Financial Services · Talent Retention Analytics

The Challenge

A quarter of the sales force leaving, and no one sure why

Rural Consumer Durable, Rural Liabilities and Rural Term Loan run on large field teams with stretching AOP targets. Exits were rising month on month, recruitment costs were climbing, and retention spend was spread evenly across everyone.

High attrition

Turnover was significant, especially in critical roles. It disrupted operations and pushed up recruitment costs.

Lack of insight

No one understood what was driving exits, so effective retention strategies were hard to design.

Untargeted spend

Money went into generic retention tactics with no clear view of their impact or fit with what employees needed.

Five layers deep

NSM, ZSM, RSM and ASM layers hid where exits clustered. A national average masked teams losing more than half their people.

HMC Approach

From a monthly attrition number to a retention plan per segment

Built with HR business partners and sales leadership, using HRMS records, AOP plans, performance data, engagement surveys and exit interviews.

Select a step to see what happened.

Discover · Join the data

Employee records, AOP plans and exit data lived in separate systems.

Joined HRMS headcount, monthly exits, band, age and tenure.
Mapped every sales officer to their ASM, RSM, ZSM and NSM.
Coded exit-interview reasons and linked survey scores to individuals.

Deliverable: One employee-month table, from April onwards.

Design · Agree the measures

Leaders agreed one definition before arguing about the number.

Annualized attrition = exits ÷ average headcount × 12 ÷ months elapsed.
AOP achievement measured at the ASM territory level.
Early attrition = exits within six months of joining.

Deliverable: Metric dictionary and dashboard wireframes.

Build · Dashboard and model

A Power BI dashboard for leaders and a risk model for HR.

Pages: Home, Attrition, Chain, Decomposition tree, Scatter AOP + Attrition.
Predictive model scoring every employee's likelihood of leaving.
Segments with a matched retention action and cost per head.

Deliverable: Power BI report, risk scores and segment playbook.

Adopt · Monthly people review

Attrition joined the sales review, not just the HR review.

NSMs and ZSMs review their own decomposition tree every month.
HRBPs act on the high-risk list before resignations arrive.
Retention budget allocated by segment, not spread evenly.

Deliverable: Review cadence and HRBP action tracker.

Scale · Beyond rural sales

The same model extends to other businesses and roles.

Monthly model refresh with new joiners and exits.
Effect of each retention action tracked against a control group.
AOP setting informed by attrition risk in each territory.

Deliverable: Refresh routine and measurement plan.

The Solution

See it, explain it, predict it, then test the fix

The Power BI report as delivered comes first. Below it is a live companion dashboard built on a synthetic sales force shaped like the real one, with five modules you can filter, drill into and simulate.

01 · The Power BI report, as built

Attrition and decomposition-tree pages

Shown as delivered. Click a page to enlarge it.

Power BI attrition page with monthly attrition, headcount by month, state map and slicers — **Attrition page.** Attrition count and annualized attrition by month, headcount by month, a state map and slicers for band, age group and department.

Power BI decomposition tree splitting annualized attrition by NSM, ZSM, RSM and ASM — **Decomposition tree.** Annualized attrition split down the hierarchy: NSM, ZSM, RSM, ASM, with a department slicer.

Exits rising monthlyFrom 37 in April to 86–89 a month by Aug–Oct; annualized attrition climbed from 18% to 29%.

Headcount growing tooFrom 2,376 in April to 2,627 in October, so replacement hiring ran just to stand still.

25% annualizedThe headline card, with every state mapped by attrition and headcount.

NSM spread 15% to 57%+The decomposition tree showed one national average hiding very different teams.

02–06 · Live companion dashboard

Talent retention cockpit

Every chart below responds to the filters. Data is a synthetic sales force of 3,000 people (April–October) built to the same shape as the real report. No real employee data is used.

Department

Band Age

02 · Exits, annualized attrition and headcount by month

Bars: exits · line: annualized attrition (YTD) · shaded: headcount · dashed: projected Nov–Mar

Attrition by state

Tile map · colour = annualized attrition · number = headcount

Band × age group

Annualized attrition; hover for counts

Why people left

Coded exit-interview reasons, split by tenure at exit

Under 6 months6–12 monthsOver 12 months

When people left

Exits by tenure at exit (months)

03 · Sales hierarchy explorer

Department → NSM → ZSM → RSM → ASM. Bar height = headcount, colour = annualized attrition. Click any block to zoom in; use the crumbs to zoom out.

Decomposition path

Press “Find the hotspot” to follow the highest-attrition branch at each level.

Most exits · ASM teams in view

04 · AOP achievement vs attrition

Each bubble is a sales team. Teams far below target lose people fastest. Click a bubble to open it in the hierarchy.

Rural Consumer DurableRural LiabilitiesRural Term Loan

05 · What predicts an exit

Model drivers as odds ratios: how much each factor multiplies the odds of leaving

Risk scores vs what actually happened

Predicted risk for people who stayed and people who left (Apr–Oct)

StayedLeft

Personalised retention strategies by segment

Current employees, by their main risk driver. Expected exits are over the next six months, from the model.

06 · Retention simulator

Choose how much of each segment to cover. The model reruns 400 times over the next six months for every current employee in the filtered view.

Budget optimiser₹80 L

No actionWith your planBars: share of 400 simulated futures

Synthetic data for illustration only: names, teams, states and scores are generated. Segment costs, effect sizes and the ₹2.2 lakh cost per exit are indicative assumptions.

Impact Dashboard

Fewer exits where they hurt most

With targeted onboarding, AOP recalibration in struggling territories and focused retention offers, attrition fell fastest in the segments the model flagged.

Annualized attrition by segment

First half of the year, before vs after

100 critical-role holders, one year on

High-performing ASMs and RSMs: who stayed and who left

StayedLeftSaved by a retention action

Figures indicative; validated metrics available on request.

Dimension	Before	After
Attrition view	One national number, a month late	Every NSM, ZSM, RSM and ASM team, every month
Root causes	Anecdotes from exit interviews	Coded reasons linked to tenure, targets and managers
Targets and people	AOP and attrition reviewed separately	Under-target territories flagged as attrition risks
Early warning	Found out at resignation	Risk scores about 60 days ahead
Retention spend	Generic schemes for everyone	Segment-specific actions, budget tested in a simulator
Accountability	HR owned attrition alone	Sales leaders review their own team's tree

Find the next resignation before it's written