Dimensions 2.0
Enhanced support for metric dimensions, with improved filtering and segmentation capabilities.
Want to learn more about this project? Reach out to me for a detailed presentation on how this project evolved, including insights into the design process, drafts, decision-making, and brainstorming stages.
Project Info
Name
Dimensions 2.0
Role
Project Lead
UX/UI Designer
Year
2024
Background
PowerMetrics provided users the ability to filter and segment data by metric dimensions, but our customer feedback highlighted a demand for expanded control over these dimensions. There is a clear need for additional functionalities to better serve both end users and business stakeholders by delivering highly relevant metrics.
"
The current filtering mechanism, by selecting members, feels really restrictive. I need more flexibility.
Customer - Technical User
"
I need to configure metrics in a way that reduces confusion for my users. I need more control over the filters and segments they can access.
Partner - Technical User
"
I need to filter by both numerical and text columns, but that’s not an option right now.
Customer - Business User
"
For some of my customers’ metrics, I need default filters that can be adjusted but always reset to the original settings.
Partner - Technical User
Proposed opportunities
Enhanced Filter Capabilities:
Expanding filter functionality to include text, numerical, and date-based filters will empower users to filter metrics beyond member selection. This enhancement will support a more nuanced analysis and data segmentation.
New dimensions types:
To increase flexibility, we are broadening the types of dimensions to include not only text but also numerical (both continuous and discrete), dates, and durations. This expansion enables more detailed metrics filtering and better aligns with diverse user requirements.
General metric capabilities:
Users will now be able to define preset filters or establish default filters and segments tailored for business user needs. Additionally, to reduce confusion and enhance usability, dimensions that are unnecessary or could lead to misinterpretation can be hidden from view.
Guiding principle: Maximize capability vs. Maximize compatibility
At the project's outset, we needed to establish a strategy: should we prioritize maximizing capability or compatibility?
Maximize capability: We support as many features as possible for each metric type, even if other metric types don’t support them. We’ll need to define rules for handling downstream artifacts when metrics are reconfigured and features are no longer supported by the new metric type.
Maximize compatibility: We support a common set of features across all metric types, ensuring smooth transitions during reconfiguration without downstream effects. This approach maintains consistent and predictable behavior for all metrics, even if it means excluding features that cannot be universally supported.
We have elected to go for maximizing compatibility.
Enhanced filter capabilities:
The introduction of new dimension types will require new and more advanced filtering methods for each type.
Key Requirements:
- Text match filters.
- Numeric filters and range selection.
- Date range filters on temporal dimensions
Text match filters
This advanced filtering method goes beyond simply selecting members. It supports various operators like "Equals," "Contains," and "Ends with," and allows for additional conditions to refine your filtering.
Numeric filters and range selection
This new numeric filtering method offers multiple options: use a range slider to select values, or create more advanced conditions for precise filtering. Useful for examples such as "Filter out companies under $10K spend".
Date range filters on temporal dimensions
Time-based filtering enables users to apply date ranges to specific dimensions, which is essential for those seeking trend analysis and historical insights. Useful for examples such as "Filter out companies with sign up date before Jan 1 2022, or show me arrival time vs departure time".
New dimensions types:
New types of dimensions to support:
- Text dimensions (Already supported).
- Numeric (discrete/continuous) dimensions.
- Time (date/duration) dimensions.
General metric capabilities:
Key Requirements:
- Enable preset filters.
- Enable default filters.
- Enable dimensions capabilities.
Preset Filters
The value of a preset filter lies in its ability to refine a metric. For example, if I have a dimension called 'user events' that includes members such as 'Churn,' 'Purchase,' and 'Extensions,' I can use a preset filter to display only 'Churn' to define the metric and for more focused analysis.
Dimensions capabilities
Control whether a dimension appears as a filter, a segment, or is hidden from users entirely.
Default Filters
A default filter sets the initial presentation of a metric while still allowing the end user to modify the filter or segment as needed.
Key business and product outcomes
- 80% of users self-identified as integrators have used the advanced filters and dimensions to set up their metrics.
- Increased customer requests to include further advanced features such as binning and grouping.
