Increase your Data Culture by Embracing these Trends

At the Milwaukee Tableau User Group meeting last quarter (which was held soon after the Tableau Conference) we asked attendees which Tableau Conference event was their favorite.

Of course, Iron Viz and Devs on Stage were highly rated, as they should be. They used to be my favorite as well. But, over the past few years, I’ve come to really appreciate the opening keynote with Chief Executive Officer, Mark Nelson, and Chief Product Officer, Francois Ajenstat. It’s full of big picture, industry trends. While it doesn’t give specifics, such as what features will be released soon, it does provides a vision of the impact data has on everyday activities and the trajectory of the company.

Tableau Tim provided a recap of this keynote on his YouTube channel and you can find the full keynote here. But I'm here to discuss what I’m most excited about, why, and how it’s impacting my 2022 plans. Let's dive in with the overall theme: data for everyone, anywhere.

Integration and Embedded Analytics


  • Data is a member of your team
  • You should be able to ask it questions
  • And it should let you know when there are issues
  • Get notified about the things you care about where you work
  • Easily search for the content you want where you already are
  • Integrate Tableau into Slack and other applications!

My Take

First, let me set the stage. In the past, I very often focused my data literacy and Tableau training efforts on business analysts who were developing graphs and dashboards in Tableau Desktop. The analysts would then share their creations with their team via Tableau Server. Great! Except, I didn’t spend much time with the consumers to show them how to navigate around Tableau Server, how to make use of features such as comments, Ask Data, or even subscriptions. Users would likely get emailed a URL and they would bookmark it and that's all that their experience would be. This trend focuses on putting data at the fingertips of consumers, in the tools they already use everyday, almost without ever knowing they're using Tableau. 

I envision our sales team needing to answer a client question and using Slack to ask a question that can be answered with a data set or dashboard on Tableau Server. Then asking follow up questions. Or a data owner that needs to be notified immediately if there is an issue with a data source being refreshed so they can resolve it. Taking things a step further, embedding Tableau into other applications that people use help give people a customized experience, and again, put data closer to where they are already working. For example, embedding Tableau into homegrown applications or tools like Salesforce, would extend the reach of data-driven decision making. Heck, even incorporating Tableau content into our internal search engine would make it easy and intuitive for users to find what they need.

Business (Data) Science


  • Streamline decision making
  • Extend the capabilities of business analysts
  • No-code/low-code tool to build predictive models
  • Connect to data and choose a target metric to predict
  • Artificial intelligence selects the best model
  • Model results can be shared

My Take

I want people to be excited to derive more insights from their data. In fact, I lead a project to think outside the descriptive analytics box and generate some business ideas that could be solved with predictive analytics. I’m a huge fan of the analytical maturity curve, as well as no-code or low-code tools and the democratization of data, so I love seeing others in my organization who are passionate about taking that next step in their data maturity. But, I’m also wary of making predictive analytics too easy for those who aren't properly trained, because if you get something wrong with a predictive model you could be making a decision based on incorrect information.

What I appreciate is that Tableau CRM (formerly known as Einstein Analytics) helps educate business users every step of the way. That said, users have to have a very good understanding of their data and what they're trying to achieve. Also - and I cannot emphasize this enough - data science is not magic. It might not find interesting patterns in your data. You may need to get more data or different features. 

As a trained data scientist, there are few things I think organizations can do to ensure success with business science. The first is to have an educational program to onboard data champions and citizen data scientists. The second would be to have a review process before a model goes into production. Perhaps, even generating a risk framework and process to calculate model drift. These thoughts apply to all tools that make it easier to create predictive models, such as Alteryx and DataRobot.

Data Management


  • With the proliferation of data, order is needed
  • Tableau Prep controls data chaos
  • Prep Conductor helps to scale
  • Add in Data Catalog and you've got a complete picture
  • Lineage allows more detailed insight into what is used where
  • Data quality warnings can be setup ad hoc or dynamically
  • Virtualization helps define governed tables of data
  • With centralized row level security
  • You can even import a business glossary from another tool

My Take

Yes yes yes yes yes! It may not be all that exciting to business users, but these capabilities make it easier to build data sets tailored to a specific need and gets us one step closer to that centralized data library we all crave. And while it may take a significant amount of effort to define every field in a data set and comment in every calculated field, you can't argue that those data sets are easier to use. 

Imagine being able to navigate to Tableau Server and see a list of available data sources. You can explore each one easily because each has been given a logical description as well as every field. Maybe you own one of those data sets and occasionally the data vendor sends the data late or with errors. Instead of worrying about notifying everyone who uses the data, you can set up a data quality warning that automatically gets applied when the data refresh fails. Perhaps one field, in particular, is of concern. You can find that field in the lineage section and see all the places it is used.

Ecosystem & Collaboration


  • There are gaps in wanting to be data driven and getting there
  • Investments in Community, Partners, and Platform are being made
  • Bring together various components into one symbiotic relationship
  • Tableau Academic is growing
  • With the need for data jobs and data skills accelerating
  • Help people and organization get value from their data faster

My Take

Who doesn't love the idea of a quick start guide or knowledge sharing? Tableau is looking to marshal its resources to get data into the hands of more people. There are many out there who not only love to help, but can extend the capabilities of the Tableau platform by creating third-party extensions and dashboard starters for common data sets. This makes it easy to get started quickly, especially if you're dealing with a not-so-simple data set. Tableau is making possible the sharing of collective knowledge, wisdom, and experience.

There is also a need to support data literacy and data culture at scale. The Data Leadership Collaborative and Do No Harm Guide are two excellent examples of Tableau's stewardship. I appreciate how much everyone associated with Tableau helps lift others up to realize the power of data. Specifically, Tableau Academic, which is helping to bridge the gap between the demand for people with data skills and the supply. This collaborative nature is only going to continue and will evolve. I'm seeing examples within the finance space, as we combine our efforts in managing data vendors.

Hopefully themes from the Tableau Conference keynote inspire your 2022 projects and goals as much as it did mine. I think Francois said it best when we he said:

Data needs to be easier to use and trust.