The Caregiver Retention section of our AI module measures and analyzes Caregiver "value". There are two major components to this measure, Caregiver Value Index and Know Your Caregivers
Caregiver Value Index - This section provides a score for caregivers on a scale of 0-1. At the bottom of the screen, you will see the individual caregiver scores, expressed as a decimal. The higher the score, the higher the value of the caregiver. It is important to note that this score is entirely independent of the Churn (Retention) Score
The AI model segments the caregivers into five categories based on their tenure with your agency and their recent income (last six months):
Champions - This is the highest level. Caregivers in this category have the longest tenure and highest recent income (High tenure, high income).
Loyal - These caregivers also have long tenure, but medium recent income (high tenure, medium income).
Promising - These caregivers have the potential to become Loyal or even Champions. They have both medium-length tenure and medium income (Medium tenure, medium income).
Low Performers - These caregivers have fewer visits or lower value visits, so while their tenure is of a medium length, they fall into a low-income category (Medium tenure, low income)
New Caregivers - These caregivers are the most recent hires and have both the lowest tenure and the lowest income. (Low tenure, low income)
The pie chart shows you the distribution of each category in your agency.
Know Your Caregiver - This section provides a score for caregivers on a scale of 0-1. At the bottom of the screen, you will see the individual caregiver scores, expressed as a decimal. The higher the score, the higher the risk of caregiver churn (caregiver leaving your employ).
This score is predicted by our AI model which considers the following factors:
Tenure - The duration of employment, measured in days. I.e. the number of days the caregiver has been employed by your agency.
Hours Spent - The total number of hours for which the caregiver has provided service to your clients. The sum of all visit hours from the hire date to the current date.
Gap First Visit - The number of days between hiring and the first client visit. This is a top driver of high churn in the home care industry.
First 3 Months Income - The caregiver's income in their first three months of employment with your agency. It measures how active they have been in your agency.
Last 3 Months Income - The caregiver's income in the most recent three months of employment with your agency.
Average Visits per Client - The number of total visits by the caregiver divided by the number of clients served. This measures how intensive their staffing is.
Age - The biological age of the caregiver.
Total Clients - The total number of clients served by this caregiver to date.
Overtime Hours - The sum of all extra hours the caregiver has worked to date.
At the bottom of the screen, you will see a list of all your caregivers and their respective scores, sorted by name alphabetically. The headers are sortable, so you can view the data points that are most meaningful to you.
Name - Caregiver's name as it appears in their Caresmartz360 profile. This is also the default sort.
Average Hours per Week - the average number of hours a given caregiver works in a week.
Hire Date - the Caregiver's hire date as captured in their profile.
Retention Score - The Caregiver's overall Retention as calculated above.
Value Score - The Caregiver's Value score as calculated above.
Segment - This column depicts the group to which the caregiver belongs. This section is independent of retention score but related to value score.
The following section is a brief overview of the calculations in place to drive the Retention and Value Scores.
SCORES AND CALCULATIONS ON CHURN DASHBOARD
Churn Score - This score indicates the Turnover/churn score of a caregiver. A high churn score indicates high chances of turnover.
Calculation Method – This score is the probability of the caregiver leaving the agency's employ. The probability is derived from a pre-trained predictive ML model which has historic data of both active and terminated caregivers.
The following is displayed in the KYC section:
0 <= turnover_probability <= 0.33
0.33 < turnover_probability < 0.66:
turnover_probability >= 0.66
Primary Contributing Factors:
The factors are derived from the agency data to understand which feature variable impacted the most to arrive at the possibility of turnover.
Tenure in days
Total on duty hours
No. of days unengaged after first visit
Income in first three months
Income in last three months
No. of visits in last three months
Average Visits per Client
Age in yrs
No. of different clients served till date
Total overtime in hours
Value Score and Segment:
Value scores are mathematical computations based on income and tenure.
An unsupervised ML model is used to do segmentation(clustering) of caregivers based on similarities. Value scores are then used in labelling and identifying the appropriate segments.
There are 5 different bands /segments (Clusters) based on service value scores.
Value score range
0.8 – 0.9
0.65 - 0.8
0.5 – 0.65
0.5 – 0.65
0.1 – 0.3