Babù

Our Babù group project journey

​

The project we were asked to do for our DHRA class seemed very complicated at first and we feared that the amount of work to be done would be too much in such a short time, however, we were not ones to give up when things got scary and we dived right into it. The task was clear: look at the dataset of Babù company, look at their network, analyze the data, and find problems within. With the support of articles and data, we had to come up with ideas of why the company struggled, how can we prevent that from happening and how can we solve current issues. 

​

The data set given to us had to be uploaded and visualised with Cytoscape, which is a program that laid out the company's network for us. After that, we colored the nodes according to their role attribute and applied a layout that uses the frequency of interaction to discriminate between nodes. That step was crucial for the whole process because without that we had no chance to see the important information, which employees are not well connected and which are overwhelmed. After trying different layouts, we chose the “Edge-weighted" spring embedded layout” (using frequency to discriminate between the nodes) because it seemed like the best-fitting layout out of the ones we could choose from.

But the visualization itself was just the beginning of our journey, next up was looking at the network structure and evaluating the connections formed in front of us. To our relief, we found no evidence of isolated nodes, which means nodes that have zero edge count. It would have been very disturbing if we found nods that were not connected at all. However, we did notice some peripheral nodes which had relatively few connections compared to other nodes in the network. Those were the ones we knew we had to focus on and find out more about. To represent the node sizes we chose two metrics that highlighted the peripheral nodes.

​

Degree centrality - This was the first metric we chose. It measures the number of edges connected to a node. Nodes with a low degree of centrality are likely to be peripheral nodes, as they have fewer connections compared to other nodes in the network. 

Betweenness centrality - This was the second metric we chose for our project analysis. It measures the extent to which a node lies on the shortest paths between other nodes. Nodes with low betweenness centrality are less likely to be on the main path of communication and can be considered peripheral. 

After creating two schemes based on those two metrics, we realized we preferred using degree centrality as the final metric for the node size. We believe the size of DM1 (central nod) was too big when using the betweenness metric and it could be misleading. So for our final structure, we decided to go with the degree centrality metrics scheme.

What did we find out?

​

When we looked at the number of connections within the network (edgecount on the table panel) we noticed that nodes DM2 and DM6 have only 2 links, DM3 and DM5 only have 3 links, DM4 has 5 links while DM7 has 11 and DM1 has 12. That showed us that the majority of digital marketing specialists are not very well connected and appear to be quite peripheral compared to the other nodes of the network. We had to take a closer look at certain nods to assess the situation and analyze the issues that may come with their lack of connections.

​

DIGITAL MARKETING SPECIALISTS DM2, DM6 and DM7 

DM2 and DM6 - have some links with other functions (marketing specialists and designers), but they are not connected between them and to their boss (DM1). This was something very alarming since it is very crucial that there is a clear communication path and relationship with their superior. DM7, on the other hand, is connected to the manager (DM7) and at the same time has the most connections with other functions. Another noticeable thing was the fact that those digital marketing specialists were not connected with the other digital marketing specialists group (DM3 DM4 DM5).

 DIGITAL MARKETING SPECIALISTS DM3 DM4 DM5

Unlike the previously described group, these marketing specialists had a connection and communication with their superior (DM1) but they were also quite poorly connected to other nods and therefore appeared quite peripheral

​

Risks the company is facing due to the way their network is connected

 

After we looked at the way people in the company are connected we found through research a few risks that the company is facing. It was important to assess both the financial and personal aspects of risks in the company.

 

• Turnover remains high; the company had at present time a very high turnover ratio and in our opinion it seems clear that if nothing changes, the high turnover ratio will definitely not decrease over time. For the company this is the key problem and it is a structural issue and not a phenomenon related to contextual factors, which means that it is very unlikely to be temporary and has to be faced and fixed.

 

• Higher costs: a high turnover ratio translates directly into huge costs for the company in many ways. Some of the most relevant are: administrative costs, hiring costs, training costs, as well as indirect costs such as the line managers’ time (he has to devolve some time to align employees to company’s objective) or overworking risks for other team members (they have to compensate the missing work of the leaving employees). [2, pp. 630]

 

• Reputational risks and negative influence on employer-branding: another way the high turnover ratio influences the company is its reputation. If a high number of employees leave your company, especially after a short time, it will negatively impact your reputation among professionals of the selected sector, making it substantially difficult to acquire new talents. And if we are honest, no one will be eager to start working for a company if their employees don't stay for long.

 

• Shortage of critical knowledge: with employees leaving, the company loses important knowledge and experience of their workers as well as their investment into their training. It is imperative that the company retains the individuals that possess the critical knowledge of interest, otherwise they face a risk of being left behind by its competitors and losing its position on the market.

 

• Contagion theory: this theory suggests that an employee’s propensity to leave a job can be influenced by whether other employees have also left their jobs or are intending to do so. Employees compare themselves and their attitudes to others to determine if they should leave. This provides both an additional explanation to the high turnover ratio, and at the same time constitutes an additional challenge, suggesting that the turnover ratio should be minimised. [8, pp. 252]

 

•  External networking behavior: the risk of losing employees to other companies is clearly connected to the way their at work social network fulfills their need for social connections. Scholars have found that employees’ behavioral frequency of building and maintaining ties with people outside the firm was positively correlated to turnover. If employees do not develop enough connections with their co-workers, there is a higher probability that they will develop connections outside of the company, leading to increased risk of leaving the company  [8, pp. 256]

 

• Peripherality: research shows clearly that the average number of connections required to reach all other employees in the organization influences turnover. And employees whose connections operate on the periphery are more likely to leave. This suggests that keeping key employees at the periphery of the social network puts the company more at risk of losing them. [8, pp. 261]

 

• Lower engagement and low job satisfaction: if we look more closely at the people aspect of a high turnover in a company, as well as a social network where key individuals are kept peripheral, we understand that it will lead to an environment that does not promote putting efforts in socializing and creating new links. For example, employees won’t know if the newly hired person will stay in the company for a long time, promoting uncertainty. All this together leads to the perception of a workplace that does not care for its employees, which has been proven to be one of the main drivers of low job satisfaction. This can manifest in the form of quiet quitting even among employees belonging to other units, with potentially catastrophic consequences.

 

• Well-being and psychological risks: a fragmented social network may lead to alienation for some employees, given the lack of clarity regarding the hierarchical structure and the congested information flows. Employees that don’t know who to talk to and who to cooperate with will feel abandoned by their company. That can impact not only their performance but potentially even their mental health.
   

When we looked at the risks before us, we could very clearly see what our goal for babù should be. Our main goal had to be to restructure the social network of this company in need. We would advise to restructure the entire Digital Marketing unit, with the following ideas and criteria in mind. Our ideal structure is one that is quite similar to the one already existing within the Sales Unit (SAL#): First, we would need a team leader, in our case DM1, that represents the main bridge with the other units, but also has direct connections with everyone within the DM unit, in order to align everyone’s work towards the main focus. Secondly, we would advise to support them with one or two secondary coordinators that will still have a number of connections with employers from other business units, which will achieve the double objective of reducing the risk of overworking the line manager, and provide a safety net in case the line manager were to leave the company (even temporarily, for things such as a pregnancy or illness). For this role, we considered DM7, who already has a surprising amount of connections with members outside of his unit, and could easily provide support in that sense. We would also like for all the DMs to be connected to each other, to the point of achieving the formation of a “clique”, reducing to the minimum the necessity of intermediaries within the unit and boosting teamwork and knowledge exchange. We also have to pay special attention to the role of manager of the unit, not only from the organizational chart perspective, but as to the leader and the manager. Teamwork and trust are crucial. Latest research says that around 35% of employees leave the company because of an uncaring leader. [5] Employees should feel valued by the employer, if they are - they are motivated to do their best. [4, pp 46] Therefore we would recommend making sure that the manager tries to create the best work environment possible.

 

But before implementing anything concrete, we must make sure that managers have awareness of the current state: this can be achieved through presenting them with some social networking analysis, explaining to them the correlation with turnover rate,  and integrating with reports about the satisfaction of the employees. [2, pp.629] 

 

What are the things we would suggest to do

 

• Learning from what works: an interesting idea could be to organize an interview with the individual identified as DM7: their case is especially interesting, given the fact that this person formed a great number of ties with other co-workers. Having conducted the interview, we can explore the directions of those connections deeper and get a better understanding of whether the employee provides support, advice, information etc. to his/her network and if this behavior is adequately monitored and rewarded by the manager. We can monitor this situation and relate it to his\her intentions to leave. [5, pp. 578]. Analyzing their approach to networking could give great insights to improve on the processes. 

 

• Improving social inclusion and organizational culture improvement: providing opportunities for informal interactions (team building, coffee breaks, short informal chats) are a great way to give employees more time to bond and develop informal connections to build socio-centric network systems. [2, pp. 633] To do that we can also use new interactive technologies to reduce the social distance between supervisors and subordinates, such as videoconferencing and internal social media (Stone, D., Deadrick, D.L., Lukaszewski, K.M., & Johnson, R. (2015). The Influence of technology on the future of human resource management). In order to promote social networking and cross-functional projects we should also concentrate on the organizational culture. An idea supported by academic studies is to de-emphasize the individual achievement in favor of a view that promotes seeing one’s work as an integral part of the contribution towards a collective goal, providing a shared mental model framework. [2, pp. 635] 

 

• Creating coaching and mentoring programmes that involve multiple people: this area is closely connected to the social inclusion and building of relationships we already talked about.  The quality of an employee’s social relationship with one primary mentor had no association with turnover, whereas the quality with an entire set of mentors at work reduced the likelihood of turnover. This can be enhanced through a restructuring of the DM unit, and reinforces the creation of ties among the various members. A secondary leader figure can be helpful in this sense, such as the ideal position we would like DM7 to cover. [8, pp. 271]

 

• Improve the on-boarding process: studies found that the ties that are created during the first few months of a new working environment are on average among the strongest. Hence, promoting the formation of new and stable ties during the onboarding process can be a great way to enhance retention. This can be achieved through different ways, such as imposing (at least for the first period) physical proximity [2, pp.632] (sharing a working space), which has also been proven to improve the number and consistency of ties [1, pp. 20]. Furthermore, early connections between newcomers and more seasoned employees make it more likely for the former to stay.

 

• Revising job description and comparing the salaries: and comparing with what people are actually doing, as well as potential mismatching during the recruitment phase, as it may be a source of dissatisfaction which in turn leads to turnover. Another important thing may be to keep up with the market benchmarks and make sure the company is being at least in line with them. Payment must be coherent with the role [2, pp. 634]. Managers need to acknowledge and reward contributions that exceed expectations, as well as behaviours that go beyond their job description (such as mentoring, helping newcomers).

 

Key performance indicators

​

• Turnover rate: self-explaining, consists of the number of DM employees who left over the number of DM employees within the digital marketing department. 

• Voluntary turnover rate: alternative measure to turnover rate, helps us get a cleaner look to what we would actually like to boost and eliminates elements outside of our control (firings, transfers due to family issues, etc.). Voluntary departures/average of total employees (only for the DM Unit)

• Degree of centrality: number of total connections, basic metric used for the analysis of social networks in multiple scenarios [2, pp. 631].

• Frequency of interaction: a metric that periodically measured over a specific time frame the amount of interactions between members. This helps to keep track on a micro level of how things are evolving and allows to notice potential issues between two nodes and intervene quickly. 

• Tie Strength: frequency of interactions over time, especially if combined with metrics such as reciprocity and/or surveys where we ask to evaluate the level of intimacy of a bond, can be used to estimate tie strength. High tie strength is highly desirable. 

• Density: number of connections a participant has over the total possible connections a participant could have. We could set a target level of density that we would like the employees to achieve. 

• Number of shortest paths between two nodes: Breadth-First-Search algorithm.

• Average length of employment: total number of years of employment for DM/total number of employees. It is an additional metric, but not necessarily the most important.

• eNPS (employee net promoter score): on a scale from 1 to 10 how likely is the candidate to recommend “Babù” as a place to work (after the responses, the survey reveals three profiles: promoters, detractors and passive).The alternative can be ESI.

 

Job Satisfaction Rate: we would like to create a metric called “Job Satisfaction Rate”, that is defined as a combination of different sub-metrics that includes elements such as (some of these have been gathered going through academic papers to get a better understanding of what has been found to be effective [1, pp. 32]): perceived effectiveness of the onboarding process; thoughts of quitting; perceived Stress; psychological Safety; perceived Job Recognition (in its many forms); High job satisfaction helps to decrease turnover and absenteeism, as well as has a positive effect on task performance behavior, contextual performance behavior and customer satisfaction. Because job satisfaction seems to be a key driver of some aspects of job performance it’s important for managers to understand just how satisfied their employees are. All of these elements can be gathered through surveys (even anonymous ones) and interviews and the data gathered should help improve the performance and satisfaction of the companys' employees.

 

Our experience with this project

 

The complexity of this project was very helpful since it allowed us to follow the whole process from data analysis to evaluation and finding solutions. Within the time we spent working on the project we learned how to cooperate, discuss and manage not only the teamwork but also the data itself. It is clear that such situations happen in real life all the time and we are sure that this project was helpful and we enjoyed working on it. The amount of work put into this project turned out to be very rewarding and enlightening.