FMCG beat productivity: bills cut, lines per call and secondary sales

Two salesmen, one number that hides the difference

An area sales manager (ASM) looking only at weekly secondary sales sees Ravi at ₹9,250 and Imran at ₹4,600 and concludes that Ravi is the better salesman. The scorecard below shows a different picture: Imran converts more of his beat, while Ravi sells a wider range into fewer outlets. They need different coaching, and the value figure alone cannot tell you which.

Where the data comes from

Beat plan (from the sales force app or a planning sheet): salesman, beat_id, visit_date, outlets_scheduled, and ideally one row per scheduled outlet.

Distributor invoice lines (from the distributor management system, DMS): bill_no, bill_date, salesman, outlet_id, sku, qty, gross_value, scheme_discount, net_value.

This is secondary sales — distributor to retailer. Primary sales (company to distributor) are a different table and should never be mixed into a beat scorecard; a distributor stocking up before a price increase says nothing about outlet demand.

Metric definitions

Scorecard code (synthetic week, scaled down)

The beats below are shortened to two or three outlets a day so the example stays readable.

import pandas as pd

# Synthetic: one week of beat plans and distributor (DMS) invoice lines
plan = pd.DataFrame({
    "salesman": ["Ravi"]*3 + ["Imran"]*3,
    "visit_date": ["2026-09-21","2026-09-22","2026-09-23"]*2,
    "outlets_scheduled": [3, 2, 3, 3, 3, 3],
})
lines = pd.DataFrame({
    "salesman": ["Ravi"]*9 + ["Imran"]*9,
    "visit_date": ["2026-09-21"]*4 + ["2026-09-22"]*3 + ["2026-09-23"]*2
                + ["2026-09-21"]*3 + ["2026-09-22"]*3 + ["2026-09-23"]*3,
    "outlet_id": ["O1","O1","O2","O2","O3","O3","O4","O5","O5",
                  "K1","K2","K3","K4","K4","K5","K6","K7","K7"],
    "bill_no":   ["B1","B1","B2","B3","B4","B4","B5","B6","B6",
                  "C1","C2","C3","C4","C4b","C5","C6","C7","C8"],
    "sku": ["Soap75","Paste100","Soap75","Shampoo","Soap75","Paste100","Soap75","Paste100","Shampoo",
            "Soap75","Soap75","Soap75","Soap75","Paste100","Soap75","Soap75","Soap75","FreeSachet"],
    "net_value": [1200, 800, 900, 1500, 700, 650, 1000, 1400, 1100,
                  600, 550, 500, 650, 400, 700, 580, 620, 0],
})
lines = lines[lines["net_value"] > 0]          # scheme freebies are not selling lines

# A productive call = an outlet billed on a day, however many invoices it produced
calls = (lines.groupby(["salesman","visit_date","outlet_id"])
              .agg(sku_lines=("sku","nunique"), value=("net_value","sum")).reset_index())

scorecard = (calls.groupby("salesman")
                  .agg(productive_calls=("outlet_id","size"),
                       lines=("sku_lines","sum"), value=("value","sum"))
             .join(plan.groupby("salesman")["outlets_scheduled"].sum()))
scorecard["productivity_pct"] = (100 * scorecard["productive_calls"] / scorecard["outlets_scheduled"]).round(1)
scorecard["lines_per_call"] = (scorecard["lines"] / scorecard["productive_calls"]).round(2)
scorecard["avg_value_per_call"] = (scorecard["value"] / scorecard["productive_calls"]).round(0)
print(scorecard[["outlets_scheduled","productive_calls","productivity_pct","lines_per_call","avg_value_per_call","value"]].to_string())

naive = lines.groupby("salesman")["bill_no"].nunique()
print("\nInvoices counted as calls (the wrong way):", naive.to_dict())

          outlets_scheduled  productive_calls  productivity_pct  lines_per_call  avg_value_per_call  value
salesman
Imran                     9                 7              77.8            1.14               657.0   4600
Ravi                      8                 5              62.5            1.80              1850.0   9250

Invoices counted as calls (the wrong way): {'Imran': 8, 'Ravi': 6}

Imran billed 78% of his scheduled outlets but almost always a single SKU — 1.14 lines per call, and Soap75 in every one of his calls. Ravi billed only 62.5% of his outlets but sold close to two SKUs per call at nearly three times the value.

What each pattern tells the ASM to do

Mistakes that distort the scorecard

Counting invoices as calls. Imran's outlet K4 received two bills on 22 September. Counting invoices gives Imran 8 calls and Ravi 6, and overstates productivity. Use distinct outlet-date pairs.

Free-goods lines. The zero-value FreeSachet line is a scheme freebie. Left in, it would raise lines per call for a salesman who sold nothing extra. The code drops lines with non-positive net value; adjust if your DMS records schemes differently.

Returns and credit notes. Expiry returns taken back on the next visit arrive as negative lines or separate credit notes. Net them in the value metric for the month, but do not let a return day make a productive call look unproductive.

Unplanned outlets. A salesman who bills an outlet not on the day's beat plan raises productive calls without raising the denominator. Decide whether off-beat calls count and apply it to everyone.

SKU codes that are the same product. A soap sold in a four-pack and singly may carry two SKU codes. Lines per call then rewards pack variety, not range. Map to a product family if range selling is the intent.

Putting it in front of the ASM

ASMs read scorecards on phones between market visits. One row per salesman, the four metrics, and one arrow per metric against the previous four weeks is enough. Put secondary value last, not first, so the conversation starts with how the value was earned. Pair each salesman with the matching row of the coaching table rather than a ranking; a league table of value per salesman rewards whoever has the richest beat.

Scorecards like this are standard exercises in the Data Analyst programme. For people already in sales roles, sales executive to sales analyst describes how to turn beat data you already handle into a portfolio project.