01

Case study · PRV-007

Forecast accuracy from 71% to 93% in one budget cycle

A driver-based model replaced 40 spreadsheets; the monthly pack now explains variance before it is asked about.

Client

Wholesale distribution group

$180M revenue · 9 depots

Industry

Distribution

Service

Performance & Analytics

Length

12 weeks

Year

2025

Statement of result · Forecast accuracy

Sample data

Before

71%

+22 pts

After

93%

Change

+22 pts

Relative

+31.0%

Timeline

12 weeks

7193M1M8

Source

Forecast vs actual, six months before and after go-live.

Stacked wooden pallets at a distribution site

Fig. · Client site — Wholesale distribution group

Sample image

Stacked wooden pallets at a distribution site

Fig. · Client site — Wholesale distribution group

Sample image

01 · The question

Why does the forecast miss by double digits every month — and why does it take two weeks to explain?

02 · What we found

—

—

Forty spreadsheets fed the forecast, and no two used the same definition of a sale.

Forty spreadsheets fed the forecast, and no two used the same definition of a sale.

—

—

Volume, price and mix were never separated, so variance could not be explained.

Volume, price and mix were never separated, so variance could not be explained.

—

—

Depots forecast in units, finance in dollars, and nobody reconciled the two.

Depots forecast in units, finance in dollars, and nobody reconciled the two.

03 · What we did

We built one driver-based model on volume, price and mix by depot, wrote a KPI dictionary with an owner for every measure, and redesigned the monthly pack around variance explanations.

Deliverable · Monthly performance pack

Sample · PRV-007

Monthly performance pack

Measure

Result

Result

Note

Forecast accuracy

Forecast accuracy

93%

93%

From 71%

From 71%

Pack issued

Pack issued

Day 3

Day 3

From day 11

From day 11

Spreadsheets retired

Spreadsheets retired

40

40

Into one model

Into one model

Wholesale distribution group

Performance & Analytics ·

2025

04 · What happened

Forecast accuracy rose from 71% to 93% within one budget cycle, and the monthly pack now ships on working day 3.

Before

71%

+22 pts

After

93%

Before

71%

After

93%

Timeline

12 weeks

Source

Forecast vs actual, six months before and after go-live.

Illustrative sample data. Provan, its people, clients, figures and sources are fictional and shown for demonstration only.

04

Next step

Facing something similar?

Talk to Anika.

Anika Shah leads Performance & Analytics. A 30-minute call is enough to tell whether a result like this is within reach for you.

Book a call

View Performance & Analytics

Your request

≤ 1 working day

A partner replies

We’ll respond within one working day.

04

Next step

Facing something similar?

Talk to Anika.

Anika Shah leads Performance & Analytics. A 30-minute call is enough to tell whether a result like this is within reach for you.

Book a call

View Performance & Analytics

Your request

≤ 1 working day

A partner replies

We’ll respond within one working day.

04

Next step

Facing something similar?

Talk to Anika.

Anika Shah leads Performance & Analytics. A 30-minute call is enough to tell whether a result like this is within reach for you.

Book a call

View Performance & Analytics

Your request

≤ 1 working day

A partner replies

We’ll respond within one working day.

Create a free website with Framer, the website builder loved by startups, designers and agencies.