Swishr Pulse - Workforce & Cost Intelligence (WFM)

Executive Dashboard

Instant staffing & cost view based on your volumes, AHT, shrinkage, and salary assumptions.

Required Headcount —

Agents (total)

Monthly Ops Cost —

USD 0

Based on salary + on-cost

Cost per Contact —

USD 0

Blended across all channels

Utilization —

Workload ÷ capacity

Staffing Mode

—

Dedicated vs blended comparison (SLA buffer applied after pooling).

Dedicated HC

—

Blended HC

—

Savings

—

Headcount by Channel

—

Cost by Channel

Monthly (allocation by workload)

Channel Breakdown

Channel	Monthly Volume	AHT (min)	Workload (hrs/mo)	Allocated HC (pooled)	Cost / Contact

Inputs

Enter your volumes, AHT, and staffing assumptions. Save & recalculate anytime.

Volumes (Monthly)

Calls / month

Total inbound/outbound handled calls.

Chats / month

Total handled live chats.

Emails / month

Total handled email tickets.

Handling Assumptions

AHT Calls (min)

AHT Chats (min)

Per chat (not per message).

AHT Emails (min)

Chat concurrency

If 2: capacity doubles for chat (approx).

Shrinkage %

Breaks, leave, training, meetings, absence.

Target utilization %

Used for guidance and alerts.

Target SLA %

Applies a simple “SLA buffer” heuristic (not Erlang).

Target answer time (sec)

Used for assumptions & warnings (informational).

Staffing & Cost

Staffing mode

Blended assumes agents are multi-skilled across channels.

Shift hours

Per agent per shift.

Working days / month

Salary / agent (USD)

On-cost % (benefits)

Adds overhead to salary (insurance, etc.).

Ops Notes

Assumptions

Baseline model

Warnings

Hourly Distribution (24h)

Split your daily workload across hours. Used to estimate peak-hour staffing (simple).

00:00–11:00

Sum: 100.0%

00:00

01:00

02:00

03:00

04:00

05:00

06:00

07:00

08:00

09:00

10:00

11:00

12:00–23:00

Tip: Make peaks match your service reality (promo hours, paydays, etc.).

12:00

13:00

14:00

15:00

16:00

17:00

18:00

19:00

20:00

21:00

22:00

23:00

Peak hour

—

Based on hourly split + utilization + shrinkage

Agents needed (peak)

—

Rounded up (simple)

Peak vs baseline

—

If > 100%, peaks may need shifts/coverage

Interval WFM Engine (MVP)

15/30-min interval staffing for voice using Erlang C, with shrinkage components. Keeps your existing estimator untouched.

Core Inputs

Interval length

Workdays / month

Calls AHT (min)

SLA target (%)

Answer time (sec)

Utilization cap (%)

Shrinkage components (%)

Sum is used as total shrinkage for interval staffing.

Leave

Sick

Training

Meetings

Other

Total shrinkage: 0%

Interval Forecast (Calls)

Paste one number per line (calls per interval). Or click “Generate from Hourly Split”.

Intervals: 0

Avg required agents (paid)

Peak required agents (paid)

Avg SLA achieved

% intervals meeting SLA

Worst interval SLA

—

Peak interval SLA

—

Interval	Calls	Agents (base)	Agents (paid)	Occupancy	SLA achieved

Notes: Voice SL uses Erlang C (single-skill, no-abandonment). “Paid” includes shrinkage (components). Use this for operational SLA staffing; the monthly planner is budgeting-level. This MVP is interval staffing only (no shift scheduling yet).

Scenario Testing

Adjust volume and AHT without changing your base model.

Scenario Controls

Volume change

All channels (%): 0%

Use this for growth/seasonality assumptions.

AHT change

All channels (minutes): 0.0

Simulate coaching/process impact on AHT.

Scenario name

Volume %

Overrides slider if filled.

AHT delta (min)

Overrides slider if filled.

Scenario Result

Headcount

—

Monthly cost

—

Cost per contact

—

Note: Scenario does not overwrite base inputs until you click Save on Inputs.

Benchmarks (v1)

Simple benchmark flags for credibility. You can edit ranges per industry later.

Benchmark Ranges

Calls AHT

Min / Max (minutes)

Chats AHT

Min / Max (minutes)

Emails AHT

Min / Max (minutes)

Shrinkage

Min / Max (%)

Cost / Contact

Min / Max (USD)

Actions

Benchmarks only affect “flags” on Dashboard.

Benchmark Flags

Tip: For real “global benchmarks”, you’ll want a benchmark library by industry (Fintech, Utilities, Retail, etc.).

Confidence & use-cases

Use cases (recommended)

Budgeting / annual planning
Hiring plan & staffing strategy (blended vs dedicated)
Scenario planning (seasonality, policy changes, coaching)
Vendor benchmark comparisons (CPC / cost per ticket)

Not suitable (without deeper WFM)

Contractual SLA guarantees without interval modeling + queueing
Intraday scheduling / shrinkage planning beyond your assumptions
High-variance arrival patterns (events/incidents) without re-forecasting

This model is a pragmatic estimator. For production-grade staffing, use the Interval WFM Engine tab + validated arrival rates.

Year Planner

Save month-by-month snapshots (volumes, AHT, pooled headcount, costs) and view the full year. Stored locally in your browser.

Tip: Set Inputs for the month → click Recalculate → then Save month.

Annual Cost

—

Annual Contacts

—

Weighted CPC

—

Peak HC

—

Month	Calls	Chats	Emails	AHT C	AHT Ch	AHT E	Pooled HC	Monthly Cost	Blended CPC	Util%	Notes

Notes for selected month

Explainer: how the staffing & cost math works

This tab documents the exact calculation logic used in the tool (workload → capacity → staffing → cost allocation), so operations and finance teams can validate assumptions.

1) Workload (hours per month)

Calls workload (hrs) = (Calls Volume × Calls AHT minutes) ÷ 60

Chats workload (hrs) = (Chats Volume × Chat AHT minutes) ÷ 60

Emails workload (hrs) = (Emails Volume × Email AHT minutes) ÷ 60

Chat concurrency reduces effective workload:

Effective Chat Workload (hrs) = Chats Workload (hrs) ÷ Concurrency

Total effective workload (hrs) = Calls Workload + Effective Chat Workload + Emails Workload

2) Capacity (productive hours per agent per month)

Paid hours per agent = Shift Hours per Day × Work Days per Month

Shrinkage factor = 1 − Shrinkage%

Productive hours = Paid hours × Shrinkage factor

Occupancy cap limits usable productive time:

Capacity hours per agent = Productive hours × Occupancy Target

Note: occupancy should be higher for async work (email) than real-time voice, but must be calibrated to your operation.

3) Pooled (blended) staffing for multi-channel teams

In a blended environment, agents handle all channels, so we size a single pooled headcount.

Workload-based pooled agents =

Pooled Agents (workload) = Total effective workload (hrs) ÷ Capacity hours per agent

If Voice SLA staffing (Erlang C) is enabled in the WFM Engine, the pooled requirement is:

Pooled Agents (paid) = max( Voice SLA Agents (paid), Pooled Agents (workload) )

This prevents “we can clear workload” while still missing voice SLA at peak.

4) Voice SLA staffing (Erlang C) — WFM Engine

Offered load (Erlangs) a = (Calls per interval × AHT seconds) ÷ Interval seconds

Erlang C P(wait) computed from standard Erlang C formula (numerically stable log-sum).

Service level for target answer time T:

SL = 1 − P(wait) × exp( −(n − a) × (T ÷ AHT) )

The engine finds the minimum agents n that meets:

SL ≥ Target SLA AND Occupancy ≤ Utilization Cap

Paid agents = Base agents ÷ (1 − Shrinkage%)

5) Cost & cost-per-contact (pooled allocation)

Total monthly agent cost = Pooled Paid Headcount × Cost per Agent per Month

Because agents are blended, channel costs are allocated by workload share:

Channel Cost = Total Cost × (Channel Effective Workload ÷ Total Effective Workload)

Cost per contact (channel):

CPC(channel) = Channel Cost ÷ Channel Volume

If a channel’s AHT is very high (e.g., email threads) or volume is very low, CPC will naturally be higher even with correct pooling.

6) Definitions used in the tool

Shrinkage: paid time not available for handling contacts (leave, breaks, training, meetings, etc.).

Occupancy: portion of productive time we allow to be “busy” (protects burnout/quality).

Concurrency: average simultaneous chats per agent (capacity multiplier for chat).

AHT: average handling time (includes ACW if you want it to).

Paid headcount: scheduled/rostered agents after adding shrinkage.

Effective workload: workload after concurrency adjustments (chat) used for pooled allocation.

Settings

This MVP stores data locally (browser). Optional Supabase integration later.

Local Storage

Your inputs and benchmarks are saved in your browser automatically when you click Save.

Restore backup

Use backup/restore to move models between machines or share with clients.

Supabase (Optional – later)

If you want multi-user + history, we’ll connect a Supabase DB and add auth like your QA system.

What you’ll get

Multiple clients/workspaces
Scenario history & versions
Imports (Zendesk, Genesys, etc.)
Permissions: Ops vs Finance vs Leadership

Note: This file is MVP-first (local). If you want Supabase now, tell me and I’ll provide the schema + code hooks.