Tableau Embedded Analytics for Your Corporation
Launching a new tool is tricky. Beyond the apparent technical challenge of ideating, developing, releasing, and maintaining a suite of tools, businesses also need to integrate these new assets into an intricate network of established business processes and ways of working – in the Tableau space, this is known as embedding. To succeed in this Tableau implementation for your business, you’ll need a strategic mindset and a keen awareness of the potential obstacles and pitfalls.
Whatever your role is in this analytical lifecycle, this blog is tailored just for you. My name is William Guicheney, a Senior Analytics Consultant at Aimpoint Digital. I’m here to share three critical, hands-on steps to ensure your Tableau (or any other BI tool, for that matter) is a resounding success!
Start With a Monitoring & Evaluation System
What gets measured gets managed.
When considering the effectiveness of the data products they produce, this simple rule often needs to be noticed by developers. We need clearly defined success metrics & a set of tools for evaluating our progress toward them to ensure our ability to make truly evidence-based decisions when deciding which dashboards to improve and which to scrap.
This data-driven approach is a core pillar of Tableau’s philosophy – and this is why their Cloud offering comes packed with a series of activity monitoring dashboards grouped into the Admin Insights Starter workbook.
This workbook is an excellent way to start thinking about the metrics you should use to monitor your rollout’s success. However, we believe producing these dashboards – selecting the data to use, iterating over their design, and reviewing them with a team of trusted developers & business owners – is a crucial first step in getting your Tableau implementation started.
So, before looking for use cases and power users, begin by building your own monitoring & evaluation system! When designing it, you should keep four questions in mind:
1. What does success mean for the implementation of your BI tool?
- Here, we recommend splitting these measures into short-term outputs and long-term impacts.
- Outputs could simply be achieving a certain number of monthly active users or a specific monthly growth in the number of views of key dashboards. Impacts are better measured from the business side and could include a certain number of hours saved, pulling and processing data each month, or a gain in revenue from better insights made available to front-line sales staff.
2. Who should be involved in the group evaluating this success?
- To keep the Tableau implementation management for your business effective and efficient, we recommend assigning only a small but dedicated team to its regular monitoring.
- The Tableau implementation monitoring should involve key developers, preferably a combination of front and back-end developers, to expedite the process of making any changes to the data warehouse needed to improve the dashboards, as well as a project leader from the business side who will oversee the collection of user feedback and identification of demands.
3. When is this group going to meet to perform this evaluation?
- Monitoring tools are only helpful if you regularly check them (or they alert you)!
- Make sure to set up a recurring weekly meeting (long enough to have time to discuss things, short enough that it happens every week) involving all key team members.
- You can use this slot to regularly invite power users you have identified and have them share their feedback directly with the implementation team.
- Pro tip: set up some Subscriptions and/or Alerts so that your core monitoring team automatically receives information from your dashboards through e-mail or Slack!
4. Which monitoring tools will they be using to execute this evaluation?
- Thankfully, Tableau’s excellent suite of integrated products makes it incredibly easy to access and build upon user usage data.
- As users create, publish, and access workbooks, this data is logged and made available to admins through a series of well-organized and clean data sources. You can view them in the Admin Insights area of your Tableau Cloud instance.
As stated above, we recommend building your own dashboards! It’s an excellent way to start putting your new tools to the test. The data is also straightforward to work with within Tableau and can be used to showcase the tool’s incredible potential for producing beautiful interactive data stories.
- To start, open Tableau Desktop, hit Data, Create a new Data Source, and select Tableau Server at the top. Once you are there, search for “TS,” and two data sources should pop up:
- TS Events stores records about each of the individual actions your users took on Tableau Online or Server,
- TS Users stores pre-aggregated user information that can be useful for determining who your users are, when they were active, etc.
- Are you a Tableau Server user? Fret not! You can easily access your usage data by connecting directly to the Tableau Server repository which is a PostgreSQL database that stores all key data about the activity on your account. Accessing this data is not the focus of this article, but we recommend checking out this article to get started if that would be of interest to you.
- We recommend working with TS Events as it stores the most crucial usage data we want to monitor.
- Here are two more pro tips to help you get started: if you are working with Tableau Online, the usage data made available to you will only include the past 90 days of data. To get around that, we recommend setting up an Extract for that data source and configuring it to be incremental using the configuration shown below.
- The Extract setup for the data source will tell Tableau to load any new rows of data whenever the extract process runs and to use the Event Date to determine which records are new.
- Over time, this will progressively build a historical dataset of all the user activity, appending each new day’s worth of data to your extract.
- Furthermore, we recommend filtering down the data in TS Events to only look at dashboard views, as this is the user behavior that will matter to us when evaluating the success of our rollout.
- To do so, set the Event Type to only include Access and the Item Type to only have Viewed, and set this filter to apply to the entire data source.
- Of course, if you are interested in measuring whether users are producing & publishing their dashboards, you should change these filters appropriately!
Once you have:
- Built your suite of monitoring dashboards,
- Created your recurring meeting to monitor these dashboards and discuss the insights they show,
- Assembled a team of developers & power users to act upon this data,
- Determined the key performance indicators you would use to evaluate this data.
You are ready to start absorbing demand for new Tableau dashboards from the business!
Please be sure to read parts 2 and 3 of this series on creating demand-oriented workstreams and building a governance system, and reach out through the form below if you would like to learn more about how Aimpoint Digital can support your business with Tableau.
