SOPHIA

Reference — Building envelope systems manufacturer × Sophia

6,000 mails a month.
Handled, filed by no one

A German family business manufactures building elements for roof and façade and sells exclusively through specialist dealers. Half the revenue arrives as email.

Half orders, half inquiries, with hard seasonal peaks. Sophia handles both: capture, check against the ERP, write back, reply.

Data plate

Customer

German family business, manufacturer of building elements for roof and façade

Size

Over 100 employees, two production sites

Sales

Exclusively through specialist dealers, strictly dealer-loyal

Production

Made to order, configured by dimension, material, coating and colour

Systems

abas ERP, inbox (Microsoft Exchange, on premise)

< 15 min

response time, even in peak season

9 of 10

orders without manual work

≥ 20 min

saved per case

~6,000

cases a month

In three lines — where it jammed, what we did, what runs now

Before

Every line item by hand

Dealers order on their own forms. In peak season, response time decides who gets the order — and that is exactly when the queue is longest.

Intervention

Sophia sits on inbox and ERP

Orders are captured and confirmed, inquiries priced and answered, both checked against the ERP.

Now

Only the exceptions

Orders and inquiries are handled in minutes. Inside sales only decides where it jams — and has time for what wins the order: the dealer on the phone.

Façade grid of vertical metal slats, abstract crop
Automated production line in a factory hall

01

Reference 01 — The company

Configured,
not catalogued

The company rolls and folds metal profiles for roof, wall and façade and supplies the whole system around them, from fixings to drainage. Nothing is sold past the dealers — everything is sold through them.

Length, material, thickness, coating, colour: every combination is its own item and its own price. What is flexibility for the customer is a calculation per mail for inside sales.

02

Reference 02 — Before and now

The inbox
was the sales team

Before

PDF line items retyped

Dozens of Excel sheets and workarounds

Every inside-sales rep doing their own thing

No cross-selling

Signals and leads going unnoticed

Pricing only through a few heads

Blocked items caught late

Backlog in peak season

Second site set up from scratch

Now

→From the dealer form, the order is created in the ERP, line by line.

→The logic from sheets and workarounds lives versioned in the platform — one way, the same for everyone.

→Every case follows the same standard, every decision is traceable.

→Cross-selling becomes a suggestion: matching products come attached to the quote.

→Signals and leads from inbox and ERP land on the table as opportunities — more revenue from the base.

→The company’s price models live in the platform as versioned logic.

→Block flags and impossible combinations surface before the quote.

→Volume is a setting, not a staffing question.

→One setup is carried over to the second entity, never rebuilt.

03

Reference 03 — At work

Done,
not just detected

W/01

Every mail gets classified

Order or inquiry, attachment read, line items extracted, customer resolved against the master. Only then does the right process start.

W/02

Order captured and confirmed

From the dealer form, the order is created in the ERP, with resolved item numbers and a checked price. The confirmation goes back as a draft.

W/03

Inquiry priced

The configuration is resolved. What is ambiguous goes to inside sales as a clarification, never as a guessed price.

W/04

Not feasible is not quoted

Block flags, missing price lines and impossible combinations are checked before they reach the dealer.

W/05

Comparison over trust

For every case the extraction sits next to the original document, and before every reply stands a human approval.

W/06

Sales researches in chat

The first use case came from the customer themselves: finding specialist dealers in a postcode area, matched against their own existing customer base.

Results — measured, not claimed

No-touch rate

9 of 10 orders captured without manual work

Cycle time

under 15 minutes, even in peak season — previously up to 2 days

Error rate

0 silent errors — deviations escalate automatically

Completed transactions

~6,000 cases a month — orders and inquiries

Inquiries answered autonomously

around 70 percent

Relief per case

at least 20 minutes saved

The real win is in none of these lines. It is the time inside sales has again for the dealer.

And the freed-up time keeps working: it goes into building the next agents on Sophia — they read signals from the market, customer news and ordering behaviour and turn them into inquiries and orders. More revenue from new and existing customers. Signal-to-Cash →

Measured from the first case, against a shared baseline, reported monthly.

Your inside sales
can do this too