Concept: Storyteller Engine
People Analytics uses an automated analytical engine called Storyteller that:
- Searches millions of combinations of data.
- Makes connections between the combinations of data.
- Surfaces the most significant results in the form of stories.
A story is a finding in the data that addresses a business question. Example: What are key
turnover trends?
Each story describes an instance
of significant under- or overperformance in a metric, along with a combination of
dimensions, trends, and drivers that contribute to the finding.
Storyteller uses statistical and heuristic methods to determine
the ranking of stories and which stories should surface after performing one of these
comparisons:
- Comparison to historical performance - Storyteller surfaces these stories when the engine detects a significant variance between the current performance of a metric against historical performance. The look-back period is usually 12 months if you have at least 13 months of data history in People Analytics, or 3 months if you have less than 13 months of data history. However, the look-back period might vary depending on the metric.
- Internal peer group comparison - Storyteller compares current performance in a given organization or location (region, country, or other location) against the performance of peers within the company and surfaces significant gaps as stories. Depending which level the story is on, the peer group changes. Storyteller might make the comparison at the company or region level, or at a different level in the hierarchy. By comparing performance to the closest peers in the same organization or region, Storyteller can identify stories that provide insights which are actionable and meaningful in relation to the dimension where the story takes place. Examples:
- For a story on FP&A suborganization in Finance at Level 3, the Storyteller engine compares performance of this dimension against the performance of the parent organization Level 2 (instead of comparing to the performance of Level 1).
- For a story in Canada, Storyteller identifies stories by comparing performance in Canada against performance at the region level, such as North America or the Americas.
Automatic Filtering
To help ensure the highest quality of insights, Storyteller automatically filters out stories where:
- The population size is 5 or less people.
- There is no active headcount for 1 or more months that are included in the story analysis. This filtering only applies to stories in these topic areas:
- Diversity and Inclusion
- Organization Composition
- Retention and Attrition
- Talent and Performance
Example: A story on a Female Diversity Trend where there is no active headcount for 1 or more of the 12 months analyzed.Example: A story on a Female Diversity Gap where there is no active headcount for the month analyzed.