Dashboards 101 – Choose a Tool and Get Started

The Case for Dashboards

Manually pulling data for a report is the bane of any analyst. Even when you have a documented process, nothing is as inefficient as generating that data repeatedly and on demand. More importantly, manual reporting leaves behind a complicated reporting process for others to learn and potentially bungle. Every manual click is another opportunity for mistakes.
Automating what you can allows you to focus attention on deeper insights. Instead of spending time copying, pasting and rechecking data, you can now use it to compare time frames, channel performance and other points of interest. Furthermore, a holistic dashboard that combines data sources makes it far easier to present companywide data in a compelling way. Tell the whole story, not 10 little stories. And do not waste your time doing it repeatedly every month.

Step 1: Assess Current Reporting

Before any other considerations, the analyst should start by reviewing current reporting structure. Take notes on how KPIs are currently being tracked and when reports are timed for release. A dashboard does not need to be a complete departure from previous reporting; it should incorporate most of the things that executives wanted to see previously. The purpose of the dashboard is to make this data more easily accessible, not to reinvent the wheel.
Depending on the scope of your dashboard, you will likely need to integrate data from several different external sources like social platforms, company databases and mailing services. Meeting these needs will be your top priority when choosing a visualization tool. Looking at the current reporting structure, and how it has evolved over time, supplies valuable insights when building dash metrics. You will also need practical help—for instance, getting permissions from team accounts for every data source you connect.
Even beyond the exploratory phase, teamwork is pivotal if you want to produce a truly compelling dashboard. A good analyst will revise their dashboard repeatedly based on stakeholder input. Ask your team how they think metrics should be prioritized, which time frames are most critical to look at, and how dashboard data should be segmented to be of most use to stakeholders. These are things an analyst can guess at, but they ultimately come down to the unique needs of the team.

Step 2: Assess Integration Requirements

Far and away the most important consideration when assessing data visualization tools is their ability to integrate with relevant data platforms. In a recent dashboard project we needed to pull client data from LinkedIn, Facebook, Twitter, Instagram, Mailchimp, Google Ads and Google Analytics. Given that so much of our data was related to Google’s suite, we decided on Data Studio as our tool of choice.
Practically speaking, you would be very lucky to find a data visualization platform that allows you to connect every data source you need directly. Most likely you will need to find workarounds for some sources of data. The point here is not to find your perfect solution, but to filter out the platforms that cannot meet your basic reporting needs. For instance, Data Studio does not have a native connection for Mailchimp, so we worked around it by calling Mailchimp data automatically into a Google Spreadsheet, which could then be uploaded directly as a data source into Data Studio. In another instance we decided to use a non-native connector to consolidate social data.
Despite these necessary workarounds 90% of our reporting in this case was based on data from Google Analytics, so Data Studio was ultimately the best choice for the job. However, your tool of choice can (and should) vary based on the data at hand. Working with your team to assess this is a critical first step in the data visualization process that many analysts skip because they are tied to a specific platform.

Step 3: Assess Cost & Complexity

Once you have narrowed your list of visualization tools to incorporate the data sources you will be working with, there are the more practical problems. Each tool has trade-offs in terms of functionality, price and complexity. An experienced analyst should research the main players in this field so that they know which tools are best for the job at hand. Two very different front runners that demonstrate this are Tableau and Data Studio.
Tableau is an enterprise solution that has obvious benefits in terms of the complexity of calculations and parameters it can handle. If the KPIs you are visualizing involve complex combinations of data, a tool like Tableau may be necessary. It gives you a high level of report customization, and its data cleaning and preparation options are top-notch, putting it above other enterprise options like SAP. Like other tools targeting the enterprise market (Sisense, Power BI), Tableau has excellent database connectors and allows you to manipulate data with querying languages like SQL and R. However, there is a steeper learning curve for Tableau, and its cost can be prohibitive for smaller companies.
Data Studio is as far from enterprise as you can get, because, well, it’s free. Although custom reporting is more restricted, new fields and filters can be created using regex to create surprising levels of customization. Data Studio is also pretty much unbeatable in terms of sharing options, and its drag-and-drop interface allows for more visually pleasing reports than you would likely get from Tableau. Furthermore, unlike many other drag-and-drop options, Data Studio has a robust set of connections simply by being part of the Google suite of apps. If, like us, you are a marketing firm that relies on Google Analytics and Ads, Data Studio is a great pick, as it was literally built to visualize those data sources.

Conclusion

You should be skeptical if someone tells you that Tableau or Data Studio is your end-all answer to data visualization. Your choice of BI tool should depend primarily on the data your team needs to visualize.
Although setup time is not trivial, dashboards can drastically reduce the time needed to generate comprehensive reports and will almost always be more efficient than long-term manual reporting. You do not need to have a finished product overnight. Start by recreating reporting data in your dashboard to see immediate value.
The highest priority of an analyst should be drawing actionable insights from data. But this is only accomplished when the analyst understands what insights are most important for dashboard stakeholders. Collaboration is key. Reporting structure, and even the tool you use to visualize your data, will most likely depend on the individual needs of the dashboard’s owner. The end product should be a reflection of your department’s reporting needs.

Keep reading in Analytics