Usage Spike on a Bill Confirms or Contradicts the Diagnosis, Decision Tree

Why this matters

You've formed a diagnosis from the inspection, and the customer has a usage spike on a bill they want explained. The instinct is to treat the spike as automatic proof your diagnosis is right, because a spike happened and you found a fault, so the two must be the same story. Sometimes they are. Sometimes the spike has a completely different cause than the fault you found, and treating them as confirmed together leads you to bill a repair that won't actually fix the bill. This tree walks through checking the spike against your diagnosis properly, in either direction. (See the related reference article for how to read bill history as evidence in general.)

Start here: does the timing line up

Compare the date the spike started against the date your suspected fault likely began, using whatever timeline evidence you have (visible wear, the customer's account of when symptoms started, service history).

If the spike's start date and the fault's likely onset are close together, that's a real point in favor of a match, continue checking magnitude and pattern next.

If the spike started well before or well after your suspected fault could plausibly have begun, that's a contradiction, not a confirmation, even though both are real and even though the fault you found is genuinely a problem worth fixing on its own. Do not tell the customer this specific repair will resolve the bill spike unless the timing actually supports it. Keep looking for what actually changed at the time the spike started.

If the timing lines up, check the magnitude

A matching timeline isn't enough on its own; the size of the usage change should be plausible for the fault you found. If the suspected fault is the kind that would produce a large, continuous load (something stuck running, a major restriction forcing much harder operation) and the spike is correspondingly large and sustained, that's consistent. If the suspected fault is minor, and the spike is large, be suspicious that a second cause is contributing, or that the fault you found isn't actually the driver of the bill increase, even if it's a real problem.

Conversely, a large confirmed fault paired with only a small usage change is also worth a second look. It doesn't automatically mean the fault is unrelated, some faults genuinely produce a modest usage effect, but it's worth explicitly asking whether the fault as diagnosed is enough to explain what you're seeing on the bill before presenting the two as a confirmed pair to the customer.

If the pattern shape matches what the fault would predict

Different fault types predict different usage patterns, not just different sizes. A component stuck permanently on predicts a flat, continuously elevated baseline across the whole billing period. A fault that only triggers under specific conditions (a certain temperature, a specific time of use) predicts a spike that shows up only during those conditions, visible on interval data if the utility provides it. If the pattern shape in the bill matches the shape your diagnosed fault would produce, that's strong corroborating evidence, weigh it as more supportive than either the timing or magnitude alone.

If the pattern shape doesn't match, for example, you diagnosed something that should be intermittent but the bill shows a flat continuous elevation, or vice versa, treat that mismatch seriously. It usually means either the diagnosis needs revisiting, or there's an additional cause layered on top of the one you found.

If a billing-side explanation is possible

Before concluding the spike is equipment-driven at all, rule out the non-equipment explanations: an estimated meter read followed by an actual read correcting it, a rate change or plan switch, a change in occupancy or season the customer hasn't mentioned. If any of these plausibly explains the spike on its own, confirm which one before attributing any of it to the fault you diagnosed, because otherwise you risk claiming credit, or blame, for a bill change your repair had nothing to do with.

If everything lines up: timing, magnitude, and pattern

If timing, magnitude, and pattern all point the same direction as your diagnosed fault, you have strong corroboration and can tell the customer with real confidence that fixing this should visibly affect their next bill. This is also useful after the repair: telling the customer to watch their next bill for the expected drop gives you a second, independent confirmation that the fix actually worked, beyond just the equipment running fine at the moment you left.

If the spike contradicts the diagnosis

If the checks above genuinely contradict your diagnosis, do not let the confirmed, real fault you found stand in as an explanation for a bill spike it doesn't actually match. Both things can be true at once: you found something real, and it isn't what's driving the bill. Say so plainly to the customer, fix what you found because it's a genuine issue, and keep the bill spike open as a separate question to pursue, rather than closing it out on a fault that doesn't fit the evidence.

Recap

1. Check whether the spike's start date matches the fault's likely onset.
2. Check whether the spike's size is plausible for the fault type you found.
3. Check whether the pattern shape (flat versus intermittent) matches what the fault would predict.
4. Rule out billing-side explanations before attributing the spike to equipment at all.
5. When timing, magnitude, and pattern all agree, that's real corroboration, tell the customer to watch the next bill as confirmation.
6. When they contradict, say so honestly and keep hunting rather than forcing a match.

References

• Trade-standard practice for cross-checking equipment diagnosis against historical usage data
• See related: What a Utility Bill Can Tell You That an Inspection Can't; The First Plausible Cause Isn't Always the Real One