Embarking on a journey into People Analytics, HR leaders are confronted with a pivotal decision — whether to invest resources in building a customised platform in-house or opt for a pre-packaged, off-the-shelf People Analytics solution.
This decision will shape various aspects of your organisation’s People Analytics journey, influencing costs, user-friendliness, scalability, speed, and even the comprehensive ability to meet business objectives in the first place. The path you choose will impact the efficiency and effectiveness of your entire People Analytics venture.
Criteria for Decision-Making
Here are the factors to consider when making this decision.
1. Tech Resources
The right technology stack and skilled personnel are essential to derive meaningful insights from HR data. Here are the key components and capabilities needed when building a people analytics function in-house:
- Data Infrastructure:
- Data Warehouse: A centralised repository for storing and managing HR data. Common choices include Amazon Redshift, Google BigQuery, or Snowflake.
- ETL (Extract, Transform, Load) Tools: Tools for extracting data from various sources, transforming it into a usable format, and loading it into the data warehouse. Examples include Apache Airflow, Talend, or Informatica.
- HR Information System (HRIS):
- An HRIS system is crucial for maintaining employee records, managing HR processes, and ensuring data accuracy. Popular HRIS systems include Workday, SAP SuccessFactors, or Oracle HCM.
- Analytics Tools:
- Data Visualization Tools: Tools like Tableau, Power BI, or Looker for creating interactive and insightful visualisations to communicate HR metrics and trends effectively.
- Statistical Analysis Tools: R or Python with libraries like Pandas and NumPy for in-depth statistical analysis and modelling.
- Machine Learning (ML) Capabilities:
- Building predictive models for workforce planning, attrition prediction, and other HR-related forecasts may require machine learning expertise. Python, R, or specialised platforms like TensorFlow or scikit-learn can be employed.
- Databases and Storage:
- Adequate storage solutions to handle large datasets efficiently. This could include cloud-based storage solutions such as Amazon S3, Google Cloud Storage, or Azure Blob Storage.
- Security Measures:
- Implement robust security protocols to protect sensitive employee data. Encryption, access controls, and regular security audits are essential.
- Integration Capabilities:
- Ensure the ability to integrate with various HR systems, applications, and external data sources to gather comprehensive insights. API capabilities and middleware solutions might be necessary.
- Data Governance:
- Establish data governance policies to ensure data quality, accuracy, and compliance with data privacy regulations such as GDPR or HIPAA.
- Scalability:
- Design the infrastructure with scalability in mind, considering the potential growth of data and the need to accommodate more advanced analytics capabilities in the future.
- Skilled Personnel:
- A team with skills in data science, statistics, data engineering, and domain knowledge in HR. This may include data scientists, data engineers, analysts, and HR professionals who understand the business context.
- Continuous Training and Development:
- Invest in ongoing training to keep the team updated on the latest technologies, methodologies, and industry best practices.
Identify what you want to achieve out of People Analytics in both the medium to long term and what resources and tech infrastructure will it take to achieve those. HRs usually struggle to get the tech resources, so honestly assess whether you will have the support from the C-suite to get the tech resources and capabilities needed to build a people analytics function in-house.
When you buy a people analytics solution from an external vendor, they typically manage all aspects of its implementation and maintenance.
2. Costs
In-house solutions may seem cost-effective initially, especially when compared to buying software that has upfront costs. But purchasing software will have no unforeseen costs later, whereas building adds up hidden expenses – usually costing much more than buying and using external software.
A McKinsey survey found that large IT projects exceed their budget nearly half (45%) of the time, and tend to provide 56% less value than initially anticipated.
Costs for building people analytics solutions inhouse come in the form of many things, including
- Software costs such as:
- Business Intelligence software licences
- Data warehouse software licences
- Data Lake software licences
- Data integration tools licences
- Maintenance costs: When you build a People AnalyticsPA system internally, it needs constant attention—fixing bugs, making updates, and improving it.
- Salaries of following professionals for the entire duration of building it, which may be in months or even years.
| Role | Skills/Expertise |
| Data Scientists/Analysts | Statistical analysis, machine learning, and data modelling skills. |
| Data Engineers | Expertise in data integration, cleaning, and database management. |
| Database Administrators | Skills in setting up and optimising databases or data warehouses. |
| Data Visualisation Experts | Proficiency in tools like Tableau, Power BI, or programming for custom visualisations. |
| Security Experts | Knowledge of data security and compliance measures. |
| Project Managers | Leadership and project management skills to oversee the entire project. |
| Training and Communication Specialists | Skills to facilitate user training and communication. |
| Domain Experts (HR) | Understanding of HR processes and metrics. |
Allocating internal resources to people analytics developments means diverting them from other business functions. For example, Tim Cook maintained that Apple should only handle the main technologies in their products and be in markets where they can make a big difference. Evaluate if dedicating your team’s efforts to creating an in-house people analytics tool is justified as a long-term strategic investment.
Time to Value
Just like you want to reduce costs, you also want to make sure it doesn’t take too long to create and use the system. This way, you won’t miss out on the potential benefits during that time.
Building a platform that has all the data integrations in place and works effectively and accurately without bugs takes months and sometimes years, creating delays in generating reports. Think about all the things you could have achieved in that time, and what your competitors may have achieved, potentially causing you to lose your competitive edge.
However, external solutions offer seamless integration with various data sources via APIs in real-time, allowing you to consolidate and analyse data from different systems such as Applicant Tracking System (ATS) and Human Resources Information System (HRIS) quickly and easily.
With the buying approach, you can create hundreds of dashboards, metrics and visualisations in weeks instead of months or years, allowing you to promptly identify and address issues as they arise.
Compliance and security
In addition to getting started with people analytics as quickly as you can, there’s also a need to ensure compliance and security. When developing an in-house people analytics tool, it is important to consult legal and compliance experts to ensure adherence to data security and privacy standards.
Depending on data type, geography, and industry, compliance might be necessary for regulations like GDPR (General Data Protection Regulation) in the EU, CCPA (California Consumer Privacy Act), ISO/IEC 27701, AI Ethics and Bias Standards, etc.
You must ensure that you can protect employee data, because, in the event of a data breach, organisations may face financial penalties, legal actions, and suspension of data processing activities. The associated reputational damage and loss of trust can even adversely impact HR goals of attracting and retaining talent.
For example, in 2023, Yum! Brands, the parent company of KFC, Taco Bell, and Pizza Hut, faced a cyber attack. While not a breach of data from people analytics software, the incident impacted employee data, prompting financial repercussions including the closure of almost 300 UK locations and increased security costs.
Given the significant repercussions and ramifications of compliance failures and data breaches, using external software may offer a safer option. External vendors, whose entire business centres around data, enforce strict compliance measures, actively monitor policy standards and possess expertise in data management.
Vendor tools also make it easy for different people to see different data due to their customisable permissions. Doing this with in-house software built over PowerBI and Excel can be tricky since you’d need data experts and data engineers to set it up and make it work. This means that the employees who contribute to setting it up might have access to the sensitive data, raising potential security and employee concerns.
Data Accessibility and Insights:
Although security and privacy are foundational and cannot be overlooked, the true utility and impact of people analytics arise from what data can you access, what insights it generates, and who you can share them with.
However, there are significant differences in data that can be accessed between a purchased software and an in-house solution, mainly concerning: historical data, benchmarked information, tailored insights, and sharing capabilities.
- Historic Data
People analytics software allows for the seamless storage, retrieval, and analysis of extensive historical workforce data, providing a comprehensive view of trends and patterns over time.
However in-house solutions often lack automated analytics features and require significant manual effort to work with and draw patterns from large volumes of historic data.
- Benchmarking data
Along with historical data of your company, you also ideally want access to industry data so that you can compare your workforce-related metrics such as turnover rates, productivity, and talent acquisition efficiency with organisations similar to yours and identify areas of improvement.
External people analytics software often provides benchmarking data and features, such as
- include pre-built industry benchmarks,
- facilitate anonymised data sharing among participating organisations to enhance benchmarking accuracy,
- allow customisation based on criteria like company size or geography,
- facilitate peer comparisons, and
- provide trend analysis and predictive benchmarking to help organisations track changes over time and forecast future trends.
However, when developing the tool internally, your access is limited to industry reports, which might lack accuracy, be outdated, and lack real-time information.
Additionally, incorporating data from external industry reports into your in-house software requires manual effort, consuming a considerable amount of time. This process also needs to be repeated periodically with fresh yearly data, introducing the need for constant updates and revisions.
- Tailored reports
While gaining access to data, whether historical or benchmarked, is one aspect, it’s equally crucial to provide access to different levels of data and insights to different professionals, depending on their role in the company.
For example:
- Leaders like to see high-level dashboards and executive summaries. This can include metrics related to overall employee engagement, talent retention, and key performance indicators aligned with broader business objectives.
- HR Business Partners (HRBPs) might need more granular insights into workforce planning, diversity, and talent development.
- Managers might require information on team performance, individual contributions, and employee engagement within their respective teams.
A pre-built people analytics software provides distinct views designed for leaders, managers, and other stakeholders. This customization ensures that each user group receives the most relevant and actionable information.
However, these tailored reports are difficult to be replicated when building people analytics in-house, without spending even more time, resources, and finances than you already have.
- Sharing and collaboration
Similar to the absence of customised reports, constructing an in-house People Analytics solution is often criticised for its deficiency in easily shareable settings without compromising security.
Those responsible for generating reports often struggle to readily share them with critical stakeholders due to a lack of easily shareable settings that also maintain security standards. It restricts insights to specific users or departments and hinders the impact and results that you can achieve with people analytics.
On the other hand, buying pre-built people analytics software promotes seamless sharing and collaboration, effectively utilising available insights across the organisation.
Ease of use
To build a user-friendly solution in-house, you need dedicated personnel to handle it. If the person who knows how to use the tool is on holiday or leaves the company, there can be a significant knowledge gap and potential disruptions in the system’s operation.
But buying a PA tool from a vendor is different. Given the simplicity of the external tool, usable by most people, and often the presence of an expert team and account manager, it ensures everything runs smoothly even amid team changes or holidays.
Scalability
Beyond being able to manage and use the software easily from the start, you should also think about what you might need from it as your organisation and People Analytics function grows. If initial insights and projects are successful, business leaders will want the data to answer more nuanced questions, drill down to the department or employee level, run multiple projects simultaneously, or simply increase data points and users.
This is when the in-house system becomes no longer sustainable as you may struggle to answer questions accurately, face delays, grapple with data inconsistencies, and have difficulty managing increased demand. However, scaling an in-house system requires additional work, resources, time, and costs that are difficult to predict accurately, and that usually hinder the momentum achieved in people analytics so far.
On the contrary, external software easily accommodates increased data and demands as it is purpose-built, designed by experts, and already scaled for various clients, ensuring seamless expansion without additional in-house adjustments.
Buy, not build
Based on the above points, we recommend you buy your people analytics software.
Building it in-house may be suitable for smaller-scale applications. But for handling intricate people analytics queries, managing multiple projects, ensuring agility and scalability, and aligning all stakeholders, purchasing an external solution is generally more efficient, cost-effective, and impactful.
How to build people analytics solutions in-house
However, if you decide that building an in-house people analytics solution aligns with your current stage, business objectives, and size, below is a general outline of the process.
| Description | Tech Resources Needed | Timeframe | |
| Define Objectives and Requirements | Clearly outline the goals of the people analytics solution. Determine the types of insights and metrics the organisation wants to derive. | – | A few weeks to a couple of months |
| Data Discovery and Integration | Identify and integrate relevant data sources such as HR systems, performance reviews, employee surveys, etc. | Data integration tools, understanding of APIs, knowledge of HR systems | A few months, depending on the number and complexity of data sources |
| Data Cleaning and Transformation | Cleanse and transform the data to ensure consistency and accuracy. This may involve handling missing data, standardising formats, etc. | Data cleaning tools, scripting languages (e.g., Python, SQL) | Several weeks to a few months, depending on data quality |
| Database and Storage Setup | Establish a database or data warehouse to store the cleaned and transformed data securely. | Database management skills, knowledge of cloud platforms | Several weeks to set up, with ongoing optimization |
| Analytics and Modeling | Apply statistical methods, machine learning, or other analytical techniques to derive insights. This step involves creating models for predictive analytics or identifying patterns in the data. | Data science expertise, statistical modelling skills, machine learning knowledge | Several months, depending on the complexity of the analytics |
| Visualisation and Reporting | Develop dashboards and reporting tools for easy visualisation of insights. Tools like Tableau, Power BI, or custom web-based solutions may be used. | Data visualisation tools, and web development skills if custom solutions are used | Several weeks to a few months |
| Security and Compliance | Implement security measures to ensure data privacy and compliance with regulations, especially when dealing with sensitive HR data. | Security expertise, compliance knowledge | Ongoing, integrated throughout the project |
| User Training and Adoption | Train end-users and stakeholders on how to use the analytics platform effectively. Ensure user adoption through change management strategies. | Training and communication skills | Several weeks |
| Maintenance and Iteration | Regularly maintain and update the system based on changing business needs. Iterate on analytics models and visualisations as necessary. | Monitoring tools, ongoing data management | Ongoing, with periodic updates and improvements |
Building a people analytics solution can indeed be a substantial IT project, and the specific details of the project can vary based on the organisation’s requirements and goals.
