Case Study 4

IMB 851

AMBER BY INFEEDO: THE CEO'S VIRTUAL ASSISTANT REVOLUTIONIZING

EMPLOYEE ENGAGEMENT

DEBOLINA DUTTA

Debolina Dutta, Professor of Practice, OB&HRM, prepared this case for classroom discussion. inFeedo cooperated and provided information to the Indian Institute of Management Bangalore in connection with the preparation of this case and it was reviewed and approved before publication by a company designate. No funding was sought or received from inFeedo for the development of this case. This case was also developed from available and permitted sources of information. This case is not intended to serve as an endorsement, source of primary data, or to show effective or inefficient handling of decision or business processes. The support and help provided by Tanmaya Jain and Ajay Ruhela of inFeedo is gratefully acknowledged.

Copyright © 2021 by the Indian Institute of Management Bangalore. No part of the publication may be reproduced or transmitted in any form or by any means – electronic, mechanical, photocopying, recording, or otherwise (including internet) – without the permission of Indian Institute of Management Bangalore.

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

Tanmaya Jain, the 24-year-old CEO of inFeedo stared meditatively into space while his key team members argued

vociferously in the virtual meeting room in Gurgaon. The quarterly meeting, which included all the organization's

business and function leaders, typically focused on product evolution and preparation of a road map. It stirred a

healthy debate on new or unanticipated customer needs. The COVID-19 scenario, which had forced many inFeedo

clients and their employees to work from home, changed the employee engagement context.

Interrupting the debate, Jain argued for incorporating his idea into the new product avatar:

Amber's (our chatbot, which enables employee voice) core vision was to provide employees with a platform to solve their workplace problems. Today, Covid pandemic has changed the paradigm of employee engagement. The top of mind concerns for employees revolves around job security, their mental and physical well-being. Eventually, this has a rub-off effect on their productivity, motivation, engagement, etc.……all things that the organization is also interested in safeguarding. I think this pandemic is going to change the ways of working for organizations and their employees forever. Therefore, in her new avatar, I feel Amber should dig deeper and our AI platform should be able to present a revised dashboard and analytics that expand beyond our current measures of employee engagement.

While the Head of Product Development, Varun Puri, spoke about the challenges he anticipated; Shyamantak Das (Director-Customer Service) felt that the existing clients' current expectations on the product provided enough challenges, which needed to be addressed if renewals were to be expected. Additionally, Das felt that the inFeedo team needed to increase the training and education of HR teams and line managers on ethically managing the insights provided by their platform. Ajay Ruhela (AVP-People Sciences) felt that the team was already overburdened with managing and making sense of the enormous amount of employee data that was being generated. Ruhela argued that employee well-being as a domain was comprehensive and could encompass multiple dimensions. While the arguments continued to rage, Bhavan Kochhar, VP, Sales of inFeedo, privately messaged Jain on the chat section, "The use of AI and Chatbots in the workplace is hotly debated on the ethics of the phenomenon, the ethics of data management and privacy. Have you thought through the legal and ethical dimensions of weaving in the COVID-19 sentiment into our analytics? Are we ready for this yet?"

MEASURING EMPLOYEE ENGAGEMENT

The world over, organizations were deeply concerned about enhancing employee engagement. A Forbes study1

estimated employee engagement to be a $600 billion problem, with an estimated $11 billion loss annually in the United States alone due to employee turnover.2 The commonly adopted practice within the industry to measure employee engagement was to conduct annual/bi-annual employee satisfaction and engagement surveys. To enhance participation and encourage employees to speak up frankly and fearlessly, most organizations planned these surveys anonymously with the help of independent third-party vendors, who collected the data and disseminated the aggregated reports. Many big and boutique firms offered this service with differing survey instruments (Exhibit 1). The popular engagement survey by Gallup (called Gallup Q12) was based on four levels of employees' development needs and employee performance3 (Exhibit 2). Other popular engagement survey models were formulated by Willis Towers Watson (Engaged, Enabled, Energized), Aon Hewitt (Say-Stay-Strive), and Mercer (Achievement-Camaraderie-Equity).4 A large number of boutique firms also offered this as a service.5 After the data were collected, the vendor firm collated the data and presented it as multiple cuts, along functions,

1 https://www.forbes.com/sites/shephyken/2018/02/25/a-600-billion-employee-engagement-problem-solved-empathy/#66d6d52fb1a3 last viewed on June 30, 2020. 2 https://www.forbes.com/sites/victorlipman/2013/01/18/why-are-so-many-employees-disengaged/#50c83d6e1e22 last viewed on June 6, 2020 3 https://www.gallup.com/workplace/285674/improve-employee-engagement-workplace.aspx#ite-285716 last viewed on May 5, 2020 4 https://www.sirota.com/employee-engagement-surveys/employee-engagement-model/ last viewed on June 30, 2020 5 https://vendordirectory.shrm.org/category/testing-surveys/employee-engagement-surveys and https://www.hrotoday.com/wp- content/uploads/2018/06/hrotJunePDF_EngagementProviders.pdf last viewed on May 5, 2020

Page 2 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

geographical regions, managers, etc., apart from reporting the organization-level scores. This was typically followed up with focus group discussions across all employee segments to understand the underlying sentiment and factors causing any dips in the trend, leading to broad-level action plans to address concerns. The HR function then followed up with the respective teams to ensure the execution of the intended action. The whole cycle could be completed in anywhere between 2 and 6 months.

A Chief Human Resources Officer (CHRO) of a large organization lamented the perceived futility of the process and the insights gained from it.

The time from data collection to action is sometimes so long and never done in some cases. This dilutes the impact of the whole exercise, which itself is fairly costly. Engagement is a flow measure, and actioning on a stock insight is akin to taking one's temperature once a year and giving a healthy/not-healthy certificate on its basis!

Many business leaders also viewed it with some cynicism and questioned this well-established practice. Opined a senior leader in an MNC organization:

We go through this rigmarole, apply broad-based interventions, and some of the critical talents still leave. If my engagement scores are high, this should not happen. Maybe, the deeper issues do not surface through these engagement survey methods. How can I know ahead of the curve which employee is disengaged and why? That way, there should be some remedial time to heal and rebuild engagement, rather than getting a shock when the resignation is tabled.

Some organizations had adopted a mood meter using emoticons to understand the daily sentiment of employees. These ranged from emoticons at the access gates, either during entry/exit time at office or both, to emoticons presented on the laptop during the login. While these were able to generate sentiment trends, actionable insights were still challenging to obtain. Raghu, a senior delivery manager in an IT firm which ran this practice of collating the mood trend for his firm, debunked the premise behind the method, stating:

I could be happy due to the smell of rain or my child's superior performance at school rather than something at work.

Employees sometimes demonstrated a low level of trust in the managers and sometimes in the organization; high levels of skepticism were seen among the younger workforce.6 There was a challenge in getting people to speak up honestly, without fear of reprisal, and not having the communication filtered either by managers or HR partners. The existing tools for measuring engagement did not help enhance trust or engagement, and the problem seemed insurmountable. Recognizing this reality, many HR functions and organizations conducted leadership-connect sessions to enable frank upward communication through practices such as "Coffee with the CEO", open house forums, etc., wherein groups or individuals spent time with the CEO or other leaders. The leadership-connect sessions were mostly random and did not cover all employees. Recognizing this opportunity space, some new firms started to provide a platform for solving employee engagement concerns, while the existing survey-providing firms looked at acquisitions to develop this capability. Aon Hewitt acquired Modern Survey, Virgin Pulse acquired ShapeUp and Global Corporate Challenge, Ceridian joined forces with WorkAngel, and Reward Gateway picked up Yomp.7 Employee engagement analytics was expected to have the most significant impact on future talent management practices in organizations.

6 https://hbr.org/2010/01/faith-in-firms-as-low-as-youd-expect, last viewed on June 6, 2020 7 https://campaigns.thestarrconspiracy.com/2016-employee-engagement-brandscape/the-future-of-employee-engagement/ last viewed on June 6, 2020

Page 3 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

GENESIS OF AMBER

As a 16-year-old student in college, Jain was deeply affected by the trend of students losing their identity and morphing into a number-based identification. The same phenomenon extended into the corporate environment, as he realized later, where every employee was synonymous with an employee code.

Even in relatively large families, everyone has a name and a distinct identity. However, with scale, be it a college, an organization, or a country, the individual's uniqueness is lost, and a number becomes the identity. In these environments, individual voice, opinions, feelings are not heard, and providing personalized attention becomes an impossible task. I wanted to provide a platform that enabled every individual voice to be heard, which is not necessarily the loudest or the group think sentiment.

In this context, Jain started his new venture, inFeedo, to develop an anonymous platform for employees within an enterprise to express their concerns. As soon as they launched their new platform, Amber (which they positioned as a Smart AI Assistant that would work toward "making the voice of every employee felt, heard and valued"), the inFeedo team quickly realized the difference between the envisioned use of the platform and how the platform was actually being used. It was evident that the platform did not serve the primary intent. Reminiscing over the trend, Puri wryly smiled and remarked:

Most of the issues and themes voted as the most significant and critical were often the most insignificant aspects of apropos driving employee engagement on an anonymous platform. We were amazed to see that the differential size and flavor of the birthday cake during birthday celebrations was voted as the top issue by 400 of the 600 employees in a firm. Similar trends emerged from other organizations. We quickly realized that the anonymized social platform only enabled negativity spewing without the desired personalization to address meaningful grievances.

The challenge was in marrying the desired vision of being the voice enabler and champion of the faceless employee while helping the client implement its HR agenda. With most HR functions having tangible metrics of cost, productivity, engagement, attrition, etc., the platform had to translate the voice inputs into measurable key performance indicators (KPIs) of the HR function. Jain felt that this was a critical decision-enabler for successfully selling the product to the HR teams, acting as both the gatekeepers for acceptance and the champions for enabling adoption. Explaining this rationale, Jain said:

If data science and HR were to elope and marry, inFeedo should be the baby.8

With an intent to enable the CEO/CHRO's agenda of making the employees heard and valued, the AI-based chatbot "Amber" was redesigned. An additional realization was that drivers of engagement differed during different life stages, with varying psychological needs not addressed by a standard and common engagement questionnaire. Rupa Chatterjee, the senior HR leader of a client organization, doubted the relevance of some of the questions in traditional engagement surveys:

It's ridiculous to ask someone who is more than five years in the organization if they have a friend at work! The concern about assimilation, onboarding experience, etc. should be the top-of-mind issues for an employee who has just joined. The drivers of engagement at different life stages within the organization would be different for employees at different stages of their careers, which most survey firms do not address.

In order to solve this issue, Amber was designed to customize the questionnaire such that different questions were asked to employees going through different career stages (Exhibit 3). These questions were statistically linked to drivers of engagement for that particular stage. The positive response from CEOs and CHROs of mid-sized and large

8 https://www.youtube.com/watch?v=FQZfLRrcomo, last viewed on May 5, 2020

Page 4 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

firms saw inFeedo slowly and systematically gaining traction in this domain.9 While organizations continued with the practice of using annual engagement surveys, the benefits of using Amber for insights were provoking the question if the engagement survey practice could be wholly replaced by Amber's dashboard (Exhibit 4).

INTRODUCING AMBER

In most organizations, Amber was introduced as the "Virtual Assistant to the CEO" with an initial mail sent (by the bot) through the CEO's email ID, encouraging employees to engage with her (Exhibit 5A & B). The CEO's sponsorship also included the time-based touchpoints initiated by Amber (Exhibit 5C). The communication informed employees that their conversations would be "confidential but not anonymous". For each organization, the permissions to view conversations were customized, with most customers opting to have the viewing rights restricted to the CEO, CHRO, or both. In larger organizations, additional permissions were given up to the function head. Employees also had the option of using an anonymous platform, apart from their open chat. The configuration of Amber's reach was decided mutually with the management team. Talking about the cadence of connecting with the employees, Jain had the following to say, "Ability to act needs to correlate with the frequency of inputs from the employees. That is why annual surveys work. If we collect inputs too frequently and then do not act on them, it dilutes the trust and efficacy of the voice platform."

Based on the employee's milestone, Amber would initiate the conversation through an email chat weblink and not through an app. Each chat started with an emoticon. Response to this triggered the subsequent questions through Amber's AI that helped the conversational flow by deploying sentiment analysis and showing empathy with user feedback and following up to understand the concerns better. The platform sent daily reports to the configured recipient within the organization on the status of the chats completed (Exhibit 6). It enabled the recipient also to view the individual chats of any specific employee. Apart from these daily mailers, the configured recipient was privy to the dashboard (Exhibit 7A, B, C & D), which indicated the percentage of employees contacted by Amber, the response rate to chats initiated by Amber, the average mood within the organization, and the benchmark comparison with the best in class firms using Amber services. Further, the platform NLP (natural language processing) and sentiment analysis allowed thematic representation of engagement drivers such as "My Career and Learning", "My Organization", "Senior Leadership", “My Organization Culture”, “My Manager”, and “My Work”. The percentage of positive, neutral, and negative responses flagged the major concern areas within the organization. An additional feature of “People to Meet” predicted who were likely to attrite among the actively disengaged people. This enabled the leader/HR partners to engage in discussions with the dissatisfied employees. In organizations where the people data flowed from the Human Resource Information System (HRIS) integrated with the inFeedo platform, the talent segmentation information allowed a quick view of critical and high potential talent that was disengaged. By “watchlisting” this talent, critical talent insights were also available. Speaking on this functionality, the CHRO of a client firm of inFeedo stated the following:

The reality is that not all dissatisfied people are employees we want to engage with and retain. Sometimes, we would be happy to see the deadwood leave. They spread more negativity by staying on in the organization. Also, complacent employees stagnate, create productivity leakages, and are detrimental to growing the organization's talent capability. Amber was accurately able to pinpoint the disengaged employees. However, having a linkage with the high potential employees (HIPOs) ratings allows me to prioritize having this conversation with the talent I would like to retain and possibly reverse an intention to attrite through positive actions that reaffirm faith.

The platform also presented the engagement level across tenures in the organization and provided information at various levels such as the business unit, geography, and even at the manager level.

The “Anonymous Bat” function, akin to a whistleblower mechanism, allowed employees to confidentially share information that they may not have been comfortable sharing publicly or in a confidential chat format. Employees

9 https://infeedo.com/customers, last viewed on June 30, 2020

Page 5 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

could initiate this reporting through any of the webmail links shared by Amber. While the employee remained anonymous, any follow-up response from the configured recipient of the information showed the name of the person responding. In instances where employees felt confident in engaging further, they could share additional details. However, most of the “Anonymous Bat” inputs did not allow for affirmative action due to the absence of specifics and traceability but provided the employees a whistleblower option.

IMPACT OF AMBER AND THE ENHANCED CUSTOMER EXPECTATIONS

The practical and applicative insights generated by Amber resulted in tremendous traction among HR heads of organizations resulting in a rapid increase in new customers as well as renewals (Exhibit 8), along with industry recognition and awards.10 inFeedo was rated among the Top10 Startups by NASSCOM,11 the premier trade and commerce body of the Indian IT industry, and it won the “Number 1 AI Startup Award” at iTech2017. The demonstrated ROI of the platform (Exhibit 9) through reduced attrition, increased engagement scores, better response rates, etc., resulted in a large number of positive client testimonials for Amber.12 The annual subscription was priced at differential rate slabs, depending on the number of employees covered by the platform. For up to 1,500 employees, the annual rate in 2019 was $11,000. A manufacturing company with 1,200 employees determined the ROI of the platform and found it to be an extremely wise investment (Exhibit 10).

With the immediate problem of personalized engagement addressed, CHROs of client firms came back with a flurry of expectations from Amber, in its next avatar. Despite better response rates than annual surveys, a few employees still did not engage with Amber. The non-response by a few employees to the initial invite to chat with Amber resulted in follow-up emails at pre-determined time gaps, again from the CEO’s desk, gently reminding the employee to speak up. While a few employees responded to the nudge, there was a correlation between the silent employees and subsequent attrition. The CHROs wanted the non-response trend to be fed into the “Talent at Risk” call-out on the dashboard.

The freedom to initiate a chat with Amber was not available to the employees. Amber was configured to connect with employees once they completed specific tenure-based milestones within the organization (15-30 days, 6 months to 1 year, 1–1.5 years, etc.) (Exhibit 3). While Amber's time-based check-in and its associated rationale were explained to the CHROs and subsequently to the employees, some employees who fell between the check-in timelines felt “discriminated” when Amber did not reach out to them but contacted a colleague. An employee who had completed 20 days had to wait for Amber for another 10 days to initiate the chat. A few employees reached out to their HR partners expressing keenness to talk to Amber proactively, rather than waiting for the appropriate milestone-based check-in to be initiated by Amber. Although the “Anonymous Bat” option was accessible to the employee after Amber had begun the chat, many of them preferred the regular chat. The CHRO of a mid-sized IT services firm expressed his concern as follows:

Employees leave after they try all avenues to address their concerns. It is only when they feel nothing works or no one is listening – that is when the decision is taken to look for opportunities outside. Amber should be accessible for these employees also, regardless of their tenure stage and scheduled check-in points.

A few CHROs were looking for this functionality, which would enable employees to voice their opinions, grievances, or even anonymous complaints at will and at a time chosen by the employee, rather than determined by their tenure within the organization. Responding to this need, inFeedo launched a new offering by Amber called “Moments That Matter”. The tenure-based milestone used by Amber helped the client understand what worked well at a regular period, recognizing critical change or impactful moments such as a change in role, induction of a new manager, performance appraisal/increment time, mergers & acquisitions, crisis management, etc., and would

10 https://infeedo.com/newsroom last viewed on June 7, 2020. 11 https://www.nasscom.in/ last viewed on June 7, 2020. 12 https://infeedo.com/customers last viewed on June 7, 2020

Page 6 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

help in establishing a connect with the employee and understand the sentiment. These events had the potential to increase employee contentment and commitment towards the organization and enhance positive feedback during the tenure-based touchpoints.

ACCEPTANCE OF AMBER AND RENEWALS

Many HR partners presented annual engagement trends from the perspective of reporting accountability to the leadership team. In other words, they were more interested in reporting the lag measures13 of performance from a previous period and presenting a trend comparison. HR partners would often like to have engagement measured on a lag scale as they measured other constructs like revenue, profit, quality, and customer satisfaction, and this became an expectation from inFeedo in terms of reporting and presenting their dashboard. In some instances, organizations would even change their engagement survey vendor. Puri challenged this logic of trend comparison as well as the measurement system, with the following argument:

It is common practice for HR functions to change their engagement survey vendor, mostly due to cost (and sometimes due to measurement) considerations. In these instances, the incumbent vendor's engagement constructs do not precisely map onto those provided by the incoming vendor. How does “Say-Stay-Strive” equate with the Q12 questions? The intent should be to get actionable insights and lead measures ahead of the curve and not lag measures of engagement.

The standard dashboard presented by Amber was something that a few clients were not comfortable with. Responding to this need for customization, Jain attributed this as a reason for some of their client losses. He further explained the reasons for the unsatisfactory experiences with Amber’s introduction in a few organizations.

Unless the launch has the CEO’s sponsorship and the leadership teams’ support, subsequent adoption, and implementation of Amber as an employee-voice mechanism would fail. If this is seen purely as an HR initiative, it generally does not work. In some organizations, we have seen the leaders viewing Amber’s insights with suspicion – sort of “is big brother watching” and reporting and calling out the poor managers. So unless the intent is there from the CEO and the leadership team to expose bad managers and change the culture, people do not speak up. Unfortunately, we see the use of Amber drop in these cases.

Some CHROs would ask for a proof of concept (POC) and a trial with a fraction of the employee population. In these instances, presenting the initiative as a POC resulted in lower adoption. Furthermore, the population brought in as an experimental group always wondered why they were singled out, while the control group deprived of the Amber experience felt resentful.

Handling the chat communication also required high levels of sensitivity. Many CHROs, while privy to the dashboard and chats, simply did not have the bandwidth to personally engage with the employees at an individual level to address their grievances. Apart from the ethical issues involved in the sharing of information, which was promised to be kept confidential, the mechanism of chat information-sharing was often viewed with suspicion and therefore led to the erosion of trust with the employees. Jain narrated a few instances where HR business partners entrusted to deal with the employee feedback challenged the employee citing verbatim the words used in the chat with Amber. The sensitivity of handling this process required HR partners and leaders to be trained in engaging with employees in a more empathetic manner, without breaking the confidentiality and trust expected from Amber.

13 While a lag measure tells you if you have achieved the goal, a lead measure tells you if you are likely to achieve the goal in the subsequent period.

Page 7 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

ENGAGEMENT DURING THE COVID-19 CRISIS

The COVID-19 pandemic that erupted in February 2020 in India and the ensuing lockdown changed the ways of working for all kinds of organizations. The resulting changes in the individual and organizational behavior-dynamics were reflected in the pre- and post-lockdown periods' engagement sentiment.14 The trends indicated a reduction in the turnover intention of the high-risk employees by about 12%. Additionally, the response rates to Amber reach-out dipped by 4%, while employees adjusted to the new ways of working and increased their online presence. There was a sharp drop in the sense of purpose, which reduced by 38%, while the reduction in manager- employee engagement was much lower at 10%. On the other hand, there was a 10% increase in recognition of employees’ contributions, notwithstanding the dispersed and remote working; rather, surprisingly, the perceived accessibility and approachability of managers increased by 41%. The pandemic also highlighted the urgent need to address mental and physical well-being, especially the employees’ needs of psychological well-being. In some instances, the perceptions of reduced connectedness and support, sense of isolation and helplessness, loss of control, uncertainty about the future, etc. were all creating significant mental stress. The concern this caused for organizations was tremendous, with many million workdays being lost due to mental illness.15 Jain felt that this was a huge opportunity space16 for inFeedo, mainly because it was addressing a core vision and purpose of his company. The inFeedo team had launched a crisis management beta update to Amber’s engagement to understand employee sentiment around the sudden workforce disruption, the business crisis, and the emerging uncertainty about the future business scenario (Exhibit 11).

The boundaries of what constituted employee engagement were increasingly getting blurred, with technology solutions available beyond B2B, B2C, and P2P offerings (Exhibit 12). Within this domain, solutions on recognition, engagement, and measurement as well as wellness and well-being were gaining traction with organizations and HR functions. Combined with these trends, the HR functions were changing their approach to technology in various ways: from digitization for regulatory and compliance adherence to strategic business benefits, from process efficiencies to deriving value, and from top-down “command and control” information flow to rights-based information transparency.

On the wellness front, organizations had to grapple with the ballooning of healthcare expenses, but the demonstrated ROI of workplace wellness programs was difficult to identify and measure. This created the paradoxical situation of organizations believing that it was the right thing to do for their employees but could not justify the concrete ROI for investing in this. When constrained by budget, HR practitioners tilted towards solutions that enhanced engagement, for example, giving rewards. Out of the estimated $75 billion engagement solutions market in 2016, organizations spent only $14.5 billion on wellness, while $46.2 billion was spent on recognition 17

(Exhibit 13). Further, insights on mental health were more difficult for HR practitioners to act upon, given the societal bias against publicly declaring mental health conditions. For organizations, there could also be a concern about the possible differential treatment of such employees by their managers.

THE WAY FORWARD

The Amber product team was excited about launching the “Exit Voice” of the employees, which would provide the genuine reasons for the attrition that rarely got captured in the formal exit interviews conducted by the HR function. However, concerns were looming large on whether this could precipitate unnecessary post-mortems and witch-hunts and how the clients would use this information.

14 https://getstarted.infeedo.com/world-of-work-covid19, last viewed on June 9, 2020. 15 https://www.forbes.com/sites/carleysime/2019/04/17/the-cost-of-ignoring-mental-health-in-the-workplace/#5c2212653726, last viewed on June 9, 2020 16 https://www.grandviewresearch.com/press-release/global-corporate-wellness-market, last viewed on June 9, 2020 17 https://campaigns.thestarrconspiracy.com/2016-employee-engagement-brandscape/the-future-of-employee-engagement/ last viewed on June 9, 2020

Page 8 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

The ethical boundaries of using AI were getting blurred, with concerns around biases, data handling, and privacy issues. These were regularly surfacing in the conversations with the CXO teams. Fortunately, inFeedo had established robust data privacy protocols based on laws mandated across the global clientele and regions they served, along with Amber explicitly calling out exactly to employees who had access to their feedback. However, as a SaaS company, inFeedo was still working on processes to sensitize the downstream HR and address specific concerns on how this data on the dashboard providing employee insights was managed. Although the opportunity landscape around presenting AI-based analytics and dashboard on employee mental well-being was huge,18 the team was concerned about labeling employees, resulting in differential treatment, causing social stigma, and hurting their self-image. What would be the hidden and emerging biases in evolving such a platform? Was this a sufficiently compelling value proposition for organizations and CHROs to pay for this service? Should Amber restrict itself to providing insights or extend the domain to providing AI-based counseling?

Further, the COVID-19 pandemic had raised uncertainty of what would change about the psychological contract for employees. Any changes in the AI to adapt to this had to anticipate the trend, and they were not sure what was temporary and what was likely to be a more permanent trend.

As the boundaries of AI applications were getting blurred, the complexities and the potential of what was presented as the use of AI in the domain of HR were changing rapidly. Jain knew that this debate would evoke strong emotions and points of view not only within his team but also with his customers.

18 https://www.openpr.com/news/1897921/mental-health-software-market-size-to-reach-usd-2760-6-mn-in-2024, last viewed on June 30, 2020

Page 9 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

Exhibit 1 Major Players in the Employee Engagement Survey Business

Source: https://campaigns.thestarrconspiracy.com/2016-employee-engagement-brandscape/the-future-of-employee-engagement/#market- share, last viewed on June 30, 2020

Exhibit 2 Gallup Q12 Employee Engagement Survey Questions

Source: Gallup Employee Engagement Survey Framework

Page 10 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

Exhibit 3 Tenure-based Engagement Questions

Source: Internal Documents of inFeedo

Exhibit 4 Comparison between Annual Engagement Method and Amber Platform

Source: inFeedo Investor Deck April 2020

Page 11 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

Exhibit 5A Introduction of Amber within an Organization

Source: Internal Documents of inFeedo

Page 12 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

Exhibit 5B Introductory Mail on Amber from the CEO’s Desk

Subject: A new employee at ****** who is not human

Dear All,

We are pleased to welcome a new one-of-a-kind employee named Amber into ******. She is an artificially intelligent assistant (BOT) who is helping the leadership team to act on the areas of improvement that surface in our culture.

Amber will touch base with a selected few, every now and then, to understand and empathize with their journey in ******** so far. Since she is not human, you can open your heart out without any fear of judgment.

Please take two minutes of your time to interact with her if you receive an email from her. Looking forward to some meaningful interactions.

Regards,

Source: Internal Documents of inFeedo

Exhibit 5C Milestone-based Connect Mail on Amber from the CEO’s Desk

Subject: Ram*, congratulations for completing 6 months* in ********. How has it been?

Hi Ram*,

As an ongoing activity in our company, I regularly connect with people on special days in their journey with us to understand their experience at *******. Just like the last time, I would love for you to have a quick 2-minute chat with Amber and share your feedback. It goes without saying that your feedback will only be visible to me and will be kept strictly confidential.

You'll also have an option to share your thoughts anonymously towards the end of the chat, in case you feel there's something other than your feedback that I should know. So, share anything and everything. Click here to begin your chat. Cheers!

Source: Internal Documents of inFeedo

Page 13 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

Exhibit 6 Daily Reporting on Chats Completed

Source: Internal Documents of inFeedo

Exhibit 7A Dashboard Analytics on Chats Completed

Source: Internal Documents of inFeedo

Page 14 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

Exhibit 7B Employee Analytics on Demographics/Location/Function

Source: Internal Documents of inFeedo

Exhibit 7C Daily Reporting on Talent at Risk

Source: Internal Documents of inFeedo

Page 15 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

Exhibit 7D Overall Sentiment Analysis

Source: Internal Documents of inFeedo

Exhibit 8 Growth of InFeedo – Revenues, Clients, and Employees Covered

2016 2017 2018 2019 Unique Customer Firms 6 40 65 108 Percentage of Customer Renewals - 100% 98% 95% Employees in Firms 4700 37000 120000 300000 Revenue (USD, in thousands) 51 240 665 1720

Source: Internal Documents of inFeedo

Page 16 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Source: Internal Documents of Infeedo

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

Exhibit 9 Customer Testimonials on Accrued Benefits

Source: Internal Documents of inFeedo

Page 17 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

Exhibit 10 Engagement Metrics in a POC with a Manufacturing Firm

Total number of employees engaged through the platform: 1200 Total number of Amber chats (including repeat chats by Amber with the same individual): 3627 Total number of chats answered by employees: 2605; Total number of chats dropped: 71; Total number of chats that did not start: 951 Current response rate (indicated on the dashboard): 73.8% Watch-listed employees (Critical talent and HIPO employees called out in the HRIS): 731; Watch-listed employees at risk: 78 Watch-listed employees’ response rate: 75.4% (1846 out of 2449 chats for 731 employees) Total at-risk employees: 145 (emerging from sentiment analysis and called out on dashboard, including 78 watch- listed employees)

Source: Internal Documents of inFeedo

Exhibit 11 Amber Chat during COVID-19 Lockdown

Source: Internal Documents of inFeedo

Page 18 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.

Amber by inFeedo: The CEO's Virtual Assistant Revolutionizing Employee Engagement

Exhibit 12 Employee Engagement – Technology Solution Landscape

Source: https://campaigns.thestarrconspiracy.com/2016-employee-engagement-brandscape/the-future-of-employee-engagement/ last viewed on June 30, 2020

Exhibit 13 Employee Engagement – Technology Solution Landscape in the United States (2016)

Company Size 2,500 TO 4,999 5,000 OR MORE ENTERPRISE TOTAL

Firms 1,848 1,909 3,757 Total Employees 6,431,910 38,878,474 45,310,384 Annual Payroll $349B $2.070T $2.419T Recognition Market $4.189B $24.844B $29.033B Wellness Market $1.396B $8.281B $9.677B Measurement Market $0.063B $0.373B $0.436B Other Engagement Market $1.047B $6.211B $7.258B Total Employee Engagement Market for Enterprise Businesses

$6.695B $39.709B $46.404B

Source: https://campaigns.thestarrconspiracy.com/2016-employee-engagement-brandscape/the-future-of-employee-engagement/ last viewed on June 30, 2020

Page 19 of 19

For the exclusive use of M. Corella Zelaya, 2024.

This document is authorized for use only by Mariana Corella Zelaya in MIS_520_Summer_6_Tomblin taught by Michael Tomblin, Westcliff University from Jul 2024 to Aug 2024.