The Timeline Between Submission and Payment
Why this matters
The single sentence that generates the most rebate-related aggravation in a shop is a well-meaning estimate of when the money will arrive. Say a number, and the customer writes it on the calendar. When it slips, and a meaningful share always slips, they do not call the program. They call you, they call you more than once, and each call costs office time you never priced into the job. The fix is not to say less. It is to say a number that is true nine times in ten instead of a number that is true half the time.
Why no program can tell you when you will be paid
A published processing window is a service target for a claim with no complications, produced by an organization whose queue depth, funding cycle and staffing you cannot see. Four structural facts make any single number soft:
- Intake and adjudication are separate queues. A claim can be received quickly and sit unreviewed, and the program's clock may or may not start at receipt.
- Funding runs on a period, not on a stream. Programs with an annual or quarterly budget behave differently at the start of a period than at the end of one.
- Payment is a batch process. Even an approved claim usually waits for the next payment run, and where the incentive is delivered as a credit on a utility bill rather than a check, it waits for the account's next billing cycle on top of that.
- Any information request resets you to the back of a queue. A three-day response from you does not buy a three-day delay; it usually buys a full cycle.
None of that makes a published window dishonest. It makes it a floor rather than a forecast.
The stages the money passes through
| Stage | What is happening | What can stall it |
|---|---|---|
| Submission to receipt | Your claim lands in the program's system | Portal errors, mailed submissions, a missing confirmation reference so you cannot tell |
| Receipt to completeness screen | Fields and required documents checked | Form version, illegible invoice, missing signature |
| Screen to adjudication | Eligibility compared against lists and prior claims | Queue depth, seasonal surges, an ambiguity that becomes an information request |
| Adjudication to approval | The decision is recorded | Nothing much, once it gets here |
| Approval to payment run | Approved claims batched for payment | Payment run cadence, fiscal period boundaries |
| Payment run to delivery | Check issued and mailed, or credit applied | Mail transit, or the account's next billing cycle for a bill credit |
Six stages, and your submission quality only influences the first three. Everything from adjudication onward moves at the program's cadence no matter how clean your paperwork was, which is why "we filed it perfectly" is not an answer to "when do I get it."
Build your own baseline
The only timeline that predicts anything for your shop is the one built from your own paid claims, per program. Keep it in the submission log and compute it from two dates you already record: submission date and the date funds were received or the credit appeared. Not the approval date. The customer does not experience approval.
Two things make the number trustworthy. Measure per program, because a manufacturer claim and a utility claim on the same job routinely differ by weeks and averaging them produces a number that describes neither. And recompute at least annually, because a program that paid quickly under one administrator can slow substantially when the contract or the funding source changes, and your baseline will quietly go stale while continuing to look authoritative.
Ten paid claims is enough to start. Below that, use the program's published window plus a wide marker and say so out loud.
Reading the distribution: the median plans, the tail promises
Two numbers come out of the history, and confusing them is the whole error.
The median is a planning number. It tells you what a typical claim does, which is what you need for sizing office workload and forecasting how many claims will still be open at season end. It is a terrible number to give a customer, because a median is the value half your claims exceed, by definition. Quote it and you have guaranteed that roughly half your customers will call you about a late rebate.
The nine-in-ten mark is a promise number. It is where you set the chase date, and it is the marker you give the customer as "if it goes past this, call me." Quote it and about one customer in ten calls early, which is a queue you can actually staff.
Twenty paid claims, read properly
One program, one shop, twenty paid claims. Weeks from submission to funds received, sorted:
4, 5, 5, 6, 6, 6, 7, 7, 7, 7, 8, 8, 8, 9, 9, 10, 11, 12, 14, 19
Median. Twenty values, so the middle is the average of the 10th and 11th: 7 and 8, giving 7.5 weeks. Exactly 10 of the 20 claims, half of them, ran longer than that. That is not a flaw in the data, it is what a median is, and it is precisely why quoting 7.5 weeks to customers produces ten chase calls out of twenty.
The eight-week temptation. Thirteen of the twenty, 65%, landed at or under 8 weeks. Eight weeks feels like a safe round answer for that reason, and it still leaves 7 of 20, 35%, calling you.
The nine-in-ten mark. Eighteen of the twenty, 90%, landed at or under 12 weeks. That is the promise number and the chase date. It is also honest about the tail: 2 of the 20, the 14-week and the 19-week claims, breached it. Ten percent breaching a nine-in-ten mark is the mark working as designed, not evidence that it is wrong, and you should expect roughly that rate rather than treating each breach as a surprise.
The spread. Fastest 4 weeks, slowest 19 weeks, so the slowest claim took nearly five times the fastest. Any single number you quote is describing a process with that much internal variation, which is the real argument for quoting a marker with a "call me if" attached rather than a date.
What the worst claim was doing. The 19-week claim sat in an information request that arrived in a personal inbox and was not seen for three weeks. It is worth pulling that one apart rather than shrugging at it, because it was not the program being slow. It was the shop's own tracking failing, and the fix lives in the mail routing, not in the baseline.
What happens if you drop it. Remove the 19 and you have nineteen claims. The median moves from 7.5 to 7.0 weeks, half a week. The tail moves more: 18 of the remaining 19, about 95%, still land at or under 12 weeks, and 17 of 19, about 89%, land at or under 11. So dropping one bad claim would let you promise about a week tighter while barely changing your planning number. That is the general shape worth remembering. Outliers move the number you promise on far more than the number you plan on, which means fixing your own tracking failures pays out mostly as a better promise to customers.
Do not actually delete outliers from a live baseline. Recompute both ways when you are diagnosing, keep the full set for the number you quote, and treat a persistent tail as a signal about your process rather than as noise to be cleaned.
What lengthens a timeline
- Season and program period. Claims filed into a peak, or right before a program period closes, meet a deeper queue. If your work is seasonal, expect your own baseline to have a slow half of the year inside it, and consider computing the mark separately for peak months once you have enough claims.
- Funding cycle boundaries. A claim approved just before a fiscal boundary can wait for the new period's payment run for reasons that have nothing to do with the claim.
- Bill credit rather than check. A credit applied to a utility account waits for that account's next billing cycle after approval, which stretches the tail and, more importantly, means the customer may not recognize it when it arrives. Tell them what form it will take, not just when.
- Any information request. Treat these as adding a full cycle, not as adding your response time.
- Inspection dependencies. Where the program requires a passed inspection, the clock the customer feels starts at their install, not at your submission, and the permit office's schedule sits inside that gap.
- A mid-season terms change. These slow everything, including claims filed under the old terms, because the processor's own staff is absorbing new rules at the same time.
What to say at week zero
Say four things, in this order, and stop. The claim was submitted, on this date. Here is the program's own published window, if it publishes one. Here is our own marker, at the nine-in-ten mark, and if it passes that date I will contact them. And here is the form the money will take, check or bill credit, so you recognize it.
Three sentences to keep out of your mouth. Any date certain. Any amount stated as approved. And any version of "it usually takes about," which is your median wearing a promise's clothes. If the customer pushes for a single number, give the nine-in-ten mark and name it as the outer marker rather than the expectation, because a customer who gets paid earlier than the marker is never unhappy about it.
References
- The current published processing window and payment method of the specific utility, manufacturer or state program, which is the only figure you may quote as theirs
- See related: How to Track a Submission You Do Not Control; The Rebate Submission SOP; Explaining Rebates and Incentives Without Overpromising