For grant-making offices in government
The cycle that ends at award is the visible part. Behind it sit open awards from four previous cycles, each with reporting due, drawdowns to review, subrecipients to monitor, and a closeout nobody has time to start. That administrative weight can be carried. The award decision, the risk determination and the disallowance stay with your officers.
The office is measured on getting money out the door, so the announcement, the review panel and the award package absorb the calendar. Everything downstream of award competes for whatever is left, which in a busy cycle is nothing.
So the backlog is not applications. It is progress reports filed and never read, financial reports whose numbers nobody reconciled against drawdowns, monitoring visits scheduled and postponed, corrective actions issued and never followed up, and closeouts sitting at ninety percent complete because the last document never arrived and nobody chased it.
Recipients are not adversaries and mostly are not sophisticated. A small nonprofit receiving its first federal pass-through has a bookkeeper, not a grants accountant. It will file late, file the wrong form, and misunderstand a cost principle — and the consequence of that lands on your office at audit, not on theirs.
The record is scattered by design nobody chose. Award terms in one system, financial data in another, the actual correspondence in an inbox, the monitoring notes in a shared drive folder named after a staff member who left in 2029. Reconstructing what an office knew and when it knew it is a week of work per award.
And the deadlines are external and unforgiving. A period of performance ends whether or not the report arrived. Funds expire whether or not they were drawn. An audit finding attaches to your office whether or not the underlying failure was a recipient’s. The clock does not care that the cycle was busy.
Every fact needed to steward an award exists somewhere, and no one place holds the current state of an award — so answering a simple question about a single recipient costs hours of assembly, and doing it across a portfolio is not attempted.
That is why monitoring degrades into whoever shouts. An officer with fifty open awards and no portfolio view manages the recipient who calls, not the recipient who is quietly failing. The quiet one is the audit finding.
So the first act is to assemble state, not to automate work. For each open award: where it is in its period of performance, which reports are due and which arrived, whether drawdowns are consistent with reported progress, which corrective actions are open and how old, what closeout is missing. Assembled once, that view usually reveals a handful of awards nobody had looked at in a year.
From there the administrative half moves. Collecting a report and checking it arrived complete and on the right form. Chasing the missing document. Reconciling two of your own numbers and reporting the difference. Assembling the closeout package. Producing the file an auditor asks for. All of this is document work with written rules, and it is what consumes an officer’s week.
What never moves is the judgement. Eligibility. Risk determination. Whether a cost is allowable. Whether to disallow, suspend, or terminate. Whether a corrective action is adequate. Those are exercises of authority delegated to your officers by regulation, they stay there, and an arrangement that quietly moved them would be a finding in itself.
Reports arriving complete and on the correct form the first time — measured by count of submissions returned for form or completeness, before and after, by recipient size.
Time between a report arriving and anybody reading it — measured by elapsed days from receipt to first substantive officer review, per award.
Awards nobody has examined in the current period — measured by a portfolio count of open awards with no officer touch inside a stated window — usually the most uncomfortable number produced.
Corrective actions open past their own due date — measured by count and age of open corrective actions, which is presently unmeasured in most offices.
Closeouts stalled on a single missing item — measured by count of awards at closeout, and the age of the oldest, with the specific missing item named.
Time to assemble a complete award file for an auditor — measured by elapsed hours from request to a complete file, timed on a real request rather than estimated.
Consistency of what recipients are told — measured by a sample of the same question answered across officers, scored against the notice of funding opportunity.
any award decision, eligibility determination, risk determination, allowability determination, disallowance, suspension or termination. Those are exercises of authority your officers hold under regulation. Nothing here scores an application, ranks a recipient, or decides whether a cost is allowable, and any arrangement that moved one of those would be a finding rather than an efficiency.
Award terms, budgets, periods and reporting calendars are read from the systems you already use. Nothing becomes a second authoritative record of an award, because two authoritative records is exactly the condition that makes an audit response impossible.
Where the award system and the financial system disagree — and they will, because they are updated by different people on different days — the disagreement is reported to your officer with both figures and both sources. It is never reconciled by choosing one, because choosing one is a determination.
Recipient correspondence is written back into the system of record for the award, so the file a monitor or an auditor opens later contains what was actually said rather than a pointer to somebody’s mailbox.
Where a recipient portal or a federal reporting system has no interface available, that is stated as a limitation and the step stays manual, rather than being covered by driving a screen and producing a record nobody can verify.
Recipient information includes financial detail about organisations that did not choose a vendor relationship with us. It stays inside your tenancy, on your retention schedule, exportable by you without a request to us, and it is not used to train anything serving another organisation.
The audit file is the product that matters here. Every action carries what was done, under which written grant of authority, by which actor, at what time, against which award, with what result — produced during the work. A file assembled afterwards to describe what probably happened is the thing auditors are trained to distrust.
On assurance: an independent SOC 2 Type II attestation is in progress and no report exists yet. Where a specific programme carries a federal privacy or security obligation, the applicable terms are agreed in writing during scoping. We would rather find a mismatch there than at a single audit.
And the boundary is written before anything runs: what is administrative, what is a determination, and what happens when a case cannot be classified. That last category goes to your officer by default rather than being absorbed into the administrative pile.
The grants officer needs the boundary in writing: exactly which acts are administrative and which remain theirs under regulation. That document is produced during scoping and is short enough to read completely, which is the only length at which a boundary document is actually read.
The CFO’s question is whether anything can touch a drawdown or an obligation. It cannot. Financial reconciliation here means comparing two of your own numbers and reporting the difference; approving a drawdown is a determination and stays with the people who hold it.
Internal audit will want the record format before anything runs, and should get it. If the record it produces would not satisfy your own auditors, that is worth discovering during scoping rather than during a single audit three years later.
And counsel needs to see the authority grant: enumerated, revocable immediately by the official who granted it, without a contract action, and covering only the administrative band.
One programme’s open awards — a single funding stream across its live cycles — assembled into current state, with no authority to contact a recipient or touch a record.
The assembly phase produces one artefact: the current state of every open award in that stream. Reports due and received. Drawdowns against reported progress. Corrective actions open and their age. Closeouts and their missing item. Awards with no officer contact inside the window.
That artefact is frequently the whole value, and stopping there is a legitimate outcome. An office that learns it has eleven awards nobody has examined this year now has a management problem it can act on, and it did not need to delegate anything to find out.
If it continues, the first grant covers one administrative act — usually report collection and completeness checking — for one funding stream, with every determination explicitly outside it.
Correct, and the determinations are not delegated here — the risk determination, the allowability call, the adequacy of a corrective action and the decision to disallow all stay with your officers under their delegation. What moves is the document work around those decisions: collecting, checking completeness, chasing what is missing, reconciling two of your own numbers and reporting the difference. If your counsel reads the boundary document and finds a determination inside it, that is a scoping error and it should be removed before anything runs.
Most do, and the honest question is not whether the system has the feature but whether the work is actually getting done. A system with a reporting calendar still requires somebody to notice a report did not arrive, chase it, check it when it does, and escalate the numbers that disagree. Start with the assembly phase against your existing system: if it shows no stalled closeouts, no aged corrective actions and no untouched awards, your system is being worked properly and you should not buy anything.
The design point runs the other way. Most of what burdens a small recipient is being told late that a submission was wrong, and then guessing what to fix. Checking completeness at receipt and naming the specific missing item is less process for them, not more, and it is the difference between one exchange and four. The reminder before a deadline is also for them: a small recipient that misses a period-of-performance date usually did not know it was approaching.
The design intends that it cannot, because the administrative band is defined by an enumerated list rather than by exclusion — anything not on that list is out of scope by default and routes to your officer, including anything unclassifiable. The record then shows which acts were taken under which written grant, so an internal audit can test the boundary directly rather than take it on assurance. And the grant is revocable immediately by the official who signed it, without a contract action.
That is a fair weighting and it should govern the scope. Two things reduce the exposure: the assembly phase delegates nothing at all, so the first thing you get is a portfolio view you keep whatever happens next, and every record produced is in your systems and exportable by you without asking us. Exit terms belong in the initial scoping documents rather than in a later negotiation, and if your internal audit will not accept the record format, that is a reason to stop before anything is delegated.