Repligen Customer story
Run seven acquired ERPs as one compliance system.
Every acquisition brought its own ERP. One data standard and one classification layer now read across all seven, so reports and audits start in one place.
- ERP systems unified71One classification layer on top
- Quarterly reporting80+ h14 hPer cycle, 82% fewer hours
- Annual USITC HTS audit9 wk11 dAcross 16,000+ SKUs
- Annualized savings~$1.2MLabor and risk reduction
Two more acquisitions have since connected in about six weeks each.
Results in the first year. ERP systems unified: 7 to 1 (One classification layer on top); Quarterly reporting: 80+ h to 14 h (Per cycle, 82% fewer hours); Annual USITC HTS audit: 9 wk to 11 d (Across 16,000+ SKUs); Annualized savings: ~$1.2M (Labor and risk reduction).
What happened at Repligen?
One data standard came first, then a unified layer and a classification service on top of it.
Repligen is a global bioprocessing manufacturer that buys one to two manufacturers a year. Each one arrived with its own ERP, so seven ERPs held parts of the truth about 16,000+ imported SKUs, and every report meant reconciling seven schemas. GingerControl first designed one Trade Compliance Data Standard with Repligen’s trade team, then built a unified data layer and a classification service that watches USITC changes. The build took nine months and rollout three. Quarterly reporting fell from 80+ hours to 14, and the annual HTS audit from about nine weeks to eleven days.
RepligenPublished June 6, 2026- Engagement
- Global classification + ERP integration
- Industry
- Bioprocessing manufacturing, global
- Footprint
- 7 ERPs, 16,000+ SKUs
- Build window
- 9 months
- Timeline
- About a year: 9 to build, 3 to roll out
Case file for Repligen: Engagement Global classification + ERP integration; Industry Bioprocessing manufacturing, global; Footprint 7 ERPs, 16,000+ SKUs; Build window 9 months; Timeline About a year: 9 to build, 3 to roll out.
Why did one report need seven systems?
Each acquired company kept its own ERP, and each ERP held part of the truth about what Repligen sold and how it was classified.
- Repligen serves biopharma and CDMO customers in more than thirty countries and buys one to two manufacturers a year.
- Some ERPs were built in-house years ago; others were packages configured by an integrator long gone.
- The brief: not another reporting tool on the side, but one central system that reads every ERP and absorbs the next acquisition.
- Acquired ERPCountry of originFree text
- Acquired ERPCountry of originCoded reference
- Acquired ERPCountry of originDerived value
- Acquired ERPCountry of originOwn convention
- Acquired ERPCountry of originOwn convention
- Acquired ERPCountry of originOwn convention
- Acquired ERPCountry of originOwn convention
Diagram of seven acquired ERPs, each storing country of origin its own way (free text, a coded reference, a derived value and other conventions), feeding one quarterly compliance report that took 80+ hours of assembly per cycle by hand.
Where did the trade team’s hours go?
Into report assembly, a nine-week HTS update every year, and data cleanup before every audit.
- Cross-ERP reporting by hand80+ hours per cycle pulling seven extracts into a spreadsheet and chasing columns that should have agreed.Unified data layerA thin connector per ERP writes into one compliance schema. Quarterly reports now generate from one place.
- The annual USITC HTS update16,000 SKUs re-checked by hand against every split, merge and new subheading, about nine weeks a year.Classification serviceWatches USITC changes, compares them with the SKU-to-HTS map and lists only the SKUs whose code may be stale.
- No shared data standardCountry of origin was free text in one ERP, a coded reference in another, a derived value in a third.Trade Compliance Data StandardOne schema, required fields, validation rules and an owner for each field. Bad records show up when written, not at audit.
Table of three pain points and their fixes: cross-ERP reporting by hand fixed by a unified data layer, the annual USITC HTS update fixed by a classification service that lists possibly stale SKUs, and no shared data standard fixed by the Trade Compliance Data Standard.
What had to be unified first?
Six data domains, one definition each, every field with a named owner, frozen before any build began.
- The scope was agreed in the first month and frozen before the build.
- Every new acquisition’s ERP is onboarded against the same standard.
- The classification layer reads the standard, not the quirks of each ERP underneath.
- Product master
- SKU naming, description structure, unit of measure, bill of materials, net and gross weight, dimensions
- Classification
- US HTS, ECCN, Schedule B, foreign HTS for major destinations, prior CBP rulings per SKU
- Country of origin
- ISO country code, manufacturing site, substantial transformation evidence, FTA eligibility flags
- Party master
- Suppliers, manufacturers and customers under one canonical ID, with EORI, MID and TIN
- Valuation
- Transaction value, assists, royalties and license fees, selling commissions, freight and insurance allocation
- Programs and documents
- Section 232, 301 and 122 applicability, ADD/CVD scope, GSP and preference indicators, CI, PL, BOL and COO rules
The Trade Compliance Data Standard with six domains: product master, classification, country of origin, party master, valuation, and programs and documents, each with the fields it defines.
What did it cost?
About $1.2M, fixed scope, paid against five milestones. The yearly savings matched it in the first year.
- Pricing model
- Fixed scope, milestones
- Paid per accepted milestone, not by the hour
- Investment
- ~$1.2M
- One-time: seven-ERP rollout + data standard
- Payback
- ~1 year
- Annualized savings about equal to the build cost
Engagement budget: Pricing model Fixed scope, milestones, Paid per accepted milestone, not by the hour; Investment ~$1.2M, One-time: seven-ERP rollout + data standard; Payback ~1 year, Annualized savings about equal to the build cost.
- Canonical data standard, frozen, with field owners
- Connectors live for the first ERPs
- One quarterly compliance report generated from the layer
- Classification service watching USITC changes
- The remaining ERP connectors
- Self-serve dashboards and ad-hoc analytics
- Destinations and programs beyond the first scope
MVP scope. In the MVP: Canonical data standard, frozen, with field owners; Connectors live for the first ERPs; One quarterly compliance report generated from the layer; Classification service watching USITC changes. Phase two: The remaining ERP connectors; Self-serve dashboards and ad-hoc analytics; Destinations and programs beyond the first scope.
- Repligen knew the number before the build began, and each payment tracked a milestone it could accept.
- Each new acquisition now onboards through the same layer in about six weeks instead of six months.
How long did it take?
About a year: nine months of build and three of rollout, with an acceptance test Repligen owned at every milestone.
- The data standard had to be signed by its field owners.
- The MVP had to produce a quarterly report that matched a hand-built one.
- The classification service had to find the right stale SKUs against a known USITC change.
- ERPs moved over in tranches, and the old process ran until each tranche passed.
- Months 1–3Assessment and data standardA quarter of past reports walked through, then the standard: six domains, an owner on every field.Deliverable Signed data standard; fixed scope and price confirmed.
- Months 4–5Unified layer and first connectorsThe canonical layer stood up, with the first ERPs writing into one schema.Deliverable First ERPs flowing into the unified layer.
- Months 6–7One report, end to end (MVP)A full quarterly report from the layer instead of seven spreadsheets.Deliverable Quarterly report produced from one source.
- Months 8–9Classification service and remaining ERPsThe USITC-watching service connected, then the remaining ERPs rolled on.Deliverable All seven ERPs unified, classification service live.
- Months 10–12Testing, training and evaluationEnd-to-end testing across all seven ERPs, team training, a check against the baseline.Deliverable Team trained, results validated, signed off.
Milestones: Months 1–3, Assessment and data standard; Months 4–5, Unified layer and first connectors; Months 6–7, One report, end to end (MVP); Months 8–9, Classification service and remaining ERPs; Months 10–12, Testing, training and evaluation.
What changed in the first year?
The trade team stopped assembling reports and started reviewing them.
Hours to assemble and analyze one cross-ERP compliance report, before and twelve months after go-live. Bar length follows total hours.
- Extract, normalize, reconcile by hand
- Generate from the layer, spot check
- Analysis
Bar chart of hours per quarterly compliance report. Before: 80 hours, 74 of them extracting, normalizing and reconciling by hand and 6 of analysis. After: 14 hours, 3 generating from the layer and spot checking and 11 of analysis.
Time to audit 16,000+ SKUs against the year’s USITC HTS revisions. Same scope, same standing team.
Bar chart of the annual USITC HTS audit: about 9 weeks before, 11 days after.
- The annual HTS audit finished inside two weeks with the standing team, with no borrowed analysts.
- Two more acquisitions connected in about six weeks each, against the six months it used to take.
- The reconciliation was paid once, at the data layer, so the team’s hours go to classification judgment.
| Metric | Before | After | Change |
|---|---|---|---|
| Quarterly compliance reporting | 80+ hours per cycle | 14 hours per cycle | 82% reduction |
| Annual USITC HTS audit | ~9 weeks | 11 days | 83% reduction |
| New acquisition ERP onboarding | ~6 months | ~6 weeks | 75% reduction |
| SKUs under classification monitoring | Sampled, manual | 16,000+, automated | Full coverage |
| Annualized labor and risk savings | Baseline | ~$1.2M | Captured year one |
Before and after: Quarterly compliance reporting, 80+ hours per cycle before, 14 hours per cycle after (82% reduction); Annual USITC HTS audit, ~9 weeks before, 11 days after (83% reduction); New acquisition ERP onboarding, ~6 months before, ~6 weeks after (75% reduction); SKUs under classification monitoring, Sampled, manual before, 16,000+, automated after (Full coverage); Annualized labor and risk savings, Baseline before, ~$1.2M after (Captured year one).