The carrier's largest customer existed in its systems as 47 separate records. One unified customer master rebuilt how JA Frate prices freight, credits sales, and measures risk.
INDUSTRY
Regional LTL and truckload freight
BUSINESS PROFILE
10,000+ shipments per month · 3-state Midwest service area · Family-owned since 1971 · 99%+ on-time rate
JA Frate is a family-owned, asset-based regional LTL and truckload carrier headquartered in Crystal Lake, Illinois. Founded in 1971 and led today by second-generation president Jill Jennings-Dinsmore, the carrier moves more than 10,000 shipments per month across Chicagoland, southern Wisconsin, and northern Indiana, with a published on-time rate above 99%.
IIn the spring of 2025, JA Frate's commercial operations team began what should have been a three-day customer profitability review. Ten weeks later, it was still running.
The reason was hidden in plain sight. JA Frate's largest customer by shipment volume, a regional furniture manufacturer, appeared across the carrier's systems as 47 separate party records with no shared identifier.
The transportation management system held shipper, consignee, and bill-to records for every load. A daily broker feed used different naming conventions, and electronic data interchange feeds from larger shippers used yet others. No customer master linked them, and the fallout showed up in three places.
Volume tier discounts trigger off cumulative shipment counts. But if one customer booked through three brokers and one direct channel under four different names, no tier ever fired and the customer paid published rates on every load.
When freight came through a broker channel, the direct-account sales rep got no commission credit even though the underlying customer was the same business.
JA Frate's true largest customer concentration was not the same customer the aged receivables report flagged. The carrier was carrying more exposure to its top accounts than its risk models showed.
JA Frate had tried twice before. A quarterly Excel VLOOKUP caught obvious cases but broke on punctuation. A desktop deduplication tool the operations team had picked up at a trucking conference had no concept of party role and sometimes linked unrelated entities.
By March 2025, the issue had reached the president, who asked for a permanent fix.
JA Frate ran a five-week vendor evaluation. The team built a gold-standard test set of 750 record pairs they had manually verified, and each candidate tool ran against the same 1.2 million-record extract from five years of TMS, broker, and EDI history.
JA Frate selected MatchLogic on the Server tier, with the Workflow Scheduler powering a nightly batch pipeline against the customer master. Five capabilities drove the selection.
MatchLogic's fuzzy matching engine, anchored by Jaro-Winkler distance, handled the routine variations in trucking party data without falsely linking unrelated companies. Every match decision carried a traceable rule the operations team could audit.
The platform let the team build rules that flagged candidate matches where the party-role mix (a parent showing up as shipper while a subsidiary appeared as consignee) suggested a possible business affiliate rather than the same entity.
The team defined three tiers: auto-link on tax ID plus a 90% name match; review queue on name plus state plus address similarity; periodic batch review for lower-confidence candidates.
The Workflow Scheduler ran a nightly job that ingested deltas from each source, scored candidate matches, and wrote unified entity IDs back to the master. The pipeline ran without supervision.
Two of JA Frate's largest customers had contractual restrictions on processing their shipping data off-site. Running MatchLogic on a server inside the carrier's data center avoided renegotiating those contracts before the project could start.
The seven-week rollout moved through three phases.
MatchLogic's solutions team ran the visual profiler against extracts from the TMS, broker staging tables, and EDI history. The profiler found that 34% of party name fields had inconsistent legal suffixes, 19% of addresses used non-USPS abbreviations, and 11% of records had truncated company names from a 32-character field limit.
The team built and validated the three-tier match definitions against the 750-pair gold-standard set. MatchLogic identified 47% more true relationships than the prior approach, with one false positive in the 750-pair validated set.
The Workflow Scheduler runs a nightly 02:00 job that pulls deltas, runs the cleansing pipeline, scores candidates, and writes unified IDs to the master. The morning profitability dashboard refreshes before the sales team logs in.
The 28-to-1 collapse came from the structure of freight data. The same end customer routinely appeared across the TMS as shipper, consignee, and bill-to on different loads, then again through any broker channel, often with subtle naming differences.
Joint-venture entities and parent-subsidiary distinctions accounted for most of the candidates that required human review. Every legacy party ID was preserved in the master record and linked to the unified entity ID, so downstream systems kept operating without re-coding while still reporting against the unified view.
Once unified IDs flowed through the data warehouse, the profitability analysis that originally took ten weeks completed in two days. The exercise found 312 customer accounts whose true shipment volume qualified for the next volume discount tier but were billed at the lower rate, and 88 accounts whose unified volume no longer justified the discounts they were receiving. The commercial team has prioritized the 312 retention conversations for the next pricing cycle, with the goal of converting recognized leakage into customer-perceived value through service tier upgrades rather than pure rate increases.
Once the master was unified, the actual concentration of revenue across JA Frate's top accounts became visible. The top 20 customers, previously reported as 28% of annual revenue, were actually 41% under the unified view.
The commercial team has restructured its key-account retention program around the corrected number, with dedicated relationship managers now assigned to the 20 unified accounts that previously appeared as more than 60 fragmented entities. The finance team has also re-baselined credit risk exposure at the unified entity level, which shifted the carrier's reinsurance and bad-debt reserve modeling for fiscal 2026.
Two analysts on Curtis Bauer's team had previously spent more than half of their working hours reconciling customer records across the three source systems by hand. That work no longer exists. The recovered capacity now supports a monthly customer health scorecard and a structured top-50 account review process the commercial leadership had wanted for years, but had not been able to staff before the master was unified.
JA Frate now prices, credits sales, and models risk against one version of its customer base instead of four. The nightly pipeline keeps the master current as new TMS, broker, and EDI activity arrives, so unification is a standing capability rather than a one-time cleanup. For a family-owned carrier competing against national networks, knowing exactly who its customers are has become a commercial advantage in its own right.
The difference was immediate. In the first month, we stopped creating about 1,400 duplicate patient records that our old system would have missed. That's 1,400 kids whose medical histories stayed intact instead of getting split across two profiles.
Founded in 2003, MatchLogic has matched over 2 billion records across 4,500+ enterprise installations, powering production-scale entity resolution pipelines.
MatchLogic combines deterministic and fuzzy matching with visual profiling and drag-and-drop cleansing workflows to identify duplicates and maintain trusted golden records.
Built for fintech, healthcare, insurance, and government teams, the platform provides REST API integrations, cross-platform deployment, and full auditability across every match operation.
.svg)
