Setting a Talent Framework
All roles in the organisation should be defined in a systematic way; each is a collection of 3-4 skills and can be the same across teams in the organisation
Identify the senior leaders who will own the roles in the organisation and are in charge of defining what good looks like for the role and designing the interview process
You can use scorecards describing observable behaviours to define the different mastery levels for each skill, providing an objective way to assess potential and existing employees
The talent bar determines the minimum required mastery levels for each skill and seniority level for a given role, setting expectations for a candidate's performance
Without a clear definition of what good actually looks like, even the best interview structure won’t guarantee consistent decisions. That’s where a talent framework comes in.
It aligns everyone -recruiters, hiring managers, and interviewers- on the skills, behaviours, and expectations required for success at each level. It ensures your hiring bar is applied consistently across teams, and becomes the foundation for performance and promotion once someone is in the door, which you can read more about in our performance playbook.
How to build a talent framework
There are fundamentally three steps to creating a talent framework:
- Standardise roles - push for assessing similar roles on a common set of skills. For example, you should expect the same quality of engineers across all teams, regardless of who they report to.
- Define skills with scorecards - create scorecards with “ideal behaviours” that you need to see in candidates for each of the skills in your organisation. Managers won’t be able to deviate from the scorecards when hiring.
- Set a clear talent bar - determine what level of skill each role in your organisation is expected to conquer. For example, you wouldn’t look for the same signs of problem-solving from a junior data analyst and your next CFO.
Defining roles
Roles are defined as a collection of skills and are universal across teams
Sales Team
Finance Team
Product Team
Role Owner
Data AnalystData AnalystData AnalystData Analyst
Skills
PythonData VisualisationProblem SolvingA role is defined as a collection of 3-4 skills, based on what is required for the job and is universal for anyone who uses the same skills in the company, regardless of the department, product or team.
For each role, you should identify a role owner. These are senior leaders in your organization who can define what “good” looks like and are responsible for designing the interview processes, including the questions to be asked. For example, the Head of Sales can be the functional leader for Salespeople and the CTO for Engineers.
Use scorecards to standardise assessments of skills
Consistency is the key characteristic of a talent framework. Without it, managers might apply different standards, compromising the quality of talent of your organisation.
Scorecards translate skills into concrete, true/false statements about observable behaviours. They are used during interviews to guide evaluation and after hiring to support performance reviews and promotions.
You should build scorecards with 2 things in mind:
- Consistent scoring - the same 5-level scale is used across all scorecards: Poor, Basic, Intermediate, Advanced, Exceptional. Poor refers to “red flag” behaviours - no one should exhibit these. Exceptional is the level that only the very best in the field should hit.
- Observable behaviours - Think about the behaviours you’d want to see in your employees and build a scorecard around them. This approach reduces manager bias by giving them precise behaviours to track in performance reviews and interviews while providing employees with clear goalposts to hit.
The talent bar
The scorecards provide useful guardrails to understand whether an employee or candidate has mastered a specific skill - but how do you know what good enough is?
This is where the talent bar comes into place. This defines the minimum required mastery level for a given skill and seniority. It serves as a structured guide that aligns expectations for different roles within the company and establishes a clear benchmark by which candidates can be assessed.
All employees are expected to practice their role’s skills at the level mandated by the talent bar for their seniority; when interviewing, you need to make sure candidates also meet the talent bar.
Example - Data Scientist
Performance score calculation: Apply methodology, statistical inference, and offline machine learning models to drive business decisions.
Talent bar:
Data Analysis skill scorecard:
- Gets lost in datasets. Struggles to find simple trends or insights.
- Chooses visualisations that are not fit for presenting the given data. Doesn’t give a clear overview, or even hides characteristics due to poor choice of dimensions.
- Comprehends and finds insights in large datasets.
- Understands business objectives, and can come up with the right metrics to measure.
- Appropriately uses descriptive statistics (e.g. sum, avg, median, rolling calculations).
- Builds simple, clear visualisations, focusing on important information.
- Provides simple business recommendations based on their analyses with demonstrated examples.
- Based on business objectives, drives analyses independently. For example, creates comprehensive dashboards focusing on business objective, without stakeholders having to define list of required calculations.
- Structures and presents complex information in a highly effective way
- Provides recommendations for major business decisions and processes with demonstrated examples
- Examines data from multiple angles, ensuring that all relevant underlying info is uncovered and presented
- Thinks of their analyses in terms of the bigger picture, can connect dots between different areas. E.g. specific products’ or processes’ contribution to company objectives, associated risks, interdependencies
- Provides trusted business recommendations for the company’s most important decisions with demonstrated examples. Authoritative knowledge of which metrics provide the right insights