Four Organisations, Four Sets of Questions

The previous article argued that organisations can build analysis on their own project data rather than waiting for a vendor to ship it. The obvious response is fair enough, but like what?

Here is the useful part. Take the same PMIS, holding the same kinds of record, and put it in four different organisations. The questions each one urgently needs answered have almost nothing in common.

That is precisely why no vendor builds them.

Take the same PMIS, holding the same kinds of record, and put it in four different organisations. The questions each one urgently needs answered have almost nothing in common.

The owner and the developer

An organisation delivering its own assets is managing capital exposure across programmes that are at different stages, and the questions are about position rather than progress.

Where does the portfolio actually stand between approved budget, committed value, certified work and forecast final cost — and how has that spread moved over the last four reporting periods? The individual numbers exist on every dashboard. The movement between them over time is what tells you whether estimates are holding, and it is rarely presented.

How concentrated is exposure by counterparty across the whole portfolio, including where the same entity appears as a subcontractor on packages held by different main contractors? Concentration is invisible at project level by definition, and it is the exposure that matters when a counterparty gets into difficulty.

Where is uncommitted contingency actually sitting, by programme and by phase, and how does that compare with the risk still open on each? Contingency is usually reported as a balance and is only meaningful against the — which is a question that crosses from the PMIS into the sales pipeline and is never answered by either system alone.

The PMC and the consultant

A consultancy delivering for multiple clients has a different problem. The projects are similar; the contracts governing them are not.

What obligations fall due in the next sixty days across every contract, given that each client's terms differ? Notice periods, submission deadlines, reporting obligations and approval windows vary by contract, and the standard answer is a spreadsheet maintained by whoever is most diligent.

How long does correspondence sit before it is answered, by project and by counterparty? Response latency is one of the earliest available indicators of a project heading toward claims, and it is measurable from the correspondence record long before anyone uses the word.

Which projects generate disproportionate variation volume relative to their value, and is that a client characteristic or a project characteristic? A firm that knows this prices differently and staffs differently.

What is the deliverable load across the portfolio over the next quarter, and where do the peaks collide? Resource planning across concurrent client commitments is a scheduling problem the scheduling tool does not solve, because it lives across projects rather than within one.

The contractor

A contractor is managing cash and margin under terms somebody else wrote, and the questions follow the money.

What is the elapsed time from payment application to certification to payment, broken down by client, and how has it moved? Cycle time by counterparty is a cash forecasting input and a commercial negotiation input, and most contractors have a general sense of it rather than a measurement.

Where does margin erode between the approved position, the certified position and the paid position, by project and by work package? The erosion is usually discovered at final account, when the causes are historic.

What is the concentration in the supply chain, and what does performance history say about the subcontractors carrying the most exposure? This is the same counterparty question the owner asks, one tier down.

What variations are open, at what value, and how long have they been open? Ageing of unapproved variations is a direct predictor of the claims position and it is derivable from records that already exist.

The government entity and the PIF company

An organisation delivering under a national framework has obligations the other three do not, and its questions are about evidence and consistency.

Where does the portfolio stand against the framework's reporting requirements right now, without a data call — because the data call is the problem. The ability to answer at any moment from the system of record, rather than by asking twelve delivery entities to submit, is the difference between governance and administration.

How consistently is practice applied across delivery agencies? When several entities run the same processes, variation in how they are actually used is invisible unless somebody measures it, and it is the leading indicator of where assurance should be directed.

Can an audit question arriving with no notice be answered from evidence, in the same week? The question is rarely about whether the answer exists. It is about whether it can be assembled before the meeting. something to a board and can still be drilled into?

The pattern

Look across the four and one thing stands out. Not one of these is a feature request. Every one of them is answerable from data the organisation already holds, in a system it already runs, and none of them appears on any vendor's roadmap because they are specific to how each organisation is structured and what it is accountable for.

None of these is a feature request. All of them are answerable from data you already hold.

That is the case for the client-built layer, stated concretely. The vendor builds what most customers need. The partner builds what the category needs. What your organisation needs, specifically, is left over — and it is usually the part that would change a decision.

Where to start

Pick the question your organisation has repeatedly wanted answered and repeatedly not answered, because the effort was never justified for a single occasion. That question is almost always the right first one, because the fact that it keeps recurring is evidence that it matters and evidence that nobody has solved it.

Answer it once, properly, with the governance conditions from the previous article in place. If the answer changes a decision, it becomes a monthly output. If it does not, you have spent an afternoon finding that out.

Pick the question your organisation has repeatedly wanted answered and repeatedly not answered, because the effort was never justified for a single occasion. That question is almost

always the right first one.

Next week: the objection that stops more PMIS decisions than price, and why it rests on a false analogy.