Scaling a PMC Firm With AI Project Intelligence — Without Adding Headcount

The PMC Scaling Problem

A typical PMC firm in New Zealand has somewhere between 5 and 25 people. The revenue model is straightforward: experienced professionals administer construction contracts, provide advisory services, and earn fees based on time or project-based engagements. The margin is healthy when senior staff spend their time on advisory work, the strategic thinking and professional judgement that clients are paying a premium for.

The problem is that senior staff are not spending their time on advisory work. They are spending it on administration. Tracking deadlines. Chasing correspondence. Checking whether the contractor's programme was submitted on time. Reviewing payment claims against the contract. Monitoring whether notices were given within the required timeframes. Updating spreadsheets.

The PlanGrid and FMI 2018 industry research found that construction professionals spend 35% of their time on non-productive activities. Looking for project information, managing conflict, and dealing with rework. That is more than 14 hours per week per person. For a senior contract administrator billing at $250 per hour, that represents over $180,000 per year in capacity that is consumed by administrative friction rather than professional value.

This is the scaling trap. Every new project requires administrative capacity. Administrative capacity is currently provided by expensive senior people or risky junior people. The firm cannot grow without one or the other.

Why Hiring More Graduates Does Not Solve the Quality Problem

The instinct is to hire. A new graduate costs less than a senior practitioner. Put them on a project, give them a checklist, and let them handle the administration while the senior people focus on the high-value work.

In theory, this works. In practice, it creates three problems:

1. The quality floor drops

A graduate assigned to administer an NZS 3910 contract does not know what they do not know. They can follow a checklist. They cannot recognise when a special condition has modified a standard timeframe in a way that creates a time-bar risk. They cannot assess whether a contractor's claim notice meets the substantiation requirements of Clause 13.3.1. They cannot identify that a variation instruction has triggered a cascade of three other obligations that are not on their checklist because the checklist was written for the standard conditions, not the 150 special conditions attached to this particular contract.

The Construction Sector Accord found that government NZS 3910 contracts carry anywhere from 7 to 310 special conditions. A graduate with a checklist designed for the standard form is not equipped to manage the complexity that special conditions introduce.

2. Senior staff end up supervising instead of advising

Every output the graduate produces needs to be reviewed. Every payment certificate needs checking. Every notice response needs sign-off. The senior practitioner has not been freed from the administrative burden. They have traded doing the work for reviewing someone else's attempt at it. In many cases, reviewing takes longer than doing, because you have to understand what the junior did, check whether they did it correctly, and correct it if they did not.

3. The PI exposure increases

The PMC firm carries professional indemnity insurance for every project it administers. If a junior staff member misses an obligation, whether a claim notice deadline, a retention release, or a variation response timeframe, the firm is exposed. The PI claim does not care whether the mistake was made by a graduate or a director. The firm is the professional, and the firm bears the liability.

A single missed obligation on a major contract can generate a claim that exceeds the fee earned on the entire engagement. Hiring juniors to increase capacity while maintaining the same risk profile is not scaling. It is leveraging.

How AI Project Intelligence Multiplies Senior Capacity

AI project intelligence does not replace the senior practitioner. It replaces the administrative overhead that prevents the senior practitioner from doing their actual job.

When we talk about "AI capability" in contract administration, we are not talking about a chatbot that answers questions about NZS 3910. We are talking about specific, operational functions that consume time and carry risk when done manually:

Obligation tracking

The system ingests the full contract, standard conditions and special conditions, and extracts every obligation. Each obligation is tagged with a responsible party (Contract Administrator, Independent Certifier, contractor, principal), a trigger condition, a timeframe, and a consequence for non-compliance. The system tracks these obligations in real time. No spreadsheet. No manual updates. No reliance on someone remembering to check.

Deadline management

When an event triggers an obligation, such as a contractor submitting a claim notice, a variation being instructed, or practical completion being requested, the system activates the related deadlines automatically. The responsible person is notified with the specific context: what is due, when it is due, what happens if it is missed, and what clause governs it.

Special conditions flagging

The system identifies where special conditions have modified the standard form. Where a notice period has been tightened from "as soon as practicable" to 10 working days. Where a retention percentage has been increased. Where a new approval gate has been introduced. These modifications are flagged at onboarding and tracked separately. Because missing a modified obligation is exactly the kind of mistake that generates disputes.

Daily intelligence

Every morning, the contract administrator receives a briefing: what is due today, what is due this week, what is at risk, and what has been missed. Not a generic report. A targeted, prioritised summary of the five things that matter most on each contract they are managing. This is the difference between reactive administration (finding out about a missed deadline after the fact) and proactive administration (catching it before it becomes a problem).

CA/IC obligation separation

Under NZS 3910:2023, the Contract Administrator and Independent Certifier roles carry different obligations. The system tracks these separately. Even when both roles are held by the same person. This builds the documentation discipline that protects the professional if a certification decision is ever challenged. For more on this distinction, see our article on the CA vs IC role split.

The Competitive Differentiation Angle

PMC tenders in New Zealand follow a familiar pattern. The submission describes the firm's experience, names the proposed team, attaches CVs, and outlines a methodology that reads like every other firm's methodology. The differentiator is usually the individuals. Their track record, their relationships, their reputation.

This is not a bad model. But it is a limited one. It means the firm's competitive position is entirely dependent on the availability of specific senior people. If your best contract administrator is already committed to another project, you cannot credibly name them on the tender. And without them, your submission looks thinner.

AI project intelligence changes the conversation. Instead of selling only the individual, the firm sells the system. The tender submission demonstrates how the project will be administered systematically. Obligation tracking, deadline management, special conditions flagging, daily intelligence briefings, CA/IC obligation separation. This is tangible. It is auditable. And it tells the client something important: this firm has invested in ensuring nothing falls through the cracks, regardless of which individual is doing the day-to-day work.

Sophisticated clients respond to this: government agencies, experienced developers, and organisations that have been through a construction dispute. They know that the quality of contract administration is the single biggest factor in whether a project stays out of dispute. The Arcadis Global Construction Disputes Report has consistently identified failure to properly administer the contract as the number one cause of construction disputes worldwide.

A PMC firm that can demonstrate systematic project intelligence capability is not just another option on the tender list. It is the firm that has addressed the number one cause of disputes before the project starts.

The Liability Angle

PMC firms carry PI insurance because they carry professional liability. The advice they give, the certifications they issue, the deadlines they manage: all of it creates exposure. And that exposure is amplified by the NZS 3910:2023 CA/IC split, because the roles now carry distinct obligations with distinct accountability.

Consider this scenario. A junior contract administrator is managing a $60M project. The contract has 120 special conditions. One of those special conditions modifies the claim notice period under Clause 13.3.1 from "as soon as practicable" to 10 working days, with a strict time bar. The contractor submits a claim notice on day 8. The junior does not process it within the required timeframe. The principal loses the ability to challenge the claim. The value of the claim is $1.2M.

The PMC firm earned $180,000 in fees on that project. The PI claim is $1.2M. The firm's premium increases. The relationship with the client is damaged. The junior did not make the mistake because they were negligent. They made it because the system they were working with did not flag the modified timeframe from the special conditions.

AI project intelligence does not eliminate professional liability. But it eliminates the category of error where obligations are missed because nobody was tracking them systematically. That is not the same as professional judgement. It is the infrastructure that supports professional judgement.

What a PMC Firm's Workflow Looks Like — With and Without

Without AI project intelligence

1. Contract awarded. Senior CA reads the contract. Makes notes. Sets up a spreadsheet.
2. Special conditions reviewed. Some flagged, some missed because the review is manual and time-pressured.
3. Obligations tracked in spreadsheet. Updated when someone remembers. Falls behind within weeks.
4. Deadlines managed by calendar reminders and institutional knowledge. When the CA is on leave, nobody knows what is due.
5. Junior staff follow the checklist. Checklist does not cover special conditions. Gaps emerge.
6. Payment certificates issued based on the CA's assessment. No systematic check against all relevant obligations.
7. When a dispute arises, the team scrambles to reconstruct the timeline from emails, minutes, and memory.

With AI project intelligence

1. Contract awarded. System ingests standard and special conditions. Complete obligation register generated.
2. Every special condition modification flagged: tightened timeframes, additional obligations, modified payment terms.
3. Obligations tracked automatically. Deadlines activate when trigger events occur. No manual updates required.
4. Daily briefing delivered to CA: what is due today, what is due this week, what is at risk. Coverage continues when CA is on leave.
5. Junior staff work within the system. The system catches what the checklist misses. Senior staff review by exception, not by rote.
6. Payment certificates cross-referenced against obligation register. Nothing certified without confirming all prerequisites are met.
7. Complete audit trail maintained from day one. If a dispute arises, the chronology is already built.

The Economics of AI-Enabled Scaling

The economics are straightforward. A senior contract administrator currently manages two to three contracts simultaneously before quality starts to slip. With AI project intelligence handling the systematic tracking, the same person can manage four to five contracts. Because the administrative overhead on each contract is reduced, not the professional input.

For a PMC firm with three senior CAs, that is the difference between managing nine contracts and managing fifteen. Six additional contracts at an average fee of $15,000 per month is $90,000 per month in additional revenue. Without hiring a single additional senior person.

The cost of the AI system is a fraction of that revenue increase. The margin improvement is substantial. And the quality floor is higher, not lower, because every contract is being tracked systematically rather than relying on individual memory and spreadsheets.

Where to Start

If you are a PMC principal or managing director thinking about this, here is a practical starting point:

1. Audit your current obligation tracking. How do your CAs track deadlines? How many are in spreadsheets? How many are in someone's head? What happens when that person is on leave?
2. Count the special conditions on your current projects. How many obligations do they create beyond the standard form? Are all of them being tracked?
3. Measure your senior staff utilisation. What percentage of their time is spent on administration versus advisory work? That gap is your scaling opportunity.
4. Assess your PI exposure. On which projects would a missed obligation create the greatest financial exposure? Those are your highest-priority candidates for systematic tracking.
5. Talk to us. Provan can run a working session with your team to map your current workflow, identify the gaps, and show you what AI project intelligence looks like on a live NZS 3910 contract.