-DSS) 4.0 compliance, reinforcing our unwavering commitment to security and ensuring your sensitive payment card data is always protected.

What is PCI-DSS and How Does it Impact Tableau Cloud?

The Payment Card Industry Data Security Standard (PCI-DSS) is a globally recognized set of security standards. Its purpose is to ensure that all companies that accept, process, store, or transmit credit card information maintain a secure environment. For businesses handling customer payments or financial data, PCI-DSS compliance is crucial for safeguarding cardholder data and maintaining customer trust.

Tableau Cloud plays a vital role as a "service provider" in the PCI-DSS ecosystem, supporting payment processing for other companies. Our customers can now operate with even greater assurance knowing that their data environment within Tableau Cloud meets these rigorous standards.

Tableau Cloud's Commitment to Top-Tier Security

As part of our PCI-DSS compliance, we're proud to share that Tableau Cloud is a Level 1 service provider under PCI-DSS. This is the highest level of compliance, signifying that we meet all 12 technical and organizational controls of PCI-DSS.

To meet this standard, Tableau performs rigorous testing and validation, including:

• An annual on-site assessment and Report on Compliance (ROC) by a Qualified Security Assessor (QSA).
• Quarterly network scans by an Approved Scanning Vendor (ASV) to identify and address any vulnerabilities.
• An Attestation of Compliance (AOC) signed by a senior executive, confirming our adherence to the standards.

Furthermore, Tableau Cloud operates on AWS's Infrastructure as a Service (IaaS) platform and Salesforce’s Hyperforce, both of which are themselves PCI-DSS Level 1 compliant. This means our foundational infrastructure is built on a bedrock of security, with AWS responsible for securing the data centers and physical controls.

The Shared Responsibility Model: A Partnership in Security

Tableau’s PCI-DSS compliance is built on the foundational principle of a Shared Responsibility Model. As the cloud service provider, Tableau Cloud manages the underlying infrastructure and platform, securing areas like network firewalls, data encryption at rest and in transit, vulnerability management, patching, and system activity logging. Conversely, our customers are responsible for securing their own application settings, data, and usage of the service through configurations of secure access, encrypting data during transmission to and from Tableau Cloud, monitoring user activity, and securing any custom integrations.

It is crucial that customers uphold their responsibility under this model for the environment to be fully PCI-DSS compliant. 

To that end, Tableau provides a bevy of tools that can help customers uphold their compliance requirements under the shared responsibility model:

• Customer-Managed Encryption Key (CMEK): This critical feature allows customers to encrypt site data extracts with their own customer-managed, site-specific key.
• Tableau Bridge: Securely connect Tableau Cloud to private network data, including on-premises databases or private cloud data, using HTTPS and WebSockets for reliable, encrypted communication.
• Activity Log: Gain comprehensive visibility into detailed log events, sent directly to your own Amazon S3 bucket. This enables in-depth analysis, auditing, and monitoring of site activities and permission changes, crucial for compliance.
• Robust Access Controls: Tableau Cloud offers a variety of features to ensure only authorized users access your data, including:
  • Multi-Factor Authentication (MFA) for strong verification.
  • Single Sign-On (SSO) for streamlined and consistent authentication.
  • SCIM for automated user provisioning and de-provisioning.
  • User Role/Permission Control to enforce the principle of least privilege.
  • Data Access Control and Row-Level Security (RLS) to restrict data visibility based on user identity.

Key Considerations for Shared Responsibility

Furthermore, in addition to customer responsibilities, it is important to call out the limitations around Tableau’s current PCI-DSS implementation. These limitations must be taken into account by customers as part of their Tableau Cloud implementation in order to be covered by Tableau’s PCI-DSS compliance:

• Customers must use CMEK to encrypt their extracts.
• Only data in the form of data extracts or live connections are supported for PCI-DSS compliance. Using embedded data sources like Excel or CSV files, which cannot be encrypted with CMEK, is not within the scope of this compliance.
• Customers are required to use their own Identity Provider (IdP) for Single Sign-On (SSO).

Understanding these specific points helps ensure that customers can effectively leverage Tableau Cloud's PCI-DSS compliant environment while meeting their own compliance obligations.

Your Trusted Partner in Data Security

At Salesforce, trust is our number one value, a principle that guides everything we do. Tableau Cloud is engineered with security at its core, committed to protecting your sensitive cardholder data. Our comprehensive approach, spanning infrastructure security, robust encryption, and detailed logging, provides a resilient environment for your critical information.

We are dedicated to being your trusted partner on the journey toward compliance and data security. 

For more detailed information, please refer to the Tableau Cloud PCI-DSS Whitepaper, visit our Tableau compliance website and resources, or reach out to your Salesforce account executive.

Want to experience this level of security and compliance? Start a free trial of Tableau Cloud.

How often have you heard, “Without good data, you won’t have good AI?” It sounds simple, but studies show that “ensuring scalable, reliable data” is one of the top analytics challenges organizations face according to the Salesforce State of Data and Analytics Report. This challenge has significant downstream effects. Organizations that cannot confidently confirm they have reliable data cannot confidently adopt AI or self-service capabilities. To scale responsibly around AI and self-service-driven capabilities, organizations need a better process to explore, improve, and validate the data sources that drive them.

This blog provides the tools to monitor your Tableau data sources more effectively and a framework to drive iterative improvements, so you can be more confident in your data.

The framework consists of four steps:

1. Monitor your data
2. Identify a meaningful objective
3. Create and promote data sources
4. Enable your users

Improving data confidence starts with a baseline understanding of your data. Consider one of your top-level Tableau Projects, and ask yourself:

• How many data sources are in that project?
• How do users distinguish trusted, governed sources from experimental ones?
• Is there supporting documentation to help new users understand the data?

If you can’t answer these questions, scaling for AI or self-service discovery may make you nervous. The first step to alleviate these concerns is to monitor and explore your data sources. Any Tableau Cloud Admin can do so by connecting the Data Source Manager Accelerator to your Tableau Cloud Site, and then sharing for other users to see.

Start by reviewing the “Know your data sources” section to see how users can interact with their data.

A published data source makes a single source of data accessible to permissioned users, enabling self-service and Tableau Pulse capabilities. Certified data sources go a step further, assuring users that the data has passed organizational governance standards and can be trusted. Embedded data sources, however, are embedded within a Tableau workbook, making them inaccessible for exploration outside of the dashboard through web-authoring or Tableau Pulse. Monitoring and managing your data assets along these categories will help you assess your readiness to embrace Tableau’s AI and self-service capabilities in your Tableau Sites.

Instead of overwhelming yourself by attempting to improve all of your data at once, focus on achieving short-term wins on meaningful objectives. To identify these use cases, take any potential initiative, a digital whiteboard, and answer the following questions for the initiative:

The initiatives you choose to focus on should confirm all of the above. It should impact relevant organizational goals, use data that can feasibly be brought into Tableau, and demonstrate that analytics can indeed drive meaningful progress. Once confirmed, you have a worthy use case to focus on. Create a designated Tableau Project and follow the next step to equip users with explorable data assets.

Learn more about creating value maps to help select meaningful objections on our website.

You’ve listed the necessary data sources for your initiative in step 2 above, now make sure they are available in your Tableau Site by referring to the Data Source Manager Accelerator. 

Here, all of your data sources are categorized as Certified, Published, or Embedded, as discussed in Step 1. Filter the Accelerator for your Tableau Project, and explore the following sections of the dashboard to learn more about your data.

Embedded Data Sources (bottom right)

On the bottom right, you see all your embedded data sources, sorted by repetitions. Remember, these data sources are embedded within dashboards, meaning they cannot be explored with Tableau Pulse or web-authoring. Review this section to:

1. Consider publishing highly-used embedded data sources to enable features and consistency
2. Redirect embedded dashboard connections to their Published/Certified equivalent, if it exists 

Published Data (bottom left)

On the bottom left, you see all your Published data sources, sorted by user activity. While these data sources enable enhanced Tableau features, they lack any indication of whether the data has passed governance checks and can be trusted. Review this section to:

1. Consider retiring unused data sources
2. Consider certifying highly-used Published data sources
3. For certification candidates, collaborate with data owners and experts to complete the certification checklist

Learn more about how to publish data sources on our website.

Certification Checklist (hover over the checklist on the dashboard)

The certification checklist is your organization’s process for vetting data sources, which ensures users understand why they can trust their data. This checklist typically includes a variety of checks to confirm accuracy of the data and the presence of supporting documentation. While we can’t directly monitor data accuracy with this Accelerator, we can track other variables to help data owners organize their certification efforts. Hover over the checklist on the dashboard to view the following six checklist items for each data source.

• Licensed Owner: Confirms the data owner is still a licensed Tableau user;
• Refreshed Last 30 days: Confirms the data source has been refreshed in the last month;
• Description: Confirms the data source description includes specific background about the data and guides users where to go with questions (format customizable by you);
• Data Source Type: Confirms the data comes from preferred sources and databases;
• Title Format: Confirms the title is legible without special characters;
• Accessed Last 90 days Confirms users still use the data;

Note: This is a starting point. Customize your checklist items by editing the “Score_” calculations in the Accelerator or adding new ones.

Once a data expert vouches for the accuracy of the data, the data owner can use this Accelerator to ensure the necessary supporting documentation is included. These combined checks, done inside and outside of this dashboard, will ensure data sources are ready to receive the Certification label in Tableau.

Certified Data (top)

At the top, you’ll find all data sources given the Certification label on Tableau Cloud, sorted by user activity and an icon confirming compliance with the checklist.

These are your prized data assets. They have been reviewed by a data expert to ensure accuracy, supplemented with supporting resources to address user questions, and have passed organizational governance standards to ensure trust and consistency. Review this section to:

1. Confirm data adoption
2. Ensure compliance with the certification checklist
3. Consider retiring unused data
4. Identify missing data sources from step 2

Learn more about how to certify data sources on our website. 

With a defined symbol to distinguish trusted data and a credible process to back it up, it’s time to enable your users. Start by informing your users via discussion forums, user group meetings, and other channels that they now have access to new, high-quality data sources. Next, teach them best practices on how to validate and explore these data assets by guiding them to free Tableau Modules on Trailhead. Lastly, regularly check the Data Source Manager Accelerator to ensure data adoption levels meet expectations. 

By implementing our four-step framework for a meaningful objective, you have successfully converted the challenge of “ensuring scalable, reliable data” into an ongoing discussion. This dialogue unlocks new features, improves users’ speed-to-insight, and creates a catalog of quality data ready to inform AI and self-service at scale across your business. Now, identify the next meaningful initiative, and repeat the process. Eventually, you’ll flip the narrative and be able to say, “we have great data, so we know we have great AI.”

Download the Data Source Manager for Tableau Cloud Accelerator on the Tableau Exchange to help monitor best practices. 

As 2024 draws to a close, we want to express our heartfelt gratitude to the incredible Tableau Community. Your boundless creativity and innovative ideas have been a constant source of inspiration, driving us to develop products that enhance and enrich your experience. It’s your feedback and collaboration that help us push the boundaries of what’s possible in data visualization and analytics.

Thank you DataFam for inspiring, voting for, and sharing with us about the features released in 2024! This year we released features that address 150+ ideas with 46,000+ points in the Forums, across Tableau Desktop, Online, Tableau Pulse, and more. And we made progress on providing visibility into the ideas on the Tableau Community Forums: This year we have updated more than 2,900 ideas.

Here are some of the features inspired by ideas you submitted on the forums:

Viz Extensions (2024.2) & Table Viz Extension (2024.3)

See this idea on Sankey Charts and Thierry Jakercevic’s idea on tables

You asked, and we delivered—unlock your creativity and speed to insight with the latest updates to table formatting in Tableau. These enhancements, including resizing, sorting, and conditional formatting, are designed to help you visualize and present your data more efficiently and clearly. Together, these features not only streamline your workflow but also enhance the overall clarity and attention to important data points and anomalies.

For developers, Viz Extensions provide a unique opportunity to expand Tableau’s analytics capabilities with custom visualizations. These can be developed using any JavaScript visualization library and easily integrated into Tableau through our Extensions API. This makes it straightforward to deliver personalized visualization tools that can be scaled to every analyst in your organization and shared with the wider Tableau Community.

"Tables in Tableau have always been in the forefront of user feedback. It's a critical means for how our users deliver analytics, and stands to have more dedicated usability improvements. The community advocacy around these themes helped us build the business case with leadership that we should make a more dedicated approach as a Viz Extension, delivering to a variety of feedback on how users specifically wanted to design and format tables inside Tableau. It provided us the quantitation such that we could discuss usability in Tableau with decision makers, and help them rationalize that the investment to this experience is a data driven decision. The ideas forum continues to be how we think about the key issues we want to explore in our long term roadmap and specifically how we can develop the foundations that allow for us to be more responsive to the feedback as we look ahead."

Wilson Po, Product Management Director

We encourage developers to join our DataDev Developer community to share feedback and ideas for new extensions. Developers can learn more about developing new Viz Extensions by learning more about the Tableau Extension API.

Accessibility: Google Fonts in Tableau Cloud (2024.3)

Idea submitted by stelloprint dev

You spoke, and we listened! Support for Google fonts is a significant win for our front-end developers. With these web safe fonts, they can be used without having to install them on all client machines. This feature helps provide a consistent experience across all viewers' browsers, while maintaining the intended look and feel of your data visualizations and dashboards.

 

"Thanks to our community’s input and ideas regarding our product support for web safe fonts. From their requests and feedback, we were able to prioritize and focus on adding new Google fonts to Tableau Cloud. These ten new fonts add new variety and options for our Cloud customers while also being web safe fonts. This adds a huge amount of value for customers needing a consistent representation of their data across browsers."

Rob Thomas, Product Management Senior Manager

Multi-fact Relationships (2024.2)

Idea submitted by Cristhian De la Hoz

The Tableau Community was ecstatic with the release of Multi-fact Relationships! This data modeling capability allows you to combine multiple fact tables and gain insights from complex data models. With Multi-fact Relationships, you can easily connect different data sources and analyze them together, without having to blend them. This is especially useful when you have dimensions that share the same meaning across tables, or when analyzing data that goes through different stages. To use Multi-fact Relationships, simply identify your base tables and create relationships between them and shared dimension tables. With this powerful tool, you can unlock new types of analysis and accelerate your time to insights. 

Say goodbye to linking fields - easily blend data sources with shared tables and trees.

 

We define Data Literacy as the range of capabilities that describe someone’s ability to explore, understand, and communicate with data. These days, when we have questions, there’s plenty of data around for us to find the answers, and having some data literacy will make that possible. Unfortunately, it’s not always structured in a way that makes it easy to use, which is why we’ll spend time exploring the different characteristics of data, the ways it can be structured, and why. Lastly, we’ll touch on how to restructure the data to make it work better for your needs.

Know what to look for in data

There are lots of words that are used to describe the details of an item: traits, characteristics, specifications, descriptions. All of them help us find the information we need to determine if something is a good fit for what we’re doing. Recently, I needed to replace a broken part on my bike. For the bike to work right, the replacement I bought had to match the specifications of the original and the only way to be sure of that was to read the descriptions of the parts I was considering buying. It’s similar when looking at data: for us to get the answers we need, the data has to have the traits that make finding our answer possible, and the only way to be sure it does is to know what qualities it needs to be a good fit.

The first trait is High Volume. More (relevant) data gives us a much more reliable result. If one person tells you a restaurant is good, you might believe them. However, if 100 people tell you a restaurant is good, it’d be hard to believe otherwise. The same goes for our professional data - more records telling us something is good or bad, gives us higher confidence in our results.

The second trait is Historical. If I visited Seattle for the first time and it was sunny the first day of my visit, it would be naive of me to say that it’s always sunny in Seattle. I’d need data stretching back quite a while to feel confident in making a statement like this and it’s no different for your analysis. The more historical data you have, the more likely you are to accurately anticipate what’s to come in the future.

Third we have Detail. Recently, the water pressure in my kitchen sink dropped significantly, forcing me to troubleshoot the situation. First I checked the water supply, then the valves under the sink, then the line going to the faucet, and finally the faucet itself getting more detailed with each dead end. I finally discovered that it was debris in the faucet blocking the water flow. It’s no different than exploring our data to understand why something is happening in our business or lives. The more detail we have in our data, the more depth and accuracy we can get in our answers and insight which we can confidently use to make the best decisions

Lastly is Consistency and Standardization. Think about the last time you were in a conversation with a friend and you were using the same words but had different understandings of what they meant. You end up with miscommunications that potentially cause issues down the road. Similarly, a lack of consistency in your data: field names, date formats, number formats, and a lack of standardization in how data is recorded results in numbers that aren’t an accurate representation of what actually happened. Imagine that you work at a cell phone store and you return five iPhones. When typing in what they were you type: iPHONE5, I phone5, iphone 5, I Phone 5, and Iphone Five. You then try to report on how many iPhone 5’s you returned and instead of your system showing five iPhone 5’s it shows one return of five different iPhones because it’s impossible for your system to know they are the same thing. Standardizing on a naming convention, or date format, or currency prevents situations like this and provides the consistency needed to get an accurate picture of your data.

Know what well-structured data looks like

It’s bound to happen. At some point you’ll come up against data that just feels difficult to work with. You may not know why, but I’d bet that it comes down to the data not being well-structured. It’s a bit of a loaded statement as the meaning of “well-structured” can differ based on the use case, and the systems/applications you’re using, however there are some basic data structure principles that should improve your experience when working with data.

First, you need to figure out how it’ll be used. Are you going to use software to visualize the data? Or are you presenting the data in a spreadsheet? Or are you preparing the data for someone else? Each method will change the way the data needs to be structured. There are exceptions to every rule, but once you know the way it’ll be used, these tips are a good rule of thumb to get you started:

Columns and rows make up the core of our structure. Just like columns in buildings, columns in data are the vertical sections of our data structure stretching up and down the page. In the example below, “Date” would be a column of information along with 7 other columns. Rows on the other hand are the horizontal sections of our data structure spanning left to right across the page. In the example below there are 12 rows of data.

If you are using software like Tableau to visualize your data, each row should be an instance or an event, and each column should represent a detail about that instance or event. In the image below, the first row (or event) is a commute. The details (columns) are the date, the time of day, the temperature, the precipitation type etc…Each row after the first row is an additional event where we collect the same types of details that we did in the row (event) above: 

This allows the software to easily search through your data when trying to retrieve the answers to your questions. The downside to this format is that it’s hard for humans to consume which is why it’s not uncommon see data structured like this when people don’t have access to data visualization software: 

The format above is human readable but requires you to manually calculate totals and aggregations, and manipulate structure. This structure will also continue to grow wider over time as new data is added, making it much harder to visually analyze without lots of scrolling. Lastly, having individual columns for each date or having multiple types of values within a single column makes it almost impossible for an application to help you analyze your data.

Want to learn more about well-structured data? Check out this Trailhead module today!

What to do if your data is not well-structured

Rather than let these issues prevent us from getting answers, there are a few simple things we can do to solve a good amount of our data problems.

Go back to the source. Restructuring your data can help your immediate problem but doesn’t solve the larger issue of how it got that way in the first place. It also creates additional data sources that may not align to your corporate standards which can be a problem when comparing results across your business. It’s always the best decision to start with your data team to see what’s possible. Remember, for them to make the right changes, they’ll need plenty of context so come prepared to explain what you are hoping to get from the data.

Make the change. In a perfect world your data team would help you fix poorly structured data, but that isn’t always possible, and not everyone has a data team. When it’s not, there are a couple of common structure issues that you can address on your own.

Split ‘em Up! The first issue is when fields that should be separate appear as a single line of information. For example if the fields [Name] and [Customer ID] showed up together as:

“SamPriddy CN1357WA” and needed to be separate fields instead, you’d perform what’s called a “Split”. A split gives the user the ability to choose which “delimiter” gets used to decide where to break the field apart. A delimiter is just the name for the character in the line of information that the user chooses to use as the breaking point. So in the Case of the example above where the field says: “SamPriddy CN1357WA” we could use the “Space” character as our breaking point, or our delimiter, since we wanted to split the name apart from the customer ID. This functionality is built right into Tableau Desktop and Tableau Prep with no coding required. Want to learn how? Check out this help article for more information.

Pivot, Pivot. The second issue relates back to the commute data example above. In its human readable form the data is wide, having a column for each date, with a mix of different types of values filling in the column below. You’d use a “Pivot” to change the data from being wide to being long instead.

This is by no means the only way to do a pivot but an example of the concept and what it means. Here’s a simplified example to illustrate the process:

The data below is formatted for human consumption. We want to use this in Tableau though and this format won’t work. We’ll need to do a pivot to get it into the appropriate structure. 

First we’ll make space by shifting our data out of the way. We’ll then make a column header for our fields that would be more appropriate as a category. In the example below we’ll copy/paste the “Product Type” fields down column A for the amount of values that showed up across their row. Because “Appliances” had four values (1 for each quarter) we’ll copy/paste “Appliances” until we have four rows of this product type. Similarly, we’ll pivot the individual quarters down and to the left, with Column B being our pivot point, changing the data from being horizontal, to vertical. We’ll then rename the column “Quarter” as it now contains all quarters, not just one specific quarter. 

Next, we continue the process until we have all of our product types duplicated for the amount of records they originally showed (4 each in this example). This is what we mean when we say the data gets “longer”. We haven’t added any data, just changed the layout. We’ll also continue copy/pasting the quarters down our “Quarter” column until each product type has a specific quarter next to it. 

Once that’s done, all that’s left to do is to fill in our values. Remember that originally Q1-Q4 were horizontal so we’ll move our values into the empty column next to the quarter they belong with. We’ll repeat this process until all of our values are where they belong in their new column which we’ll name “Values.” Depending on your data, the name should fit whatever best fits the values. 

After the final column is populated we’re ready to connect to this data to answer some questions. 

This example is shown to illustrate the concept, however it’s more likely that you’ll do a pivot in a tool like Tableau Desktop, or a data preparation tool like Tableau Prep. In those cases, the tools have functionality that eliminates the manual work shown in the example above, but ultimately gives you the same results. Want to learn how to do pivots in Tableau? Check out this help article for in-depth instructions.

What next?

Be a sponge, and practice, practice, practice! The best thing you can do is continue to learn, and get hands on as much as possible. Check out our Data Skills homepage for access to tons of Data Literacy resources. Visit Tableau Public to download and use Tableau Desktop for free as well as free learning and sample data sets! If you prefer webinars, check out the Datafam Discovery Kit for more resources and updates on upcoming live webinars.