How Panasonic is Improving Quality of Hire and Candidate Experience with People Analytics
Panasonic Corporation Operational Excellence Recruit & Career Create Center
- Takashi Sakamoto: Manager, Planning Department / Strategic Planning Division
- Shoto Miyatani: Planning Department / Strategic Planning Division
- Masahiro Ishiguro: Recruiting Department / Recruiting Division
Situation Before Collaborating with Panalyt
— Please tell us about the role and mission of the Recruit & Career Create Center (RCC).
Miyatani: RCC provides shared services for the Recruitment division and career support to each operating company within the Panasonic Group. We propose and support various measures to improve the Employee Experience (EX), offering a wide range of services, from improving recruitment processes at the time of hiring, to helping employees intentionally navigate their careers after joining the company.
Sakamoto: We launched the People Analytics Lab in the fiscal year 2020. Our goal was to create a system to quantitatively visualize the performance of people and organizations with data, empowering Human Resources to directly contribute to the management and quantify the business impact of their people decisions.
Panasonic is a world-renowned industry leader in manufacturing, but behind the scenes, human resources development has been “invisible” – it has been difficult to measure the effectiveness of human resources development efforts and initiatives.
In the past, talent development was centered on an approach based on KKD (Japanglish; meaning gut-feeling, experience, and courage) but with the increasing difficulty of acquiring and training quality talent, it is necessary to identify means to support, enable and develop talent in a consistent manner. Panasonic is trying to break away from KKD dependence and instead utilize people data to improve make informed talent decisions in a consistent, repeatable manner.
—What was the initial situation towards the realization of that mission?
Miyatani: RCC divides the recruitment flow into five stages ( Awareness, Empathy, Attraction (~Offered), Hire, Assignment / Development), and aims to improve the Employee Experience( EX, Candidate + Post-Hire Experience) at each stage. Initially, I wanted to utilize data collected across the entire recruitment flow to grasp the current situation, identify issues, and verify the effects of decisions made, but I was frustrated because I couldn’t easily analyze the data and find a chance for a good initiative.
Ishiguro: I felt the need to process data more efficiently and rationally when formulating various measures related to recruitment. It was difficult to track KPIs for each flow, and in addition, we were not sure if we were tracking the right KPIs. We could have a rough idea of issues/ challenges affecting the employee experience during the recruitment flow, but we couldn’t identify where and what the core issues were and how to solve them.
— Why do you think the initial results were not able to match expectations?
Miyatani: There were many requests to analyze all available datasets and identify insights, i.e. boil the ocean, without clarifying the issues and hypotheses to be verified – which kept us from performing the data analysis effectively to support business goals.
Sakamoto: We were trying to analyze our existing people data in isolation from the business, and thinking about the insights and initiatives we could propose by just looking at the data, but this approach did not lead to results.
After 1 year of trial and error since establishing the People Analytics unit, we realized we couldn’t progress further without consulting external experts like Panalyt.
Achievements from the collaboration with Panalyt?
— Why did you choose Panalyt to be your strategic partner for People Analytics in 2021?
Sakamoto: Panalyt’s vision and capability to deliver “the financial statements for human resources” set it apart from other vendors supporting personnel data analysis. I believe that finance/ business and HR divisions should contribute to management on an equal footing, but for that to happen HR also needs to make discussions and divisions using numbers.
The lack of quantitative, data-driven decision-making in the people area has resulted in the CHRO being excluded from the executive board. I believe that Panalyt is the best partner to pursue talent enablement and empowerment with data technology, just as Panasonic has empowered the industry with engineering technology.
Miyatani: While several RCC team members, including me with a data science background, had experience in the area of “personnel data analysis”, I felt a big barrier to promoting the utilization of people data and making it relevant to the business.
Due to a change in group structure, RCC had to make a fundamental shift from a conventional cost center to a profit center that independently provides value to each operating company across the Panasonic Group.
Panalyt provided us with not only the ability to view and interrogate our integrated raw people data across multiple systems, but their team’s past experience with People Analytics helped guide us in deriving actionable insights from strategically built dashboards and visualization, and ensure RCC is able to deliver insights that match the needs of the business and even go a step further with predictive insights to mitigate future risks.
— It’s been half year since starting your collaboration with Panalyt. How did you proceed?
Miyatani: Using Panalyt’s “Small Start, Quick Win” approach, we proceeded in the following four steps:
(1) Identify the stages and issues to focus on initially amongst the five recruitment flows( Awareness, Empathy, Attraction (~Offered), Hire, Assignment / Development) that are RCC’s responsibility,
(2) make hypotheses from the data and information on hand,
(3) collect and analyze the data necessary for testing the hypotheses,
(4) Make proposals to the site based on the suggestions obtained from the analysis results.
We were able to run this cycle in half a year, and the proposal was adopted by the site.
— (1) How did you specifically identify the issues?
Miyatani: We decided to focus on the impact of recruiter involvement on recruitment outcomes (first-year engagement/performance/attrition). Hypothetically the recruiter management seemed to be the easiest solution to handle, compared to other ones. Revisiting the interview process, training contents, assignment process would require more internal stakeholders with much more time and effort to proceed.
The quality of the recruiter’s selection, activity, and performance was found to be fluctuating, and these processes/decisions had historically been driven by gut-feel and intuition, rather than being operated in a data-driven way. If the KSF and best practices were identified (and consolidated in a scalable manner), we expected it would drive a significant impact on the total Employee Experience (Candidate + Post-Hire Experience).
— (2) What hypotheses did you derive from the data and information initially available to you?
Miyatani: I was able to get further suggestions by adding a few questions to our monthly pulse survey.
The results from these surveys allowed us to start with the initial hypothesis of “the involvement of recruiters in the decision-making of potential hires joining the company and the motivation after joining the company cannot be ignored.”.
I then organized in-depth interviews with the recruiters to refine my hypothesis. As a result, it became clear that many of the factors that create variations in the recruiter contact experience are in the recruiter selection process and team structure.
—(3) How did you test the above hypothesis with your integrated people data using Panalyt?
Miyatani: Before we started our collaboration with Panalyt, we had initially identified more than 3,000 data items related to recruitment activities. The Panalyt team’s expertise in People Analytics then helped us narrow this down to 15 essential data points by identifying issues and refining hypotheses. We tested most of the above hypotheses by using a technique called “recruitment outcome analysis” for these 15 items. In particular, it was interesting that the impact of the COVID-19 pandemic on online recruitment activities was noticeable.
— Was there a hypothesis that was disproved?
Miyatani: We found that the “number of candidates per recruiter” did not have as significant an impact as we had expected from the recruiter in-depth interviews.
We originally assumed that if the number of students in charge per recruiter increases, it increases the burden and the level of dedicated support for each student is compromised. But the result wasn’t that. I assume we overlooked the confounding factors; e.g. the capacity of recruiting activities differs by the recruiter*.
*Note that the recruiters are not the defined role at Panasonic (and most Japanese companies). HR divisions ask the line managers for the “volunteers” (with a certain shortlist) every year, and the “voluntary recruiters” need to spend spare time for recruiting activities outside their regular works.
—(4) What proposals did you make to the business based on the results verified above?
Miyatani: I made two major proposals. First, we defined multiple KPIs based on the above hypothesis verification results, and worked with the Panalyt team to create a “dashboard” template that allows us to visualize the progress and share it with the entire team.
–How did you feel about this proposal from the perspective of the business?
Ishiguro: I felt the “dashboard” would be well-received by our stakeholders and positively impact the business.
In the past, recruiter activities were not linked to the progress of recruitment and public relations activities, and we had no way but to accept what the recruiters report once or twice a month without questioning, so we could not identify the real issues.
Panalyt facilitates our team with the tools to have data-informed discussions with recruiters in a near-real-time, and use data to align on potential issues and solutions. I can imagine tons of use cases; monitoring the drop rate of the candidates by funnel stage, by source, by a recruiter; calculating the necessary volume of candidates from additional campaigns; reallocating the students in charge among the recruiters, to name a few.
— What was the other proposal?
Miyatani: Since it was found that the experience of activities as a recruiter and the management of leaders affect the student candidate experience, we proposed to create a program to create content that promotes the improvement of recruiter team performance and explicit knowledge of know-how with the best practices obtained from the analysis and listening efforts.
— How did you feel about the second proposal?
Ishiguro: We always had a gut feeling that we needed to spend time and effort on creating such content and programs, but we couldn’t back this up with facts and put it into practice with confidence. This year, we were able to validate the need for better training content with data, so we were confident about our decision.
–What were the key learnings for your team through the analytical process?
Miyatani: In addition to the suggestions and outputs (suggestions) obtained from data analysis and hypothesis verification, we were able to learn the process of further evolving the hypothesis by tweaking the hypothesis verification cycle a little and learning from it.
In particular, I think that the game was almost decided at the stage (①②) before collecting the data.
In the past, instead of “starting with issues and hypotheses,” we were vaguely trapped in the image that “collecting data seems to be difficult.” It was not possible to collect the data after identifying the data and information necessary for hypothesis testing, and the eventual decision(s) was based on an “answer match” of experience and intuition.
Fortunately, Panalyt had extracted and integrated our people data to a large extent, so there was almost no need for any manual effort to collect, analyze, and suggest the data after making a hypothesis.
— Would you have been able to achieve the same results for this project without Panalyt?
Miyatani: Along with enabling us with the technological capability to derive holistic employee insights at scale in a timely manner, the Panalyt team provided us with practical approaches and knowledge on how to ensure our People Analytics efforts provide value to and gain traction with our end-users in the business.
A typical example is an issues/hypothesis-driven data definition/collection/analysis approach. In data analysis, it is easy to fall into the trap where the introduction of tools and visualizations become the purpose, but Panalyt’s approach of “Small start, quick win” in this initiative helped us create a process that allows us to easily and quickly run the hypothesis verification cycle by ourselves, and even enable our internal stakeholders from the business to do the same with their populations.
Access to people data and insights in a consistent manner accessible to multiple stakeholders has allowed us to use data to align on the issues we should focus on, and has helped us overcome feelings of resistance and misunderstanding around people decisions, for which the reasons were previously “invisible” to the rest of the organization.
–Please give us a final comment on the achievements of the partnership with Panalyt so far and future developments.
Sakamoto: Over the last six months, the way we see and relate to the people involved in the company has changed significantly.
The RCC group has been reborn as a voluntary value-providing group, and it was recently announced at the RCC General Assembly that “data utilization will be the pillar of 1-1s”. The number of consultations around “Can we verify these issues and hypotheses with data” is increasing, and we can already feel the positive response to our new approach from the business.
Miyatani: While tracking the effectiveness of the content proposed this time, I would like to pick up the needs of new business units/ locations and expand the range of activities.
In addition, as one of the individual themes, we would like to further understand how communication affects performance and engagement through analyzing the metadata (no content) of how employees collaborate with each other, for which we look forward to the insights derived using Panalyt’s Relational Analytics / Organizational Network Analysis (ONA) capabilities.