_1777711377206.png)
Action Tesa: how a ₹2,800 Cr panel maker cut freight 14%
A teardown of the 90-day rollout — what changed in the indent process, the auction floor, and the settlement workflow.
Action Tesa is one of the largest decorative-panel manufacturers in the region. Six plants. ~₹2,800 Cr in annual freight spend across primary and secondary movement. Until ninety days ago, they ran the entire freight operation on Excel, three regional ERPs, and a 240-person WhatsApp group. By the end of the rollout, freight cost was down 14% and dispute resolution was down from three weeks to same-day.
This is the teardown. What changed, what didn't, and the bits that almost broke the project.
Where they started
Three things were broken at the same time. Each one made the others worse.
Indent allocation was a phone tree
Each plant's dispatcher kept a personal list of "preferred carriers" in a notebook. When a load came up, they called the top 3. If those didn't pick up, they called another 3. Median time-to-allocation: 47 minutes. Carriers who weren't on anyone's notebook never got loads, even when their rates were better.
Rates drifted upward, quietly
Without a competitive auction, the same 14 carriers won 78% of the loads. They knew they were not being benchmarked, and their quotes drifted up about 8% over 18 months relative to market.
Settlement was a 21-day reconciliation war
Every freight invoice required matching against an LR (lorry receipt), a POD (proof of delivery), and the original indent. With four document sources and no shared identifier, the finance team spent three weeks per cycle reconciling and disputing.
"We weren't losing on cost because we were bad negotiators. We were losing because we were negotiating against ourselves, six plants apart."
What we changed (in order)
Days 1–14 — Indent into one place
Every plant's indent flow now lands in the same Traqo workspace. Same template, same fields, same SLAs. Plants kept their existing ERP — we ingested via webhook. The dispatchers' notebooks went into a shared "preferred carrier" list with version history.
Days 15–45 — Auction floor live
Spot loads went to a WhatsApp-based auction (see our other post). Contracted lanes stayed contracted, but carriers were benchmarked monthly against winning bids on similar lanes. The benchmarking conversation alone shaved ~6% off contracted rates in the first month.
Days 46–75 — Settlement reconciliation
POD capture moved to WhatsApp + OCR. Invoice ingestion moved to email + parsing. A three-way match (indent ↔ LR ↔ POD ↔ invoice) ran nightly. Disputes that used to wait 21 days were flagged within 24 hours of POD capture.
Days 76–90 — Carrier scorecards
With all the data in one place, monthly carrier scorecards became automatic. Two underperforming carriers were dropped. Three new ones were onboarded from the auction data — they had won spot bids consistently and earned contracted volume.
What actually moved the numbers
The auction line is the biggest. But the auction alone wouldn't have stuck without the scorecard loop in days 76–90 — without that, contracted carriers would have just slowly drifted prices back up. The two pieces only work together.
Where it goes next
Action Tesa is now rolling the same workflow into yard operations and inbound. The playbook compresses each time — the second plant onboarded in 11 days, the third in 6. By the sixth, the dispatchers were training each other and we were on a back seat. That's the bit we're proudest of.
- 1The biggest savings came from price discovery, not negotiation tactics. Carriers behaved differently when they knew they were being benchmarked.
- 2Settlement automation only works if POD capture is upstream of it. Don't try to fix billing without first fixing the document.
- 3When a contracted carrier complains about the auction, your response defines whether the project survives the next quarter.
Writes about how the world's largest shippers actually run freight — the real workflows, the stuff vendors don't put in slides.
More from the team
After 14 customer interviews, every transporter chose WhatsApp over a slick web portal. Here's the data — and what it taught us about adoption in emerging-market logistics.
Our OCR pipeline turns smudged, curled, sometimes wet paper PODs into structured JSON in 30 seconds. Here's the architecture — and the failure modes nobody warns you about.
Most TMS dashboards drown ops teams in red. We rebuilt the control tower around the five decisions a dispatcher actually makes before lunch.