Case Study: What an AI Assessment Found Inside a 60-Person Logistics Firm

The managing director of a 60-person freight forwarding company (anonymized; details shared with permission) came to us with a familiar brief: "Everyone says we should use AI. Nobody can tell me where."

Four weeks later the answer was specific: eleven opportunities, ranked, with the top three worth an estimated 290 staff hours per month. Here's how the picture emerged.

Week 1: Where the hours actually were

Interviews with five department leads produced a process inventory weighted by time. The surprise wasn't in operations — it was administration: quoting, booking confirmations, and customs documentation consumed more combined hours than dispatch itself.

Week 2: The shortlist takes shape

Three processes stood out on the impact-versus-complexity matrix:

• Quote preparation. Staff assembled quotes from carrier rate sheets, emails, and a legacy TMS — 45 minutes each, around 340 per month. AI-assisted drafting from structured rate data cut the estimate to 12 minutes.
• Customs document checks. Manual verification of shipping documents against customs requirements, with errors costing both fines and customer goodwill. Document extraction plus rule-based validation flagged discrepancies before submission.
• Booking status emails. Roughly 70 inbound "where is my shipment?" emails daily, each answered by hand despite the answer existing in the TMS.

Week 3: Numbers instead of impressions

Each shortlisted opportunity got a business case with the company's own figures — salaries, volumes, error costs. The quote automation alone projected a payback under four months. Readiness scoring also surfaced a blocker: carrier rate data lived in seventeen spreadsheet formats. Cleaning it became roadmap item zero — the unglamorous step that made everything else feasible.

Week 4: The roadmap

Quick wins first (status-email automation, live in three weeks), the quote system second, customs validation third — sequenced so each project's savings funded the next.

What the founder said afterwards

"I expected a technology report. What I got was the clearest picture of our own operations we've ever had." That's the part assessments get little credit for: the AI opportunities are the deliverable, but the operational X-ray is what leadership teams keep using.