WEBINAR HR Analytics aplicaciones y casos de exito

Introduction to the Webinar

Welcome and Overview

• The webinar is organized by RH, focusing on "Ser Analytics," its applications, and success cases.

• Sonia Rodríguez Sobrino, a consultant in Talent Analytics at the Institute of Knowledge Engineering, is introduced as the speaker.

• Attendees are encouraged to ask questions during the presentation for later discussion.

Understanding Talent Analytics

What is Talent Analytics?

• Talent analytics is described as an integrated data analysis methodology that provides evidence for quality decision-making regarding personnel.

• The goal is to improve individual and organizational performance through objective information rather than subjective opinions.

Importance of Data in HR

• A report from 2017/2018 indicates that approximately 78% of companies recognize the importance of analytics in HR.

• About 40% of these companies claim they have necessary skills and technology for data analytics but often lack sufficient data to start projects.

Current Trends in HR Analytics

Organizational Readiness

• Around 70% of organizations with over 100 employees have teams or individuals beginning talent analytics projects.

• There’s a notable increase in LinkedIn profiles featuring skills related to HR analytics, indicating growing interest and demand.

Linking Data with Business Actions

• Approximately 50% of companies are starting to connect their HR data with business outcomes, showing a trend towards integrating analytics into strategic planning.

Analytical Maturity Levels

State of Analytical Development

• An IBM survey reveals that about 60% of companies are at an initial stage regarding analytical capabilities; only 25% have developed a strong data culture across departments.

Data-Driven Decision Making in Human Resources

The Demand for Data Integration

• There is a clear demand to link data with business operations, supported by high-level management in some companies.

• Many organizations have trained personnel who are aware of the challenges that hinder growth, particularly regarding clarity on internal versus external data analysis.

Barriers to Progress

• Key barriers include technology and infrastructure limitations, as well as a lack of a data-driven culture within organizations.

• The need for more information from human resources is driven by professionals' requirements to manage businesses effectively; understanding individual employees is crucial.

Importance of Employee Data

• Gathering employee-related information is essential for enhancing engagement and retention, especially among millennials who seek personalized services.

• Organizations must utilize analytics to identify problems and leverage data insights for solutions, focusing on both profiling and predictive projects.

Types of Analytical Projects

• Two main types of analytical projects are identified: profiling (to understand employee characteristics and prevent turnover) and prediction (to anticipate candidate success or potential employee exits).

• Examples include identifying suitable candidates during recruitment or assessing risks associated with employee turnover.

Project Focus and Methodology

• Projects should be guided by specific questions or problems; this focus determines the type of analysis required.

• Different analytical approaches depend on whether the goal is historical analysis or future predictions, which may require complex methodologies like optimization simulations.

Understanding Different Levels of Analytics

Descriptive Analytics

• Descriptive analytics is the most basic level, focusing on describing the current state and functioning of systems. It involves fixed information about what is happening.

Predictive Analytics

• Predictive analytics goes beyond description, aiming to anticipate future events, such as fraud detection in banking transactions.

Prescriptive Analytics

• Prescriptive analytics relates to recommendation systems that suggest optimal strategies for achieving desired outcomes. This level is complex and less commonly applied in organizations.

Methodology for Implementing Analytical Projects

Steps in Project Implementation

• The recommended approach includes starting with a pilot project in a specific area before expanding it based on initial success.

Step 1: Problem Identification

• Clearly define the problem or question at hand, involving all relevant stakeholders to ensure comprehensive understanding and agreement.

Step 2: Research Design

• Develop a research design outlining procedures and data collection methods, which helps identify potential deviations during project progress.

Step 3: Data Identification

• This phase can be extensive; data may be scattered across different databases or difficult to access. Quality and completeness of data are crucial for effective analysis.

Step 4: Advanced Statistics Application

• Utilize various analytical techniques including inferential statistics and complex models like neural networks for predictions.

Integration and Collaboration in Data Projects

Finalizing the Project

• The outcome may vary from delivering a simple report to automating intelligence solutions integrated into company systems for daily use.

Importance of Multidisciplinary Teams

• A diverse team comprising data architects, business analysts, IT specialists, and legal advisors is essential for project success from inception through execution.

Ensuring Data Quality

• Guaranteeing high-quality data collection is vital as it directly impacts the validity of analysis conclusions. Poor quality leads to unreliable results.

Types of Data in Human Resources

Data Collection Challenges

• In HR, while there are many available data types (e.g., resumes, performance history), they often lack expected quality levels. Awareness around proper storage and processing has increased recently.

Analyzing HR Data

• Proper analysis and interpretation of HR data are necessary for making informed decisions that benefit organizational goals.

Insights on Employee Selection and Retention

Importance of Psychological Evaluation in Recruitment

• The discussion emphasizes the need for psychological evaluation tools in employee selection processes to gather personal information about candidates.

• It highlights that when recruiting for specific roles, such as telecommunications engineers, profiles can be very similar, making it crucial to have additional evaluative data.

• Contextual information like economic conditions and social climate is essential for improving prediction models related to employee retention.

Business Metrics and Predictive Analytics

• The project aims to integrate business metrics such as financial data and accident statistics into predictive analytics models to enhance their effectiveness.

• A case study is introduced focusing on predicting employee turnover rates within a software factory facing a 30% voluntary resignation rate.

Analyzing Factors Influencing Employee Turnover

• The project's objective was to identify significant variables influencing employee departures and develop a model for predicting potential resignations.

• A combination of descriptive inferential analysis and complex predictive analytics was employed using existing company data without future projections.

Key Variables Identified in Turnover Analysis

• Important factors leading to employee turnover included salary increments, training opportunities, recognition through awards, and client engagement levels.

• Insights from the analysis provided actionable recommendations for management regarding salary reviews and retention strategies based on identified variables.

Financial Implications of Retention Strategies

• The project also focused on quantifying potential savings from improved retention rates by applying the predictive model effectively.

• Understanding the financial benefits of retaining employees is crucial; it helps justify investments in analytical projects aimed at reducing turnover.

Optimizing Selection Processes with Data

Challenges in Processing Applications

• Another case discusses optimizing the selection process for an internship program overwhelmed with applications, leading to missed opportunities with qualified candidates.

Developing Success Prediction Algorithms

• The goal was to identify success factors for candidates early in the selection process using algorithms that predict candidate success based on their resumes.

Integration with Existing Systems

• The solution involved integrating this algorithm with the company's Applicant Tracking System (ATS), allowing recruiters immediate access during candidate evaluations.

Methodology Employed in Candidate Assessment

• A longitudinal study collected curriculum data alongside personality assessments (Big Five model), aptitude tests, and competency evaluations tailored for commercial roles.

Analysis of Predictive Models in Recruitment

Statistical Significance in Descriptive Analysis

• The descriptive analysis revealed statistical significance in certain variables, indicating that success is not due to chance. Successful individuals scored higher on the assessment and exhibited greater general intelligence.

• Variables such as prior work experience and internships were identified as important indicators for successful candidates.

Effectiveness of Predictive Models

• A predictive model was developed with a high accuracy rate; it could identify nine successful candidates from 300 resumes, suggesting potential for doubling this number with improved methods.

• The current model required reviewing 300 resumes to select ten candidates, while the new model aims to reduce this number significantly.

Complexity of Predictive Analytics

• Complex analytical models are often difficult to interpret, functioning like "black boxes" where variable weights can vary significantly between cases.

• Emphasis is placed on providing more digestible analytics for clients by focusing on frequently occurring variables that contribute to success.

Case Studies in Employee Profiling

• A case study involved profiling employees based on business needs, aiming to identify characteristics that influence profitability and absenteeism rates.

• Key factors affecting profitability included age, group occupation level, and salary; these were analyzed alongside absenteeism data.

Modeling Absenteeism and Accidents

• Two predictive models were created: one linear for predicting days absent due to illness and another non-linear focused on the probability of sickness-related absenteeism.

• There was also interest in exploring accident prediction models, which would require more historical data for effective implementation.

Enhancing HR Dashboards with Business Metrics

• Current HR dashboards often lack comprehensive analytical information linked directly to business outcomes; they typically focus on past or present metrics without forecasting future impacts.

• The goal is to connect HR metrics with broader business strategies by transforming raw data into actionable insights that reflect performance effectively.

Prediction of Student Success

Objectives and Methodology

• The focus is on predicting student success by identifying variables that differentiate successful students, defined as those who complete the course satisfactorily and apply their learning in the workplace.

• A model was developed to classify students based on their similarity to a success profile, utilizing historical data and competency evaluation tests for comprehensive insights.

• Successful students scored higher in competencies such as analysis and knowledge application; an ethical-social values assessment also indicated that successful individuals exhibited certain predominant values.

Model Effectiveness

• The model demonstrated high accuracy in detecting successful students, improving detection rates from 7 out of 150 applicants to over double that number.

• A visual representation showed a comparison between random chance (red line) and the predictive model's effectiveness (green line), highlighting the model's practical utility.

Key Considerations for Implementation

Alignment with Business Strategy

• It is crucial to align projects with the overall business strategy, breaking down silos within human resources to ensure integration with broader organizational goals.

Commitment and Data Sensitivity

• Stakeholder commitment is essential; participants must be dedicated not only to project continuity but also to handling sensitive information responsibly.

Quality Over Quantity of Data

• Emphasizing quality data collection over speed is vital; having accurate data leads to more meaningful conclusions than rushing through data gathering processes.

Inventory of Available Data

• Creating an inventory of existing HR data can provide insights into potential projects, helping identify what information is available and its ownership.

Managing Data Quality Challenges

• Recognizing that many HR data points may be flawed or manipulated requires a disciplined approach towards collecting reliable information for effective decision-making.

Team Composition and Proactive Analysis

Multidisciplinary Teams

• Collaboration among psychologists, data scientists, and other professionals enriches project outcomes by integrating diverse perspectives and expertise.

Proactive vs. Reactive Approaches

• It's important to adopt proactive strategies rather than reactive ones when analyzing data related to employees' performance and well-being.

Pilot Projects for Testing Ideas

• Starting with small-scale pilot projects allows organizations to test concepts effectively before scaling them across different departments or roles.

8 Important Tips for Employee Retention

Overview of Key Points

• The speaker thanks the audience and invites questions, indicating a willingness to clarify any doubts regarding employee retention strategies.

• Acknowledgment of technical issues during the presentation; encourages attendees to submit their questions for further discussion.

Predicting Employee Turnover

• Discussion begins on how to predict employee turnover, emphasizing the importance of data collection as a foundational step.

• Initial data sources include company records such as salary history and employee demographics, which are crucial for building predictive models.

Training Predictive Models

• Explanation of how predictive algorithms classify employees based on their risk of leaving the company, integrating with existing systems for real-time alerts.

• Importance of understanding individual employee circumstances (e.g., salary stagnation over years) to inform retention strategies.

Economic Impact of Retention Strategies

• Emphasizes quantifying the financial benefits of retaining employees versus costs incurred from high turnover rates.

• Highlights efficiency in recruitment processes by using predictive models to focus on candidates likely to succeed within the organization.

Measuring Risk Factors

• Discusses methods companies can use to assess turnover risks, including monitoring performance declines and salary adjustments.

• Suggestion that direct communication with employees can provide insights into job satisfaction and potential reasons for leaving.

Data Requirements for Predictive Modeling

• Outlines necessary information for creating effective predictive models, including performance evaluations and training histories.

• Stresses the need for comprehensive data collection from various sources to identify high-potential employees effectively.

Analysis of Data and Its Impact on Diversity

Concerns About Discrimination in Data Analysis

• The discussion highlights concerns regarding data analysis potentially leading to discrimination in processes like selection and training, particularly affecting diversity.

• The speaker emphasizes the exclusion of sensitive demographic data (gender, nationality, age) from predictive models to avoid bias.

• It is noted that information that could discriminate against specific groups is not included in prediction models.

Standardization of Predictive Models

• The time required for organizations to standardize predictive model development varies based on leadership support and existing data culture.

• Companies with a strong data collection culture can implement changes more quickly than those without such practices.

• Without dedicated personnel pushing for change, the process may take months or even years.

Starting with Predictive Analytics

• Smaller companies (PYMEs) are encouraged to identify specific problems they face as a starting point for implementing analytics.

• Emphasis is placed on using available employee performance data to make informed decisions rather than pursuing overly complex analytics projects.

Importance of Employee Insights

• Understanding employee performance can help identify high performers and develop strategies to retain them within the company.

• Utilizing financial decision-making frameworks for human resources can enhance overall organizational effectiveness.

Training Needs for Data Analysis

• Discussion includes necessary training for technical aspects of data analysis, highlighting roles such as data scientists and psychologists with statistical knowledge.

• A combination of mathematical expertise and psychological insight is deemed ideal for effective analysis in HR contexts.

Addressing Legal Barriers

• The conversation touches upon potential legal barriers related to data protection laws but notes minimal encountered issues when collecting necessary data.

Data Utilization and Client Trust

Importance of Client Consent and Data Anonymity

• Emphasizes the necessity for clients to have confidence that their data can be used effectively for specific objectives.

• Highlights the importance of obtaining consent from evaluators when collecting personal data, ensuring agreements are in place for anonymous data handling.

Handling Qualitative and Quantitative Data

• Discusses the application of models using qualitative variables, particularly focusing on unstructured information like free text.

• Notes that while structured information is easier to work with, processing natural language or free text is also feasible through various methods.

Processing Intangible Information

• Introduces Natural Language Processing (NLP) as a growing method for analyzing textual data by converting it into numerical vectors.

• Addresses challenges in gathering intangible information, which includes aspects like innovation capacity that are not easily quantifiable.

Key Factors for Successful Implementation

• Suggests that having a solid team dedicated to pushing initiatives forward is crucial for success in any organizational model.

• Stresses the need to allocate sufficient time for data collection, advocating a gradual approach rather than rushing through processes.

Engaging Stakeholders and Final Thoughts

• Underlines the importance of involving stakeholders from the beginning to ensure project viability and relevance.

• Concludes with an encouragement to explore valuable personnel data within organizations, emphasizing that people are central to organizational success.