Maintaining Explorations
Communication
As explorations are tools that you are building for your Data Consumers, it's really important to have a proper communication process with your stakeholders to keep them in the loop of any changes / enhancements that you have done or are planning to do.
As well, to create impactful exploration, you'll ask to gather from user feedbacks to make sure that you configure the Exploration in the right way.
Hence, you have to create a communication channel with your end users. Depending on how your company is structured, here are some good ideas that you can follow:
Create an internal newsletter / Slack channel to discuss all changes related to your Explorations and send notifications to your Data Consumers whenever there is something interesting to share with them
Keep a "changelog" document in which you log all changes done on the Explorations configurations so that people can discover all the new things and understand the changes that are happening
Send periodically User Feedback forms / NPS survey to your Data Consumers to hear their voice and be sure to build something that they want
Share examples of great usage of Explorations that you can produce by yourself and that you saw across your organization, this will give ideas to people on how to use the exploration you created
Deprecation process
Whenever you have to make something disappear from an Exploration, you should follow a deprecation process in order to make things properly and give way to your Data Consumer to adapt to the change.
Check the usage of what will be deprecated: by checking the logs and the dashboards/question using the dimensions/metrics that will be deprecated, you will identity who will be impacted the most and you should priorize them in your communication to be sure they understand what will happen
If needed, design a "migration strategy/alternatives" for Data Consumers to help them understand how they should adapt their workflows
Mark the dimension / metric as being DEPRECATED in its name in the Exploration to avoid future use and to make sure that people will migrate their queries to the proper dimensions / metrics that are not marked as deprecated
Define a deprecation schedule, with the date of the deprecation and the date of the final deletion of the dimensions / metrics. The deprecation period length should give enough time for Data Comsumer to migrate their work while still being short enough to move quickly
Communicate clearly to all stakeholders the deprecation schedule, the reason behind the deprecation and make sure that they got the message
A few day before the "delete date", make sure that people have done what they needed to let you delete what you should
At the communicated date, delete the dimension / metric and log its deletion in the exploration changelog
Depending on your organization and to your current adoption rate, a more lightweight process could be followed, but keep in mind that people need security and stable interface to make productive work.
Breaking their workflow will produce frustration and will lower their trust into your data solution.
Update process
Similarly to the deprecation process, whenever a change of calculation of a metric or of how a dimension is build, a proper process has to be followed so that you onboard people on the change.
Refining documentation
Documentation is a never ending story, every time that someone ask a question to your support, check if the answer was present in the documentation.
What is properly explained?
What it easy to find the answer?
Was the documentation available to the Data Consumer when he got the question?
The documentation is layered across different interfaces and each one should be properly filled to ensure a global adoption of your Explorations:
Naming: The naming of your Exploration / Dimensions / Metrics is really important as it's the first thing that people see. It should use a vocabulary that everyone understand accross your organization.
Metrics / Dimensions description: How things are calculated and how they should be used, alongside their business meaning and impact should be properly documented in the Metrics / Dimensions description.
Exploration description: How the overall Exploration is positioned in the global context of your organization should be added in the Exploration description. Some examples of the questions that this Exploration can answer should also be added to help people understand how they could use it and if it's for them.
Global vocabulary / glossary: You should maintain somewhere a list of your Business concepts definitions. Like "What is a Sales Qualified Lead"? This should be shared across your entire organization and will help people understanding the data
Get user feedback
The more you talk to your users, the best insights you'll have on what are their data needs. As well, they will indicate you the proper direction to build the best Exploration that will increase their Data Adoption and increase the impact of Data analysis across your organisation.
Some ideas on how to collect user feedback:
After having trained someone on how to use an Exploration, ask for their direct feedback
Send polls / forms to your most frequent users
Create a "Submit ideas" process to be share across the organization so that people can share their ideas on how to make everything progress
Last updated