How a 200-truck brokerage cut quote response time by 20 minutes in 90 days

The brokerage

A privately held Midwest U.S. freight brokerage, roughly 200 trucks under contract, with a 14-person sales desk handling inbound quote requests for dry van and reefer moves across the lower 48. They came to us with a familiar problem: inbound volume was outpacing rep capacity, and the team was losing loads on response time alone.

The pre-pilot baseline

Before deployment, we ran a one-week passive baseline on their inbox. The numbers were in line with what we see across most mid-market brokerages:

• Median time from quote request received to reply sent: 23 minutes.
• Roughly 18% of quote requests went unanswered for more than two hours.
• Win rate on responded quotes: 41%.
• Average broker hand-time per quote (including context lookup and draft): approximately 6 minutes.

What we deployed

A standard FreightSurge pilot configuration: the Outlook add-in for every broker on the desk, retrieval over their last 24 months of historical loads, integration with their existing TMS for live capacity, and read-only access to their DAT account for spot benchmarking. No data left their environment. Setup took less than a day.

The 90-day results

• Median quote response time: 23 minutes → under 3 minutes (20 minutes faster).
• Quote response rate inside two hours: 82% → 96%.
• Win rate on responded quotes: 41% → 56% (+15 points).
• Broker time saved per day: approximately 2 hours per rep.
• Drafts that brokers sent without edits: 64% (a leading indicator we watch closely).

What surprised the leadership team

Coachable workflows

Because every draft is logged with the AI suggestion, broker edits, and final outgoing reply, the sales manager could see — for the first time — exactly where individual reps were diverging from house pricing logic. Coaching conversations went from anecdotal to data-driven inside two weeks.

Customer-tier visibility

The plug-in surfaces a customer's tier, lifetime revenue, and recent loads on every inbound quote. Brokers stopped accidentally giving spot pricing to repeat customers, and started recognizing at-risk accounts based on volume drop signals the system flagged.

A pipeline view that actually works

The leadership team got a live read on quote volume by lane, win rate by rep, and average margin by customer. None of this required a separate BI project — it was a byproduct of every quote being structured at draft time.

What we would do differently

The first three weeks of pilot, brokers were over-editing AI drafts because they did not yet trust the lane history retrieval. We addressed it by surfacing the specific past loads the suggested rate was based on directly above the draft. Trust is the actual feature, and showing the work is how you build it.

If you are running a brokerage at this scale and want to see what a quarter of measured improvement looks like on your real inbox, talk to us — we will set up a passive baseline week first so the numbers are yours, not ours.