SystemBilling, Credit & Collections

Revenue Assurance Analytics

Utility revenue assurance that finds unbilled and under-billed revenue on named premises, every week, with the evidence for every back-bill

See one case, screen by screen ↓
Watch with transcript →

Built forRevenue assurance analystsRevenue assurance managersFinance controllersInternal auditors

The Problem Today

The network map, meter data and billing are not joined, so premises consume for months with no one paying.

1–5%
of revenue lost between meter and cash, in industry estimates published
12.5 h
to build the evidence for one back-bill case by hand our estimate

Four systems nobody joins

  • Network map
  • Meter data
  • Accounts
  • Tariff
  • Each month unfound is lostIn Great Britain, not-at-fault back-bills stop after a year
  • A sample, once a yearMonthly reports run without the network map

The Solution

  1. Each premise tracedTo its transformer, feeder and substation, with its energized date
  2. Three records joinedConnected and consuming, with no account? Each gap filed in its bucket
  3. Priced, and the cause datedValued month by month at tariff, with the event behind each cluster
  4. The manager signsInside the back-billing rule, evidence attached, prepared by a colleague

Built-in controls

  • Billing records are changed only by people
  • Suspected tampering goes to revenue protection
  • Each figure opens the record behind it

What You Get

Leaked revenue is billed back inside the rule, and fewer gaps reopen once their causes are fixed.

1.7 h
to build one back-bill case's evidence, down from 12.5 h our estimate
  • Unbilled premises found soonerConnected, consuming premises with no account, caught in the weekly run
  • Each cause handed to its ownerWith how many premises it refills a week, and the fix
  • Found, billed, then collectedEach case reconciled to cash for the audit committee
In short

What is Revenue Assurance Analytics?

Revenue assurance is the work of finding electricity and water that is delivered but not billed, or billed too low, and recovering it. Revenue Assurance Analytics does it for each premise each week by joining the network map, meter data, accounts, service orders and the tariff master, and puts each gap on a named premise with a value at tariff.

How to choose AI solutions for utilities →

Updated

The problem

Revenue assurance gaps in the seams between network map, meters and bills

Revenue leaks in the seams between the network map, meter data, customer accounts, service orders and the tariff master. A tower is handed over by its developer and its tenants are not set up as customers. A meter stops talking and is billed on a frozen estimate. A current-transformer ratio is changed in the field but not in billing. A move-out is left energized. A reconnection fee is not charged. The network map knows the premise is connected, meter data knows it is consuming, billing knows nobody is paying, and nobody joins the three.

So the revenue assurance team runs about twenty exception reports on billing and meter data once a month, without the network map. Premises that are connected and consuming with no account turn up only when someone complains, a meter reader notices or internal audit samples. Building the evidence for one back-bill (trace, reads, account history, tariff and rule) takes an analyst most of a day. And each month a gap goes unfound brings it closer to the back-billing limit, after which it can no longer be billed.

typical1–5%of revenue lost between meter and cash, in industry estimates
typical12monthsback-billing limit in Great Britain when the customer is not at fault: energy used before that cannot be billed
Where the hours on one back-bill case goestimated
By hand12.5 hours
With the solution1.7 hours
  • Spotting the gap across meter data, billing and the network map3 → 0.2 h
  • Tracing each premise to its transformer and feeder1.5 → 0.1 h
  • Pulling the reads and the account history1.5 → 0.2 h
  • Pricing at tariff, tier by tier, inside the back-billing rule1.5 → 0.2 h
  • Finding why it leaked3 → 0.5 h
  • Assembling the evidence pack and getting sign-off2 → 0.5 h

Estimated split for one leakage case, from spotting the gap to a back-bill ready to sign, in hours, by hand and with the solution.

How it works

The weekly utility revenue assurance run, premise by premise

What comes in
Data in5 sources · network map, meter data, accounts, service orders, tariff
Agents at work
Network tracerevery premise
Meter data reader
Then
Account matcher9 checks, 8 buckets
Then
Revenue estimator
Root cause finder
Then
Case writerevidence pack
A person decides
Revenue assurance managerapproves the back-bill
What comes out
Leakage waterfall and map
Cases with a recoverable amount
Recovery tracked to cash
One case, step by step

A tower energized with no accounts, found and billed back inside the rule

The sample utility supplies electricity and water to 1,004,600 premises and billed $9,860.0M in the last 12 months. One Monday the weekly run finds a new ring of premises on one residential tower. Here is what happens next, screen by screen, in the working solution.

A worked example on sample data. Names, IDs and figures on the screens and in this section come from the demo.

  1. 01Monday, 06:51

    The year, drawn as one waterfall

    Six agents · the weekly run

    The run takes 06:10 to 06:51. The network tracer traces 1,004,600 premises to 9,412 transformers; the meter data reader reads 1,762,100 meters and finds 3,860 silent for 30 days; the account matcher runs 9 checks and files 17,607 gaps in 8 buckets. The waterfall runs from $9,922.3M delivered to 9,893.4M metered, 9,860.0M billed and 9,821.5M collected, each step down a named bucket.

    “$62.3M delivered in the year never reached a bill” · root cause finder: “New: one residential tower, 186 premises”

  2. 02Morning

    Consuming, no account: up 186 this week

    Revenue assurance analyst

    The analyst opens the bucket that moved: 1,240 premises, $3.1M a year at tariff, $2.2M back-billable under rule BR-7. Each premise is a dot on the map of the utility’s eight districts, and a new ring sits on the tower: 186 premises, $612k a year. Beside it, how the gap is found: network map connected and energized, meter data at least 30 kWh, or the set water volume, in 30 days, billing with no active account.

    Top root cause in the bucket: “Developer handover file rejected silently” · $1.1M

  3. 03One premise

    One apartment: connected, consuming, not billed

    Network tracer and meter data reader

    The trace runs from the substation through the feeder, the tower’s transformer and its rising main to one apartment, energized under a service order seven months ago. Its electricity meter has recorded 7,940 kWh since, and both meters show 100% actual reads. Billing shows no account, premise status “installed, not billed”, and the developer’s handover file rejected two days after the tower was energized.

    Priced at the tariff version in force: $3,228 to date. “No month went above the first tier, so no higher rate applies.”

  4. 04The case

    The case, with the three-way join for each apartment

    Case writer

    The case opens on the join, premise by premise: each of the 186 on the tower’s transformer, 158 consuming and 28 near zero (unlet), none with an account. Reads run from energization to the run date, priced at 29.5 cents/kWh and the water tariff, with the unbilled value by month for electricity and water.

    “186 premises · $612,000 a year at tariff · back-bill $370,600 from energization to date”

  5. 05The money

    Priced at tariff, inside the rule

    Revenue estimator

    Electricity, 919,700 kWh: $271,300. Water: $99,300. The back-bill is $370,600, $352,070 of it recoverable at 95% collectable, and $612,000 billed each year from now. The developer is billed from energization, because clause 9.2 of its connection agreement makes the developer the account holder until a tenant takes over the supply.

    Rule BR-7: developer-held supply, “from energization” · field check not needed: the largest premise is $3,228

  6. 06How it leaked

    A handover file, rejected silently

    Root cause finder

    Day one: the tower is energized, 186 premises. Day two: the meters start reading daily. Day three, 11:42: the developer’s handover file is rejected, because the tenant reference is blank on 186 of 186 rows, and the rejection notice goes to a mailbox nobody reads. No account is created. The same cause sits behind 412 premises across 6 developments.

    “Proposed fix: a rejected handover file raises a work item for billing operations within the hour.”

  7. 0707:18 · approved

    Approved, with the checks on the same screen

    Revenue assurance manager

    The analyst sends the case for approval. The revenue assurance manager signs for a back-bill of $370,600 to the developer, and 186 accounts in its name until tenants take over. Before the signature, three checks show: inside rule BR-7, evidence pack complete with trace, reads, account history and tariff version, and prepared by someone else.

    Agents find, price and draft; only named people approve a back-bill.

  8. 0807:24 · co-signed

    Two signatures, 186 accounts open

    Finance controller

    Above $250,000 a back-bill needs two signatures, so the finance controller co-signs after the manager. The back-bill is posted to the developer, billing operations reloads the handover file with the tenant references, and 186 accounts open, billed from the run date. The bucket falls back to 1,054 premises.

    “Billed. Back-bill of $370,600 posted to the developer; 186 accounts open, billed from the run date. The bucket is back to 1,054 premises.”

  9. 09The cause

    The cause behind the bucket, with its owner

    Billing operations lead

    Fixing a case recovers the money once; fixing its cause stops the bucket refilling. The root causes screen ranks 31 causes from 312 clusters by value a year and shows how much each refills a week. “Developer handover file rejected silently” refills 31 premises a week, worth $1.1M a year, and billing operations owns it.

    Root cause finder proposes causes and owners; it does not close a cause.

  10. 10Month end

    Found is not recovered

    Finance controller

    Recovery reconciles each case to billing for the audit committee: $49.7M found a year, $26.9M billed back this year from 2,960 closed cases, $18.4M collected (61% of the $30M target), and $21.7M billed each year from causes that were fixed.

    Wrong multiplier cases brought in the most: $7.4M paid.

Who it’s for

Five roles in utility revenue assurance, from the meter to the cash

One leak, five owners: how each worked the month before, and how they work it now.

Revenue assurance analyst
Prepares cases and back-bills
Before
Runs monthly exception reports on billing and meter data, with no network map, and spends most of a day building the evidence for one back-bill.
Now
Opens the buckets that moved this week and the premises behind them on the map, such as the 186 that changed this week, each with its proof.
Revenue assurance manager
Approves back-bills up to $250,000
Before
Sets which buckets the team works and owns the recovery target, and before signing a back-bill wants the network trace, the reads and the rule on one page.
Now
Signs with the rule check, the evidence pack and who prepared the case on the same screen.
Finance controller
Co-signs large back-bills, books recovery
Before
Reports leakage to the audit committee, where found is not recovered: what counts is billed back and cash in.
Now
Sees found, billed back and collected by bucket and month against the $30M target, and downloads the audit committee pack.
Internal audit manager
Tests back-bills each quarter
Before
Tests the back-billing rule and segregation of duties on a sample, and expects each number to open the record it came from.
Now
Sees each closed case tested against rule BR-7 and the approval limits, and draws a stratified sample with its evidence packs.
Billing operations lead
Owns account set-up fixes
Before
Owns account set-up and handover files, and needs to know which file broke and when, so the cause gets fixed and not only the case.
Now
Owns the causes that refill the buckets, with how much each refills a week and what the fix is.
The agents

Tracing, matching, pricing and root-cause agents for revenue assurance

For every premise, six agents join the network map, meter data and billing, price each gap and write the case; the revenue assurance manager approves.

Network tracer

Traces each premise to its transformer, feeder, zone and substation from the network map, with its energization date from the energizing service order.

  • Never changes the network map
  • A premise without a trace is listed, not guessed
Meter data reader

Reads consumption, read status, communication failures and estimates for each electricity and water meter, and flags meters with no actual read for the set number of days.

  • Never edits or replaces a read
  • Estimated reads are always labeled
Account matcher

Joins premise, meter, installation, account, tariff and service orders, runs the nine leakage checks and files each gap in one bucket, recording the source rows of each join.

  • Never opens, closes or changes an account
  • One premise sits in one bucket only
Revenue estimator

Prices each gap month by month at the tariff master (tiers, fuel cost adjustment, fees), applies the back-billing window of rule BR-7 and a collectability factor.

  • Never issues a bill
  • No figure without its tariff version and period
Root cause finder

Groups gaps into clusters (a tower, a firmware batch, a rule change), finds the event behind them in service orders, import logs and change records, and names the cause, the process it lives in and a proposed owner.

  • Never closes a cause
  • A cause names its evidence
  • Proposes an owner; the manager assigns
Case writer

Opens a case for each cluster worth more than $5,000 a year, with the three-way join, the value by month, the BR-7 window and the cause, and drafts the billing correction and the customer letter.

  • Never posts a back-bill
  • Never contacts a customer
  • Suspected tampering goes to revenue protection
Revenue assurance manager

Approves; finance co-signs above $250,000.

Ask in plain words

A leakage bucket, a premise or a back-bill case, asked in plain words

Ask in plain words about a bucket, a cluster, a premise or a case. Answers come from this week’s run and point to their records.

Why did consuming, no account jump this week?

Because of one tower. 186 premises were energized under one service order and have read each day since, but the developer’s handover file was rejected two days later with the tenant reference blank on all 186 rows. That is $370,600 unbilled to date and $612,000 a year.

Is the tower back-bill inside our back-billing rule?

Yes. Clause 9.2 of the developer’s connection agreement makes the developer the account holder of record from energization until a tenant takes over the supply, and rule BR-7 lets developer-held supply be billed from energization. The period is 211 days. At $370,600 it is above $250,000, so the finance controller co-signs after the revenue assurance manager approves.

Which multiplier case is worth the most?

A cold store. Its current transformers were changed to 400/5 but billing still holds 200/5, so only half its energy is billed. A field technician has confirmed the nameplates. That is $1,284,000 a year; under BR-7 the utility can back-bill 12 months, $1,284,000.

How much have we recovered this year?

$18.4M collected from $26.9M billed back this year, against a target of $30M. The best month brought in $3.9M since the weekly run started. Multiplier cases brought in the most, $7.4M.

The screens

The leakage waterfall, a bucket on the map, the case and recovery

From the working solution, on its sample data. Pick one to see it large.

The leakage waterfallThe year from delivered to metered, billed and collected, each step down a named bucket with its premises and this week’s change, and the weekly run agent by agent.
A bucket on the mapConsuming, no account: 1,240 premises as dots across eight districts, clusters ringed, with how the gap is found, its root causes and its clusters.
One premiseOne apartment traced from substation to premise, seven months of electricity and water reads, no account in billing, and the unbilled value priced tier by tier.
A leakage caseThe three-way join for each of 186 apartments, the unbilled value by month, the recovery under rule BR-7 and the sign-off.
How it leakedThe timeline from energization to the rejected handover file, the cause across 412 premises with its owner and fix, and how each agent found the case.
Approve the back-billThe manager signs a back-bill with the rule check, the evidence pack and who prepared it on one screen.
Billed, accounts openTwo signatures recorded with their times, the back-bill posted to the developer and 186 accounts open.
Leakage casesEach open case by bucket, premises, value a year, back-bill, stage from Found to Recovered, BR-7 window and owner.
Root causesEach cause with its bucket, where it leaks, premises, value a year, how much it refills a week, its owner and its fix: here the developer handover file, 31 premises a week, owned by billing operations.
RecoveryFound, billed back and collected against the $30M target, month by month and by bucket, and revenue now billed each year from fixed causes.
Leakage checksNine checks, each with the sources it joins, its threshold, hits this week, value a year and field-confirmed rate.
Audit evidenceEach closed back-bill tested against rule BR-7 and the approval limits, approvals by limit, and a random sample with its evidence packs.
Leakage rulesWhat counts as leakage, the back-billing rule, the approvers, the run schedule and the tariff version.
Governance

Segregated sign-off and back-billing rules behind each revenue assurance case

Each figure opens its recordEach number links to its source rows: the network map version, the meter read batch, the account record and the tariff version it was priced on.
Back-bills held to the ruleEach back-bill’s period is computed against rule BR-7 for its cause (12 months for the utility’s own error, from energization for developer-held supply), shown on the case and tested by internal audit.
Segregation of dutiesThe analyst prepares, the revenue assurance manager approves, and the finance controller co-signs above $250,000. Nobody approves a case they prepared.
Agents change no recordThe agents find, price, explain and draft. They do not change an account, a read or the network map, and a back-bill goes to billing only after the named approvers have signed.
Tampering goes to revenue protectionSuspected tampering is not back-billed from here: it is sent to revenue protection with its evidence, and this solution makes no customer contact.
Each step on the audit trailEach send, approval, co-signature, posting and fix is recorded with who did it and when, and each change to a setting is versioned and recorded.
Configuration

Leakage thresholds, back-billing windows and approvers in revenue assurance

What counts as leakage, the back-billing rule, the approvers and the run are settings, versioned and recorded on the audit trail.

SettingDefaultChoose from
Consuming — a premise with no account counts once it uses at least30 kWh in 30 daysSet with − and + between 10 and 60 kWh; the screen says how many premises would then be in the bucket
Meter never reads — no actual read for30 days21 days · 30 days · 45 days
Stopped estimate — an estimate below the premise’s own same month last year by40 %30 % · 40 % · 50 % below
Zone balance — flag a transformer zone above the sum of its premises by more than4 %3 % · 4 % · 5 %
Back-billing rule BR-7 — the utility’s error, customer not at fault12 months6 months · 12 months · 24 months, with the same number of monthly installments offered
Field check before a back-bill on a single premise above$50,000$25,000 · $50,000 · $100,000
Approves back-bills up to $250,000 a caseRevenue assurance managerRevenue assurance manager · Billing operations lead — the finance controller co-signs above
Weekly run — every premise, every checkMonday 06:00Monday 06:00 · Sunday 22:00 · Daily 05:00, plus a daily check on premises energized in the last 90 days
Connections

Network map, meter data, billing, service orders and the tariff master

Network map (GIS)connectivity from substation to feeder, transformer and premise, with energization
Meter data managementelectricity and water consumption, read status, communication failures and estimates
Customer information and billingaccounts, premises, installations, bills and account history
Service ordersenergizations, CT and meter changes, move-outs, reconnections and their fees
Tariff mastertiers, fuel cost adjustment and fees, pinned to one version per run
Import logs and change recordshandover files, rejected imports and rule changes, where a cause is found
What it changes

Revenue found, billed back and collected through revenue assurance

Each figure carries its label: a target the solution is built to, an estimate, a typical published result, or a proven one.

demo
$49.7M
a year of unbilled and under-billed revenue found and put on named premises, the sample's figure
Found by monthly reportsa few exceptions
With agents17,607 premises, meters and orders
target
100%
of back-bills checked against the back-billing rule, with trace, reads and tariff attached
2,960 back-bills
tested by internal audit
target
7days
to find a leak: every premise checked every week, not a sample once a year
Yearly audit sample≈ 12 months to find
With agents7 days to find

“demo” = seen in the working solution, on its sample utility data · “target” = the design goal, measured in the live solution · “estimated” = our estimate of the hours on one case (12.5 h by hand, 1.7 h with the solution, the sum of the hours split) · “typical” = published figures: industry estimates of meter-to-cash leakage at 1–5 % of revenue (Renewable Energy World; POWER); a 12-month back-billing limit for supplier error (Ofgem, 2018). Water context: AWWA M36 apparent losses. Names on the screens are sample data.

Questions

Utility revenue assurance questions from analysts, finance and audit

What is revenue assurance in utilities?

Revenue assurance is the work of finding electricity and water that is delivered but not billed, or billed too low, and recovering it. Revenue Assurance Analytics does it for each premise each week by joining the network map, meter data, accounts, service orders and the tariff master, and puts each gap on a named premise with a value at tariff.

How does revenue assurance software find unbilled consumption?

Nine leakage checks compare two or more sources for each premise, for example connected in the network map and consuming in meter data, with no active account in billing. Each gap is filed in one bucket, priced at tariff, grouped into clusters and opened as a case with its evidence. Each check shows its field-confirmed rate, and a check below 60% is sent to revenue protection instead of billing.

What kinds of revenue leakage does it find?

In the working solution: meters that do not read, consumption with no account, wrong CT multipliers, stopped estimates, consumption after move-out, wrong tariffs, service orders not billed, unexplained zone balances and suspected tampering. Suspected tampering is sent to revenue protection with its evidence and not back-billed from here.

Does it keep back-bills inside our back-billing rules?

Yes. The revenue estimator applies the back-billing window for each gap’s cause (in the sample utility, 12 months for its own error and from energization for developer-held supply), and the approver sees the rule check before signing. Internal audit sees each closed back-bill tested against the rule.

Does it work for water as well as electricity?

Yes. It reads electricity and water meters and prices both at tariff: in the sample case, 919,700 kWh of electricity worth $271,300 and $99,300 of water. For water these are apparent losses, water consumed that does not become revenue.

Who signs a back-bill, and can an analyst approve their own case?

No one approves their own case. The analyst prepares, the revenue assurance manager approves, and the finance controller co-signs above $250,000. A back-bill goes to billing only after those signatures. The agents do not change an account, a read or the network map, and do not contact a customer.

Can we set our own thresholds, windows and approval limits?

Yes. What counts as consuming, how long a meter can go without a read, the stopped-estimate and zone-balance thresholds, the back-billing windows, the field-check limit, the approvers and the run schedule are settings. Changes are versioned and apply from the next run.

What data does a first run on our own premises need?

Your back-billing rules, approval limits, tariff master and a sample of network, meter and billing data. The Agentic Solution Engine builds and deploys it from those requirements and documents, and it goes live once each quality gate has passed. We will walk you through it on your own premises first.

Point utility revenue assurance
at your own network and billing data.

We’ll run Revenue Assurance Analytics on a sample of your own network, meter and billing data.