One honest email a month about AI on the spot desk. Benchmark data included. No hype, no "future of freight" think pieces — just what we measured, what broke, and what it costs in dollars.
WE'D RATHER PUBLISH SLOW THAN PAD. EVERY PIECE BELOW IS LISTED WITH ITS REAL STATUS.
How fast 50 brokerages actually answer a quote
Abstract. Every brokerage believes it answers quotes "fast." Almost none has measured it. We are mystery-shopping 50 small and mid-size US brokerages with realistic spot-quote request emails — common dry-van corridors, plausible weights, ordinary shipper language — sent during business hours and after hours, then timing every reply to the minute. The result will be the first response-time benchmark a 5–30 seat brokerage can actually compare itself against.
Mystery-shop protocol. Each brokerage receives a small number of realistic quote-request emails over several weeks, from a plain shipper-style address. No phone calls, no recordings, no impersonation of any real company. Requests are ordinary enough that answering them is just… doing the job.
Consent-respecting. Results are published in aggregate only. No brokerage is named, ranked, or identifiable. Any brokerage that asks to be excluded is removed, full stop. We're brokers' corner, not gotcha journalism.
Latency buckets. Replies are timed into buckets: under 15 min · 15–60 min · 1–4 hr · 4–24 hr · next business day · never. Business-hours and after-hours (7 PM–6 AM local) windows are reported separately.
Median business-hours reply
—h —m▏
PLACEHOLDER · PUBLISHES WITH STUDY
After-hours reply rate
—%▏
PLACEHOLDER · PUBLISHES WITH STUDY
Requests never answered
—%▏
PLACEHOLDER · PUBLISHES WITH STUDY
NUMBERS STAY BLANK UNTIL THE FIELDWORK IS DONE. WE DON'T PRE-WRITE CONCLUSIONS.
Six pieces in the pipeline, listed with honest status tags. Filter by topic. Each publishes to the list first.
Every hour a quote sits, your win rate slides — so reps cut margin to claw loads back. That discount never shows up as a P&L line, but you pay it every week.
Add up the rep hours spent dialing carriers to ask "where's the truck?" Across a 10-seat shop it totals a full salary — for a job no one was hired to do.
FIGURE IS A MODEL — ASSUMPTIONS PUBLISHED WITH THE ARTICLE.
Most enterprise AI projects get walked back. That's not a reason to sit out — it's a free list of mistakes you get to skip. We go through the failure modes one by one.
INDUSTRY STAT — SOURCED AT PUBLICATION.
Day one, a rep approves every single send. Months later, routine lanes can earn auto-send — and the client controls every rung. Why we built it that way, on purpose.
Re-used MC numbers, callback area codes that don't match the truck, a "driver" replying from three states away. The pattern is in the check-in replies — if something is awake to read them.
What each TMS actually exposes for read and write-back, where the docs end and the workarounds begin, and what that means for your integration timeline.
DRAFT = WRITTEN, BEING FACT-CHECKED. QUEUED = OUTLINED, RESEARCH UNDERWAY. PUBLICATION DATES ARE ESTIMATES, NOT PROMISES.
| Event | Where | Topic | When | Status |
|---|---|---|---|---|
| Broker Carrier Summit | Dallas, TX | Response latency: what the benchmark found | 2026 | PLANNED — NOT YET BOOKED |
| TIA Capital Ideas Conference | TBD | A year of evals on a live spot desk | 2027 | PLANNED — NOT YET BOOKED |
| Your brokerage's ops meeting | Video, anywhere | 20-minute desk teardown, no pitch | ON REQUEST | OPEN INVITATION |
HONESTY LINE: WE HAVEN'T KEYNOTED ANYTHING YET. THESE ARE ROOMS WE INTEND TO EARN — LISTED HERE SO YOU CAN HOLD US TO IT.
Benchmark data. What broke on the desk and how we caught it. Unsubscribe is one click and we don't guilt you.
SUBSCRIBER COUNT: 0. YOU'D BE FIRST. WE'RE FINE SAYING THAT OUT LOUD.
The pilot is the real field note: 30 days on one inbox, every send gated by your reps, hard before/after numbers at the end. $2,500, credited against setup if you continue.