Duration: ~1 year from P&G pilot to scaling across 40+ customers (P&G pilot: 3 months, then expansion) Role: PM leading the initiative with 1 engineering manager + 4 backend engineers Company: Loadsmart (Digital freight brokerage platform) Outcome: Scaled email quoting channel from $3M → $17M annually | Conversion rates from 1% → 30-40%

The Problem (That Nobody Saw)

Loadsmart's digital brokerage was humming. The platform automated freight quoting for customers who integrated via API or used the web portal.

But there was a massive blind spot.

Many large enterprise customers—the ones with the highest volumes—were still sending quote requests via email. Or worse, they were using their own Transportation Management Systems (TMS) and we had to scrape their loading boards manually.

The process looked like this:

1. Customer sends email quote request (or we scrape their TMS portal)
2. Broker manually reads the request
3. Broker manually enters shipment details into Loadsmart
4. Broker manually generates a quote
5. Broker emails the quote back to the customer
6. Customer manually enters it into their system

Each step took time. And in freight brokerage, time is money.

The result? We could only handle a handful of these requests per day. Response times were slow. Win rates were abysmal (~1% conversion). And we were leaving millions on the table.

The Insight

I dug into our operational data using SQL queries to understand the actual bottleneck.

What I found: A massive portion of our enterprise revenue potential was stuck in manual processes.

The math was simple: