SAP

Author

Sid Bhattacharya

Sid Bhattacharya

Sid Bhattacharya is Vice President, Global Presales and Head of Demo Product Management team with SAP SuccessFactors where he works with solution experience, product management, operations, and engineering groups to define the long-term demo roadmap. He is a passionate technologist and has interests in analytics, machine learning and is focused on bringing the best-run demo experience for SAP’s customers and prospects.

Keep in touch

Subscribe for the latest news, updates, tips and more delivered right to your inbox.

Subscribe to updates

Category

Insights

Connect with us

Experience Management (XM) is the process of monitoring every interaction people experience with a company in order to spot opportunities for improvement. In this blog, we will use XM techniques to analyze a Best-Run Bike company use-case and come up with a recommendation model for flight risk prevention and top talent retention.

experience management

The Best-Run Bikes company manufactures bikes and sells them online and through a few physical stores. Best-Run Bikes is looking at customer service attrition and noticing an increase in turnover in certain locations and in certain job roles. They are turning to Employee XM techniques to blend the operational and experience data to discover how to attract, develop and retain employees.

XM can help you understand:

  • Correlations between employee experience and customer experience
  • Factors that drive employee engagement
  • Evaluation of learning platforms with sales effectiveness
  • and more…

What is currently unknown is:

  • Are employees who are leaving critical to the company (are they top performers)?
  • Do the employees who are leaving solve critical support problems and are they experts?
  • Are ex-employees joining competitors? If yes, why?
  • Have exiting employees provided feedback on why they are leaving?
  • Were there any early signs based on employee surveys that indicate a possible employee turnover?
  • Is there any feedback provided in exit interviews and surveys that would help create a strategy to retain critical employees in the future?
  • What are the short term and long term strategies to retain top performers?

We will start to look at operational and experience data across various silos: HR, Finance, CRM as well as external sources. The purpose of the analysis to identify “what” is the employee churn problem and “why” is there a problem and finally coming up with a recommendation model.

XM Data Analysis

Employee Experience chart
Fig: Performance metrics indicate “Tech Experts” with “Outstanding or Successful” ratings have been leaving the company in the past 6 months.

The above view shows that the top performers are leaving the company and impacting the overall productivity of the company (“Tech Experts” with “Outstanding or Successful” ratings have been leaving the company in the past 6 months). This analysis was done by combining HR data (job profiles, performance data, learning data) with CRM service data (tickets).

Further analysis of the employee turnover data shows that the top performers also worked on the most complex tickets which means that losing them will lead to losing highly specialized.

Analysis of employee turnover
Fig: Service ticket metrics indicate “Tech Experts” work on most of the tickets and solve the most complex tickets across other users.

These analyses tell the “what” side of the employee churn story. To understand the “why” we need to look at the employee experiences captured at various moments of the employee.

Fig: Exit Interview survey data

Employee data indicates that most of the employees who left the company in the past 6 months have been in the “Tech Expert” role with “Austin and Cleveland” the locations with the most turnover. The right side also indicates that the #1 reason for employees leaving is “Voluntary resignation to competitor”

Employee Surveys

Employee survey analysis
Fig: Employee survey data

Employee surveys for high performers in “Tech expert” role indicate the areas employees have ranked higher than others.

Further analyzing the anonymized employee surveys, the topics that should serve as early detectors; those top performers were generally unhappy with their managers and the overall approval process in the company wasn’t adequate, leading to the experience gap:

“I need to get approval to do anything out of the standard. If I was not competing so heavily with other manufacturers and dealers, this would not be an issue. When a prospect asks me to match or beat a competitor, it takes days and weeks to respond.” 

“No, I always have to seek approval from my manager and they go to their manager.” 

Next, generating a natural language bi-gram frequency display in SAP Analytics Cloud provides a general view of verb terms that are captured in the survey comments.

negative employee sentiment chart
Fig: Bi-gram frequency distribution of negative sentiments in employee survey comments
Radar chart
Fig: Top 5 bi-gram link radar chart

“You know what needs to be done to correct a customer issue, but you have to go through too many channels to get the job done.”

Using a neural network based word similarity model, we can not only see the frequency of the words but also see a similarity between key terms.

In the view shown below, negative (in red) and positive (green) terms are shown in cells that are most similar to the header words. Based on this, it shows that across the column for “Customer” you can see “issues” listed as a negative; similarly “Communication” column shows that the words “zero”, “seriously” and “rude” are used in the similar context all of these indicating negative sentiments.

Using a word2vec similarity model and displaying as a sentiment grid/table in SAP Analytics cloud, it’s easy to see what employees think about the most when asked about “compensation/pay”, their “manager”, “customer”, “vacation” etc. It shows the correlation between the employee sentiments with what matters the most to them.

SAP analytics cloud allows analysts to perform analysis like these on the fly and blend data from multiple data sources. They can also blend external data like salary data to validate if the employee churn related to, voluntary resignation to competitors, is related to compensation.

Market salary data
Fig: Market salary data

Market salary data for similar positions and locations indicate the market median salary is much higher than the average median salary for a “Tech expert”.

Conclusion

By analyzing two-quarters of attrition data we were able to understand that:

  • Top employees are leaving the company, tech experts with outstanding performance ratings.
  • The trend is more prevalent in cities like Austin, Cleveland, and Philadelphia.
  • There is a big productivity impact as the top performers also worked on the most # of tickets and the most complex ones.
  • These employees have scored high on customer satisfaction based on their work.
  • The top performers have called out some common areas that need to be addressed – back-office processes like approvals, better tech. support with mobile apps and others.
  • Market salary analysis reveals that competitors are providing better compensation than what is currently offered at Best Run Bikes.

Blending the employee experience data with operational data allowed for:

  • Enriching the data set by bringing in influencing experience metrics (survey scores, feedback, and comments).
  • Uncovered problems that were not known before, through text classification.

SAP Analytics Cloud earns a top ranking from BARC

See how SAP Analytics Cloud performed in the world’s largest survey of Business Intelligence software users.