Managing 200 Subcontractor COIs
What Breaks When You Scale

The Math That Breaks the Spreadsheet

A single subcontractor requires a certificate of insurance covering four to six distinct policy lines: General Liability, Workers' Compensation, Commercial Auto, Umbrella/Excess Liability, and often Professional Liability and Builder's Risk. Each of those policies carries its own carrier name, policy number, coverage limits, effective date, and expiration date — and those expiration dates are rarely aligned. A framing sub's GL may renew every January 1, but their auto policy runs March-to-March, and their umbrella renews in September.

At 200 subcontractors with an average of four tracked policy lines each, a general contractor is not managing 200 COIs. The actual load is 800 policy-level data points that update on independent annual cycles — roughly 15 new or changed entries every week, every week, in perpetuity. Each entry requires someone to open a PDF, read 10 to 15 fields across multiple certificate sections, verify that coverage limits meet contractual requirements, check for additional insured endorsements, note any discrepancies, and type the result into a tracking system.

At five minutes per certificate — the conservative estimate most construction firms cite — that is 67 hours of data entry per year just to keep the spreadsheet current. And that number assumes every COI arrives correct on the first submission. In practice, 45-55% of initial COI submissions contain errors or deficiencies that require follow-up with the subcontractor or their agent — each cycle adds days of back-and-forth before the data entry can even begin.

Three Breaking Points Every Growing GC Hits

The spreadsheet does not fail all at once. It fails in three distinct stages — and at each stage, the nature of the failure changes. Cross-referenced data from multiple COI platforms makes the pattern clear.

Breaking Point One: ~25 Subcontractors — "One Person Can Still Hold It Together"

At 25 subcontractors, the spreadsheet is strained but functional. A single project coordinator can track who has what coverage, remember which subs are late with renewals, and personally send follow-up emails. The weekly time commitment is 3-5 hours. The key risk at this stage is not scale — it is concentration. One person holds the entire compliance picture in their head, and if they leave, the institutional knowledge leaves with them. The NAHB reports that the average new single-family home uses 24 different subcontractors — many residential GCs sit precisely at this threshold without realizing it.

Breaking Point Two: ~75 Subcontractors — "Errors Compound Faster Than You Can Catch Them"

At 75 subcontractors across three or more active projects, the spreadsheet passes a structural limit: one person can no longer hold the mental model of who has coverage, when it expires, and what endorsements are required per project. Renewals are missed. A sub's GL expires mid-project and nobody notices until the next audit. The coordinator stops proactively verifying coverage limits and starts simply recording whatever certificate arrives — a shift from compliance management to clerical filing.

This is the range — between 25 and 150 subcontractors — that Billy identifies as the zone where most COI-related claims and audit failures originate. The organization still feels small enough that manual processes seem acceptable, but the data volume has long since crossed the line where a single human can maintain accuracy.

Breaking Point Three: ~200 Subcontractors — "The System Is the Liability"

At 200 subcontractors, manual COI tracking is no longer a productivity problem. It is a systemic risk. A Phoenix general contractor discovered this when a subcontractor's workers' compensation policy had expired 47 days before an OSHA inspection triggered by a minor injury on site. The resulting liability claim, regulatory fines, and project delays cost $2.3 million — more than five years of the sub's policy premiums.

The critical difference at 200 subs is not volume alone. It is that the GC has multiple active projects with different contractual insurance requirements per project, different endorsement requirements per project type, and subcontractors who work across multiple projects with coverage that may be adequate for one but insufficient for another. The spreadsheet was designed to record what a single certificate says on a single date. It was never designed to track whether that coverage is sufficient across a portfolio of active projects with shifting requirements.

The Annual COI Lifecycle at 200 Subcontractors

Understanding why manual tracking breaks requires understanding what the COI calendar actually looks like at scale. The lifecycle has five phases that repeat continuously — not once a year, but in overlapping waves across the subcontractor roster.

Phase 1: Collection (ongoing). Before a sub begins work on a new project, the GC requests a current COI. The sub contacts their insurance agent. The agent issues the certificate — sometimes same day, sometimes a week later, sometimes after back-and-forth when the agent realizes the sub's policy limits don't meet the contract requirements. At 200 subs, this collection phase never ends. Somewhere between 3 and 8 new certificates arrive every week.

Phase 2: Review (5-10 minutes per COI). Someone opens the PDF and checks: Is the named insured the correct legal entity? Do coverage limits meet contract minimums? Is the general contractor listed as additional insured? Are waiver of subrogation and primary/noncontributory endorsements included? Does the certificate list all required coverage types? Are the dates valid? The review takes longer when something is wrong — and nearly half of all initial COI submissions require corrections.

Phase 3: Data Entry (5-7 minutes per COI). The reviewed certificate's fields — carrier, policy number, coverage types with limits, effective date, expiration date, additional insured status, endorsements — are typed into the tracking system. With 10 to 15 fields per COI across multiple policy sections, this is repetitive, error-prone manual work. At 200 subs, it consumes roughly 70 hours per year — and that is assuming every certificate arrives complete and correct.

Phase 4: Monitoring (continuous). Expiration dates are tracked. At 200 subs with 4+ policy lines each, there are 800+ dates to monitor, spread across every month of the year. Best practice is 60-day advance notice with follow-ups at 30, 15, and 7 days before expiry. Without automated reminders, this monitoring either does not happen or consumes an employee's full attention.

Phase 5: Renewal (cyclical). Before a certificate expires, the GC contacts the sub. The sub contacts their agent. The agent issues a new certificate. The review-and-entry cycle begins again. At 200 subs, this cycle is not an annual event — it is a weekly operating rhythm.

The spreadsheet can handle Phase 1 and Phase 3 at small volumes. It cannot handle Phase 2 (no way to enforce contract requirements automatically), Phase 4 (no automated reminders), or the compounding effect of all five phases running simultaneously at 200-sub scale.

COI Tracking Software Solves Monitoring — Not Data Entry

The construction industry's answer to COI tracking at scale has been dedicated compliance platforms: myCOI with its insurance-industry compliance logic, Billy with Procore Side Panel integration and construction-specific endorsement review, Jones with enterprise-grade workflows across a network of pre-verified vendors, BCS with AI-powered certificate analysis and a 78,000+ vendor database. These platforms handle Phases 4 and 5 of the lifecycle effectively — automated expiration alerts, renewal workflows, compliance dashboards.

The gap is Phases 2 and 3: review and data entry. COI tracking platforms monitor certificates once they are in the system. But someone still needs to open each incoming COI PDF, read the ACORD 25 form's multiple policy sections, extract the relevant fields, and type them into the platform. The tracking software eliminates the monitoring bottleneck. It does not eliminate the data extraction bottleneck — and at 200 subs, that bottleneck is processing approximately 15 new or updated certificates per week, each requiring 10-15 typed data points.

For a general contractor evaluating compliance options at 200 subcontractors, the question is not "should I buy COI tracking software?" The question is two-part: (1) how do I get certificate data into whatever system I use, and (2) how do I monitor it once it is there? The industry conversation focuses almost entirely on the second question.

The cost dimension reinforces this gap. COI tracking software typically ranges from $3 to $80 per vendor per year depending on features and service model. At 200 subcontractors, that is a $600 to $16,000 annual line item — before factoring in the internal labor cost of reading certificates and entering data into the platform. The platform automates the alerting. It does not automate the reading.

Extraction vs. Tracking: Why the Distinction Matters at Scale

This distinction — between data extraction (reading fields from a PDF) and data tracking (monitoring expiration dates and compliance status) — is the organizing principle for COI management at scale. Both functions are necessary. At 200 subcontractors, you need both. But treating them as one problem — and buying a single solution — is how GCs end up with a $500/month platform that still requires someone to spend Wednesday afternoons typing policy numbers.

Data extraction is a document-reading problem. You have 200 PDF certificates that arrived from 200 different insurance agents, each formatted slightly differently, each containing four to six policy sections, each needing the same set of fields pulled out: Subcontractor Name, Carrier, Policy Number, Coverage Type, Limits, Effective Date, Expiration Date, Additional Insured status. The ACORD 25 form provides a standard layout, but agents frequently use proprietary letterhead or embed the certificate in a cover email PDF, and the standard fields — effective date, expiration date, policy number — are the same data points regardless of layout.

Data tracking is a calendar-and-compliance problem. You need to know which certificates expire next week, which ones are missing endorsements, which subcontractors are out of compliance across which projects. This is what COI tracking platforms do.

The extraction step has historically been the harder problem to automate because it requires reading unstructured documents — a task that has only recently become reliable with AI. The tracking step has been solvable with software for decades (it is essentially a database with date-based alerts). Yet the industry's solution bias runs the other direction: buy the tracking platform, and figure out the extraction later.

A Practical COI Workflow for 200 Subcontractors

The solution to COI management at scale is not replacing the spreadsheet with a platform — or replacing the platform with a spreadsheet. It is separating the two problems and solving each with the right tool. Here is a workflow designed for a general contractor managing approximately 200 subcontractors across multiple active projects, using existing tools plus one extraction step:

This workflow separates extraction from tracking. The extraction step — reading 200 ACORD forms and typing 2,000+ data points — becomes a batch operation that takes minutes instead of days. The tracking step — monitoring 800+ expiration dates — can be handled by the existing spreadsheet (with conditional formatting and calendar reminders) or by a COI tracking platform, depending on budget and scale.

What makes this practical at 200 subs is the column-name extraction approach: instead of training a template for each insurer's COI layout, you define the fields you need once — Policy Number, Carrier, Effective Date, Expiration Date, GL Each Occurrence Limit, WC Statutory Limits, Additional Insured — and the AI locates each value anywhere on the document by understanding what it means, not where it sits on the page. This is the same mechanism explained in detail in our guide to extracting COI data to Excel — and it is what makes the extraction step scale linearly with file count, not complexity.

For GCs who need more than a spreadsheet for monitoring — particularly those managing compliance requirements that vary by project — the structured data from the extraction step can be directly imported into COI tracking platforms like Billy, BCS, or myCOI via CSV. The platforms handle the monitoring; the extraction step eliminates the typing. This is the practical middle ground that most discussions of the COI tracking email-and-spreadsheet problem overlook: you do not need to choose between a $0 spreadsheet and a $500/month platform. You can combine AI extraction with whichever tracking method fits your current scale.

Frequently Asked Questions

At what subcontractor count does manual COI tracking actually break?

Most COI tracking platforms and insurance compliance guides — including ExpirationReminder and Billy's scale analysis — converge on 30-50 subcontractors as the point where Excel becomes unmanageable. The specific number depends on your project count and policy complexity: a GC with one project and 30 subs may still manage manually; a GC with five projects and 50 subs almost certainly cannot. The better indicator than headcount is time: if you are spending more than five hours per week on COI data management, you have outgrown manual tracking regardless of subcontractor count.

Can AI extraction handle the ACORD 25 form's multi-section layout?

Yes. The ACORD 25 Certificate of Liability Insurance contains a standard structure with separate sections for General Liability, Automobile Liability, Umbrella/Excess Liability, Workers' Compensation, and other coverage types — each with their own carrier name, policy number, effective date, and expiration date. AI extraction identifies each section by its header label and extracts fields within that section contextually. This is different from template-based OCR, which would require a separate template for every variation of the form layout. The AI reads the document the way a human does: by understanding what each section label means and locating the corresponding values underneath it. For batch processing of certificates, see our guide on batch-tracking 50+ subcontractor COIs.

Is AI extraction a replacement for COI tracking software like myCOI or Billy?

No — it serves a different function. COI tracking platforms handle monitoring, expiration alerts, compliance dashboards, and audit trail generation. They are excellent at Phase 4 and Phase 5 of the COI lifecycle. AI extraction handles Phase 2 and Phase 3: reading certificate fields from PDFs and converting them into structured data. At 200 subcontractors, a growing GC may need both. But many GCs find that once the data extraction bottleneck is removed, a well-structured spreadsheet with calendar alerts handles monitoring adequately at the 50-150 sub range — delaying the need for a dedicated COI platform until the compliance complexity (multi-project endorsements, OCIP/CCIP wrap-up programs, varying per-project requirements) genuinely requires it.

How do I handle subcontractors whose different policy types expire on different dates?

This is one of the most common sources of manual tracking failure. A sub's GL may renew January 1, their auto March 15, and their umbrella September 30. The extraction approach handles this by treating each policy line as a separate row in the output spreadsheet — so the same subcontractor appears on multiple rows, each with a different policy type and its own expiration date. When you sort the spreadsheet by expiration date, you see all policies expiring next month across all subcontractors, regardless of type. Calendar alerts can then be set per row rather than per subcontractor. This per-policy-line tracking is what makes the difference between a 40% compliance rate and the 95%+ rate that automated systems typically achieve.

What does COI non-compliance actually cost a general contractor?

The costs arrive in three forms. First, audit premium recapture: when your GL carrier's year-end audit cannot produce a valid COI for a subcontractor, those sub payments are reclassified as uninsured, and the carrier charges you at the sub's trade rate — frequently $4,000+ per uninsured sub per audit. Second, uninsured claims: when a sub's coverage lapsed and an incident occurs, the liability transfers upward to the GC's own policy, affecting your loss history and Experience Modification Rate for years. Third, regulatory penalties: OSHA fines, stop-work orders, and project delays. We examined these costs in detail in our analysis of what COI non-compliance actually costs.