Categories
Creativity Innovation User Experience

How bots helped reduce the volume of Microsoft HR tickets by ~35% and response time by ~25%

Legacy HR tools with disjointed interfaces and functionality delivering a less-than-optimal experience was not cost effective

At Microsoft, employees reach out to Global HR Operations (GHRO) to get policy clarifications, tool help, process clarification to avail employee benefits etc. In Americas and Europe, the volume of such requests can be up to in excess of ~50K per annum. Today, depending on the complexity of the issue, it can take 2 to 48 hours to get an answer from GHRO. Often, employees do not provide complete information at the beginning, and it takes more effort for GHRO to gather necessary information and to provide relevant answers.   GHRO has dedicated teams with necessary subject matter expertise across regions to provide timely answers. Once the employee logs an issue, a ticket is created in the CRM which is attended by Operation Advisors (OAs) who connect with the employee to gather necessary information, extract suitable responses from several sources and provide the clarification. At times, this can be manually intensive.

Analysis of employee ticket volume by category reveals that at least 40% relate to OAs providing details that is already available in various sources such as knowledge bank documents, CRM Database, HR Web, SAP, third party tools etc., which otherwise may not be readily and easily accessible to the employees in a consolidated format.

What was the anticipated consequence

Legacy HR tools with disjointed interfaces and functionality, delivering a less-than-optimal experience that was not cost effective. Lack of realignment of HR strategies to corporate strategic and cultural changes – mobile first, cloud first.

What was my role

I saw an opportunity to deliver business impact by improving productivity of HR OAs and employee experience. I drove the decision making process including development of business case, solution architecture, two-phase approach (pilot and rollout).

What action was taken

This challenge lends an opportunity to automate the process with the help of a virtual assistant such as bot and enhance employee experience with real time response and query resolution, which will significantly reduce the current lead time of up to 48 hours to get an answer. In case the bot is unable to provide an answer, it can help consolidate all the necessary information captured as part of conversation and provide the same to the OA, which will significantly improve OA’s productivity. As more and more questions are answered by the bot, it learns, and the answer relevance / accuracy increases. All these will contribute to improvement in NSAT, CSAT, ABDR, Bottom Box % and RTI (Response Time Indicator) measures across the board for GHRO by at least ~25 – 50%.Before bots can answer employee’s questions, they must learn about the data sources and be context aware to provide employee friendly answers. Bots interface with cognitive services such as spelling check, language understanding intelligent service (LUIS) that must be trained to work for a specific domain like human resources.

To address this need, I proposed a two-stage approach –

STAGE 1

Enable GHRO operation advisors to train the bot and help them realize productivity benefits. Integrate bot with knowledge sources, embed the bot within CRM so that when the operations advisor opens a ticket, the bot can read the question, determine the most relevant source for information and surface the answer. The operations advisor can then either use the answer as-is or edit to make it more relevant. In the process, the bot can continuously learn to make the answers more employee friendly. This will set the stage for rolling the bot to be directly used by employees

STAGE 2

Rollout to employees by region to enhance employee experience. Once the bot is ready to be used directly by employees, the same will be integrated with slip stream (the new HR online portal) using a federated bot-of-bot architecture.

What was the decision making process

The decision making process involved –

  • Building a business case
  • Developing the solution architecture
  • Implementing a two-phase approach – pilot & rollout

What was the outcome

Pilot was rolled out successfully to the UK geography. Reduced volume of tickets by ~35% and response time by ~25%. This data was measured over a 6 month time period.

What has most value

Processes assist in management tasks and standardization and free employees up to focus on more creative endeavors.  Determining to implement any process at all is the first step.  After that decision is made, the next step is to identify the optimal amount of process to implement. – Martin L Abbott & Michael T Fisher, co-authors of the book The Art of Scalability.

Two methods for determining the optimal amount of process are 1) migrating from small to large through periodic changes 2) letting the team decide on the right amount. Maintenance of processes is critical to ensure organizations do not outgrow their processes.

Design Thinking self-service solutions like bots integrated with cloud help improve productivity and user experience. Use the cloud to keep information safe and deliver service anywhere on any device. This has project been showcase of how Microsoft runs its own business powered by Azure at scale.

How do you determine the optimal amount of process in your organization ?

How do you measure its effectiveness ?

Do share in the comments!

Like this article ? How about giving it a like and share ? Thank you!

By Ramesh Chandak

Principal Content Developer & Technical Program Manager

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s