10 Steps to a Data Dashboard

SIMPLE CLOUD SOLUTIONS

1-866-802-9335

©2019 by Simple Cloud Solutions

10 Steps to a Data Dashboard

January 19, 2018

Modern dashboard solutions sure are sexy. It is very difficult to avoid diving right in with these tools. Especially with their templated widgets which show web and social media activity in minutes. Those are easy, but the rest require careful planning and focused effort.

 

Resist the urge to dive in! Slow down and follow these steps with key stakeholders:

 

Each step will list an "Output". You should generate this before moving onto the next step.

 

1. Why do you measure?

 

Pull way back and really ask yourself why you measure activities. Precisely what is the benefit to the organization?

 

There are generally two motivators: accountability to a funder or manager and/or continuously improving products or services. I prefer the latter motivation, but the former is certainly important to excel on as well.

 

Use this framework and get specific on how measurement will improve the organization. It should be high-level at this stage. Write it down and agree!

 

Next you must clearly identify who the audiences are. Likely the audience will vary depending on the level of reporting and specific measures. The best measures will serve all audiences and will allow drill-down detail.

 

Outputs:

  • Clearly defined rational for measuring

  • Target audiences

 

2. Define your programs

 

Programs are the highest level segments of a reporting system. This should be pretty straightforward. A program might be a Human Resources, a Community House, or Primary Care Outreach. It's a good idea to tag programs with Departments as well, to enable roll-up where possible.

 

Output:

  • A simple Program List with tags such as departments

 

3. List program objectives and reporting requirements

 

Recall that the two reasons to report are to satisfy reporting requirements from an authority and also to continuously improve.

 

To that end, start with mandated measures. Those are easy. Next, and more importantly, list the highest level program objectives (no metrics yet).

 

This Program List should be a formal and tightly controlled document. Changes should be tracked and minuted at the senior management level. We'll come back to that in section 6.

 

Output:

  • Official Program List (OPL) which now includes program objectives and reporting requirements.

 

4. List your data sources

 

List all possible data sources, grade their quality, and identify steps required to prepare them to act as "sources of truth". This includes data entry methods and instructions to ensure future data quality, even when staff turn over.

 

We have all heard "garbage in, garbage out"; well this step contributes to avoiding that unfortunately common outcome.

 

Output:

  • Official Data Sources (ODS)

 

5. Define your metrics

 

For each measure, you should have the following:

  • Program and Objective: (From steps 2 and 3)

  • Explicit calculation including time periods (e.g. monthly, quarterly, etc)

 

Less is more! I can't emphasize this enough. Be selective and consider what will have the greatest impact. Where possible, choose metrics that satisfy all audiences. A simple example is all client contents. This is easy to roll up for the Board, and drill-down for Program Directors.

 

Outputs:

  • Official Metrics List (OML)

 

Celebrate! You are on your way!

 

At this point you have 3 draft documents that are the foundation of the next phase:

  1. Official Program List (OPL)

  2. Official Metric List (OML)

  3. Official Data Sources (ODS)

 

6. Establish Controls via Policy

 

Create a policy that requires senior management approval. Have that policy reference the 3 key documents produced above and state that changes are to be approved. 

 

It is a good idea to create a space on your intranet where these documents are posted for reference.

 

The Official Programs List can also serve as a great baseline for other efforts such as Communications planning.

 

Output:

  • Policy - Program, Data and Reporting Controls

 

7. Optimize Data Sources

 

Create a Request for Quote (RFQ) which contains all of the issues and required changes to prepare your data sources to function as the source of truth. Collect bids and move ahead with correcting the issues in a sustained way. Meaning, don't just clean-up the historical. Ensure that data entry standards are set for the future.

 

It is recommended that the person or company selected to complete this work be retained for future related work. Such continuity will help.

 

Tip: At this stage, aim for a reasonable level of data quality to support business decision making. Meaning, broad trends, which, when combined with experience, can support planning and reacting. A common error is to seek the level of accuracy required for scientific research. That is not necessary when building a system for business decision making and will stifle the process.

 

Output:

  • Cleaned historical data which can be connected to the Dashboard

  • Updated Official Data Source document with a Data Entry Guide section

 

8. Design the Dashboard and Data Model

 

Now it's finally time for the big payoff. Your data is clean, your Official Program List, Metric List, and Data Sources are established and controlled. It's time to get some beautiful, and more importantly, helpful visuals going.

 

Since you've done your homework, this step is easy. Write a Request for Quote, and attach your 3 key documents as appendices. Seek the services of someone with experience developing Dashboards.

 

Leaders in this field include:

 

Klipfolio (my favourite, and right here in Ottawa!)

Microsoft BI

Google Analytics | Data Studio

New Relic

Domo

Tableau

 

A key requirement you must include is automation. I strongly discourage solutions that require periodic intervention beyond front-line data entry. This is possible with good design and you should demand that prospective developers demonstrate their ability to automate the refreshing of data sources. It's a good idea to get used to Google Sheets as a simple yet highly effective data source.

 

Finally, thoroughly test with a small pilot group of users before going live. This is key to a successful launch. You have the opportunity to really wow the audience if you're careful and patient to get it absolutely right on day 1.

 

Output:

  • An amazing, beautiful and automated Dashboard! Yeah!

 

9. Go Live and Celebrate!

 

You have built something great, and likely at a very reasonable cost! Congratulations!

 

Tell everyone! Hold formal training and education sessions, hold lunch and learns, integrate into your intranet where appropriate. You've given your organization super-human insight into its operation!

 

 

10. Control and Keep Fresh

 

Don't stop at the celebration! In the policy created in step 6, you must include a procedure that all changes to the 3 key documents must be reflected in the Dashboard and its data model.

 

Using a Statement of Work, define the ongoing / as needed maintenance of the Dashboard and Data Model. All approved changes are then routed to the assigned contractor.

 

This routine will ensure the quality and effectiveness is retained following the launch. This is critical and cannot be excluded. 

 

Get yourself an inexpensive TV and put your Dashboards in a prominent location for staff and visitors. You will blow them away!

 

Congratulations, Data Hero.

 

Please reload

Featured Posts

Telus EMR Mobile - A Great Step Forward

February 21, 2018

1/2
Please reload

Recent Posts

August 8, 2019

February 6, 2019

December 10, 2018

Please reload

Archive