Mi-ORG’s digital app, LEAP, designed to aid servicemembers in transitioning to civilian employment, has experienced substantial user engagement since its launch a year ago, with over 50,000 downloads globally. Despite its popularity, the average user rating has not met expectations, indicating a disconnect between user expectations and the app’s performance. As Mi-ORG approaches its annual product review, there is a pressing need to enhance LEAP’s user satisfaction ratings, given the significant financial investment made in the product.

The availability of comprehensive user data from the past year presents a unique opportunity for Mi-ORG to critically analyze the app’s performance and user interactions. This report, conducted by the special projects team under the leadership of Thomas Knowles, utilizes descriptive and predictive analytics to extract actionable insights from the user data, aiming to:

1. Evaluate the app’s effectiveness in aiding servicemembers’ transition,
2. Identify the most popular and impactful services, features, and user behaviors,
3. Highlight the gaps between the app’s current performance and its strategic objectives,
4. Propose data-driven recommendations for improvements and updates.

Through this comprehensive analysis and the subsequent implementation of its recommendations, Mi-ORG aims to refine and optimize LEAP, ensuring it not only meets but exceeds the expectations of its user base. This endeavor is crucial not only for the servicemembers who rely on the app but also for Mi-ORG, as it seeks to protect and maximize its investment in this digital solution.

 

Introduction

 The transition from military to civilian life is a critical juncture for servicemembers, fraught with challenges and uncertainties. Recognizing this, Mi-ORG developed LEAP, a mobile application designed to streamline this transition by offering tailored job recommendations, skill translation tools, and networking opportunities. Since its inception, LEAP has been well-received in terms of subscription numbers, indicating a significant demand for such services.

However, the disparity between the high subscription rate and the app’s lower-than-expected user ratings necessitates a thorough examination of LEAP’s performance, usability, and overall effectiveness. The app’s success is paramount not just for the servicemembers relying on it for support, but also for Mi-ORG, which has invested heavily in this product.

Thomas Knowles and his team have been tasked with conducting a comprehensive analysis of LEAP, utilizing the wealth of user data collected over the past year. The aim is to uncover insights that can guide strategic decisions, enhance user satisfaction, and align the app’s functionalities more closely with the users’ needs.

This report presents the findings of this analysis, offering a clear and actionable roadmap for improving LEAP’s performance, ensuring that it not only meets but exceeds user expectations, and ultimately solidifying Mi-ORG’s reputation as a trusted ally for servicemembers in transition.

 

DESCRIPTIVE ANALYTICS

In this section, we aim to condense our analysis to reveal clear understandings of the data's underlying patterns and trends. When appropriate, we will incorporate visual representations. Additionally, we will assess any limitations or potential biases within the data.

Examining the data, it's apparent that a substantial portion of employees falls within the 0 to 8 years of service range. The largest group, with close to 250 employees, has served for a period between 0 to 4 years, while the next interval, 4 to 8 years, comprises approximately 225 employees. This suggests a relatively high turnover rate or a large influx of new employees within this organization.

However, beyond the 8-year mark, the number of employees starts to decrease noticeably. For instance, the group with service lengths between 8 to 12 years includes over 175 employees, but the count significantly drops to slightly above 50 employees for the 12 to 16 years bracket. Moreover, for service lengths exceeding 16 years (i.e., [16, 20), [20, 24), [24, 28), [28, 32), and [32, 36) years), there are very few employees in each of these categories, with the count appearing to be less than 50 for each bracket. This indicates a more pronounced retention rate among employees who reach the 8-year mark or longer.

In summary, this chart highlights a concentration of employees with relatively short service lengths, gradually diminishing counts as service lengths increase. It's particularly noticeable that the organization experiences a drop in employee counts after 8 years of service, suggesting potential challenges in retaining employees for extended periods, with only a select few remaining in the workforce for more than 16 years.

The graph below reveals that as you move away from this central weight range, the number of records significantly diminishes. On the lighter end, there are very few individuals weighing below 110 pounds, and on the heavier end, there are few records exceeding 210 pounds. This suggests that these weight extremes are less common within the dataset, and most individuals have weights that fall within the 130 to 180-pound range.

The dataset, as depicted by the histogram, demonstrates considerable variability in weight. Weights range from below 110 pounds to over 210 pounds, highlighting the diversity in the weight data. Despite this variability, the distribution appears to be relatively symmetric, with a slight right skew. The right tail of the histogram, representing heavier weights, is longer and tapers off more gently compared to the left tail, which suggests that while heavier individuals are less common, they still contribute to the dataset's overall distribution.

In summary, the "Weight" histogram provides valuable insights into the weight distribution of the dataset. It showcases a central tendency with the most frequent weights falling in the middle, while the variability and the presence of some outliers at the weight extremes are also evident, helping to characterize the overall weight profile of the dataset.

The scatterplot below shows employees with longer lengths of service (15-35 years), the distribution becomes more dispersed. However, a slight trend is evident where employees with longer tenures appear to have completed more interviews. This is most evident with data points hovering around 10-14 completed interviews.

Interestingly, there are several data points throughout the 20-30 years of service range that have a relatively low number of completed interviews. This may indicate specific individuals or cohorts who, despite their longer service, have not engaged in as many interviews as some of their counterparts.

In summary, while there is a general trend indicating that those with longer service tend to complete more interviews, there are exceptions. The dense concentration of points in the 0-10 years range with fewer interviews completed suggests that newer employees or those with shorter service lengths may either be less likely to be interviewed or are in roles where interviews are less frequent. The outliers present in the dataset, particularly those with longer service years but fewer interviews, highlight the need for a deeper investigation to understand the underlying reasons for this pattern.

 

Feedback on Advertising

When considering potential changes to the LEAP app advertisement based on your descriptive analysis of the app's users, it's essential to ensure that the ad effectively communicates the app's value proposition to the target audience. Here are some suggestions for improving the advertisement:

1. Emphasize Military-Friendly Features: Highlight any features or aspects of the app that specifically cater to the needs and preferences of military personnel. This could include features such as deployment planning tools, communication resources for families, or mental health support tailored for the unique challenges faced by service members.
2. Incorporate Testimonials: Incorporate real-life testimonials or quotes from users within the military community. These testimonials can provide social proof and enhance the credibility of the app's effectiveness and relevance to the target audience.
3. Highlight User Success Stories: Showcase success stories of individuals within the military who have benefitted significantly from using the app. This can inspire potential users and demonstrate the practical impact and positive outcomes that the app can facilitate.
4. Invoke a Sense of Community: Illustrate how the app fosters a sense of community among military personnel, creating a platform for mutual support, information sharing, and camaraderie. Highlight any social or interactive features that encourage users to connect with others in the military community.
5. Utilize Military Aesthetics: Consider incorporating military-themed visuals or imagery to resonate with the target audience. This could include subtle references to military symbols, colors, or themes that create an immediate connection with military personnel and their families.
6. Clear Call-to-Action: Ensure that the ad includes a clear and compelling call-to-action, prompting viewers to take the next step, such as downloading the app or visiting the app's website for more information.

By incorporating these suggestions into the LEAP app advertisement, you can create a compelling and resonant message that effectively appeals to the unique needs and interests of the military community.

 

Business Analytics Summary

This stage holds crucial significance for Mi-ORG, as it enables the organization to base its decisions on data, leading to a competitive edge in the market. Moreover, it plays a pivotal role in the realms of marketing and financial management. Furthermore, the utilization of existing data for predicting future trends and outcomes accentuates the significance of employing business analytics (Lee et al., 2022).

The current analysis involves the utilization of a classification tree. Roth (2016) argues that the classification tree algorithm consistently seeks out the most effective data split feature. The tree is initially trained on a set of labeled data and subsequently utilized for predictions by traversing nodes until reaching the terminal node. The data is bifurcated into two sets: the Training and Test sets, both of which represent sample data with predetermined outcomes.

The training data is employed to train and configure the algorithm, while the test data ensures that the model is capable of handling diverse possible datasets. Out of the 500 user records, 375 (75%) were utilized to train the classification tree, with the remaining set aside for testing the model at a later stage.

In the case of the training data, a target variable must be defined. In this instance, the variable denotes whether the user reported securing a job through the LEAP app. The classification tree is tasked with categorizing users into distinct subsets based on their success or failure in using the app to secure employment. The classification tree diagram is presented below.

The prioritization of variables by the algorithm was noteworthy, placing PTS diagnosis as the most crucial factor, followed by rank when separated, the highest level of education achieved, and finally, the type of discharge as the least influential. The node-based analysis reflects the following:

Node 1 highlights that 55% of users did not secure a job via the app, marking the node as "no job." Consequently, if the algorithm were to predict job success, it would likely suggest negative outcomes, considering the majority's failure to find employment through the app. This implies that only 45% succeeded in utilizing the LEAP app to secure employment, defining the app's success rate for future projections.

The classification tree emphasizes that the PTS diagnosis serves as the critical variable for distinguishing successful and unsuccessful job seekers, leading to a division between those diagnosed with PTS and those without. Without PTS, Node 2 becomes the next point of consideration.

Node 2, somewhat contrasting Node 1, leans towards a positive outcome, suggesting that users would likely succeed in utilizing the app to secure employment, with a 51% job success rate compared to the 49% unsuccessful rate among non-PTS users.

Further analysis delves into the education level, segregating those with a BA or doctorate from those with high school degrees, AA, or Masters. For those with a BA or doctorate, the algorithm predicts a 74% success rate in utilizing the app, with the remaining 26% falling under the "no job" category (Node 4). On the other hand, users with high school degrees, AA, or Master’s exhibit a "no job" prediction, much like Node 1, as illustrated by Node 5.

The analysis continues with the algorithm isolating individuals with a high school, AA, or Master's degree and utilizing rank as the primary segregating variable, distinguishing the Field Grade Officer from other categories such as Enlisted, junior NCO, Company Grade Officer, and General Flag Officer. This progression leads to Node 10, where the overall prediction suggests a 74% success rate for users utilizing the LEAP app. However, in Node 11, the algorithm predicts a lower success rate of 47%, indicating that 53% of users in this category may not secure a job through the app, representing 69% of all app users.

 

 

NNode	Criteria to belong to this node	Is this node predominantly a Job or No Job node?	What percentage of people found jobs in this node?	What percentage of our user base is in this node?
44	No PTS, Education BA/Doctorate	Job	74%	8%
110	PTS Diagnosis, High School Education, Field Grade Officer	Job	79%	4%
111	PTS Diagnosis, High School Education, All Other Ranks	No Job	47%	69%
22	No PTS Diagnosis (No additional criteria provided)	Job	51%	81%

 

RECOMMENDED CHANGES

Mobile job seekers face a multitude of challenges when using apps or websites to apply for jobs on their smartphones. Some of these challenges are rooted in the platforms themselves. As Zhao (2018) elaborates in his research, mobile job seekers are often highly focused on securing employment, but they encounter inconveniences such as poorly optimized websites and apps, intrusive pop-up ads that obstruct vital page sections, text that strains the user's eyes, and content and forms that extend beyond the screen. These issues can significantly frustrate users of these apps and websites.

Identifying aspects of the LEAP app that are underperforming or not aligned with Mi-ORG's strategic goals is crucial for ensuring continuous improvement. Based on the information provided, here are some areas that might need attention:

Underperforming Aspects:

1. User Engagement: If the user engagement metrics are below expectations, it might indicate that the app's current features or user experience are not effectively capturing and retaining the attention of the target audience.
2. Monetization Strategy: If the app is not generating expected revenue, it might suggest that the current monetization strategy is not fully optimized or that the value proposition is not adequately communicated to potential paying users.

Short-Term Solution:

To address any immediate concerns, it is advisable to focus on enhancing user engagement through a streamlined and personalized user experience. Implementing in-app notifications or personalized recommendations based on user preferences could encourage users to interact with the app more frequently. Additionally, introducing limited-time promotional events or rewards for active users might help boost engagement in the short term.

Long-Term Solution:

For sustained improvement, it is crucial to invest in continuous research and development to stay attuned to evolving user needs and technological advancements. Developing a comprehensive user feedback mechanism, such as surveys or user testing groups, can provide valuable insights for prioritizing feature updates and enhancements. Furthermore, fostering a culture of innovation and agility within the development team can ensure that the app remains adaptive and responsive to changing market dynamics and user expectations.

Implementing a long-term strategy that focuses on fostering a robust community within the app, integrating additional value-added services, and exploring strategic partnerships to expand the app's reach and functionality can contribute to the sustained growth and success of the LEAP app in alignment with Mi-ORG's strategic goals.

By addressing the short-term and long-term solutions, Mi-ORG can ensure that the LEAP app remains relevant and competitive in the ever-evolving digital landscape, ultimately contributing to the achievement of the organization's strategic objectives.

 

Reference:

Nainawat, S. (2021). IMPORTANCE OF ADVERTISING IN BUSINESS COMMUNICATION. International Journal of Advanced Research in Commerce, Management & Social Science, volume 04(02). https://www.inspirajournals.com/uploads/Issues/248805225.pdf

Song, W., et al.(2012). Understanding User Experience of Mobile Video: Framework, Measurement, and Optimization. Mobile Multimedia-User and Technology Perspectives.https://www.researchgate.net/publication/221922885_Understanding_User_Experience_of_Mobile_Video_Framework_Measurement_and_Optimization

Roth, D. (2016). Decision Trees. CS 446 Machine Learning. https://www.cis.upenn.edu/~danroth/Teaching/CS446-17/LectureNotesNew/dtree/main.pdf

Zhao, D. (2018). The Rise of Mobile Devices in Job Search: Challenges and Opportunities for Employers.Glassdoor Economic Research. https://www.glassdoor.com/research/app/uploads/sites/2/2019/06/Mobile-Job-Search-1.pdf

Lee, C., Moslehpor, M., & Cheang, S. (2022). Predictive Analytics in Business Analytics: Decision Tree. Journal of Advances in Decision Science, 26(1): 1-30. https://www.researchgate.net/publication/357447580_Predictive_Analytics_in_Business_Analytics_Decision_Tree