Best Practices: Creating a Workspace – Stratifyd

Before building a workspace, consider both your short and long-term goals. You'll want to think through your data sources, historical data, metadata parsing, and data categorization needs. We've assembled a list of considerations for creating a new workspace below. These are intended as a starting point only, and we'd love your input on additional factors to consider.

Which data sources am I interested in learning from? Do I need to combine multiple sources to get the full picture?

You have the ability look at sources side by side, or combine using ETL. To see all feedback in one stream (run analyses on combined text field, combine metadata) you'll want to use ETL. If checking for discrepancies between sources is important, keep streams separate.

How far back do I need to look for trends? Are historical trends important to my analysis?

Including historical data in a dashboard can be very helpful to track trends. However, a dataset too large can hinder the ability for models to display the new invigorating stories. To ensure we're capturing the latest topics, it may be best to use different widget, tabs, or dashboards for recent vs historic trends.

Which metadata will help your teams analyze?

Many times, starting an analysis using metadata can help you zero in on important data (product name, star ratings, type of request). Having this information readily available will make it easier for your stakeholders to begin analyzing. You can display this in widgets or stories.

Who needs to see this data?

Thinking of the different teams who would benefit from this data can help guide us to build taxonomies, dashboards and widgets. As we're building workspaces, continually thinking of our end users will help us calibrate with all the above questions.

Which widgets will bring your data to life?

Displaying both original data and Stratifyd analyses is very important to fully represent your data. Being able to start an analysis using metadata, and then checking for analysis trends and vice versa will show different reactions to various products or services by varying subsets of our customer base (depending on the data you're working with).

What is the best practice for sharing workspaces?

Sharing workspaces broadly with 'View Only' access is a great way to determine who's interested in which data. Then, grant 'Ownership' of workspaces to certain people upon request. This will ensure everyone who needs the workspace will have access, but you won't run the risk of having changes applied without your knowledge.

How much data should I show per tab?

Showing certain subsets of data (ex: just metadata) or certain time periods (a specific month or quarter) will show smaller, digestible amounts of data at a time. This also allows you to pull out smaller volume, trending conversations from the larger dataset. Configuring your workspaces to display only pertinent data will help your data become more widely consumed.

Further questions?

We're here to help! Don't hesitate to contact us for further assistance via chat or submit a ticket!