WEBVTT NOTE IDER in-house training materials. Captions follow measured phrase audio timings. Hypothetical worked cases; figures are instructional examples. 00:00:00.000 --> 00:00:05.580 A business reports twelve million Ghana cedis of revenue and strong growth. Your assignment is to 00:00:05.580 --> 00:00:08.719 determine what that figure says about durable commercial performance. 00:00:08.839 --> 00:00:15.005 We will not begin by applying a convenient valuation multiple. We will reconstruct what was earned, 00:00:15.005 --> 00:00:20.400 identify which activity is plausibly repeatable, examine customer retention and trace the route to 00:00:20.400 --> 00:00:20.786 cash. 00:00:20.906 --> 00:00:27.236 Every figure in this IDER in-house clinic is hypothetical. Keep separate columns for accounting 00:00:27.236 --> 00:00:29.044 corrections and commercial adjustments. 00:00:29.164 --> 00:00:34.925 That separation prevents a legitimate one-off sale from being described as an accounting error, and 00:00:34.925 --> 00:00:39.149 prevents a recurring invoice from being accepted without evidence of delivery. 00:00:39.269 --> 00:00:45.814 By the end, you will have a practical investigation sequence and a defensible explanation of unresolved 00:00:45.814 --> 00:00:46.223 risks. 00:00:46.893 --> 00:00:53.618 Four numbers can all be different without any contradiction. Reported revenue is the starting claim. 00:00:53.618 --> 00:00:58.550 Correctly recognised revenue reflects the appropriate treatment of actual contractual delivery. 00:00:58.670 --> 00:01:05.274 A recurring analytical base removes activities you do not expect to repeat under an explicit definition. 00:01:05.274 --> 00:01:09.815 Customer collections measure cash received, including amounts earned in earlier periods. 00:01:09.935 --> 00:01:15.392 Ask which question each number answers before comparing them. A signed order may support future 00:01:15.392 --> 00:01:18.666 demand but does not automatically establish current earned revenue. 00:01:18.786 --> 00:01:25.708 Equally, receiving an advance does not prove that a service has already been delivered. Build a dictionary 00:01:25.708 --> 00:01:30.593 that specifies period, currency, tax treatment and inclusion rules for each view. 00:01:30.713 --> 00:01:34.928 Most avoidable disputes begin with those definitions being left implicit. 00:01:35.598 --> 00:01:43.410 Request a transaction-level extract that can connect the commercial and financial systems. You need stable 00:01:43.410 --> 00:01:49.659 customer identifiers, contract references, invoice and credit-note identifiers, service periods and payment 00:01:49.659 --> 00:01:50.180 links. 00:01:50.300 --> 00:01:55.265 Keep the original extract immutable and perform transformations in a reproducible working layer. 00:01:55.265 --> 00:02:01.759 Before analysing cohorts, reconcile row counts and totals to the ledger for the same period and scope. 00:02:01.879 --> 00:02:08.283 Investigate missing identifiers and duplicate invoice lines. A many-to-many join between invoices and 00:02:08.283 --> 00:02:12.223 payments can multiply revenue while appearing technically successful. 00:02:12.343 --> 00:02:17.835 Test the expected relationship before joining and retain unmatched records for review. 00:02:17.955 --> 00:02:24.941 This transaction spine is your audit trail: every attractive chart should be traceable back to an 00:02:24.941 --> 00:02:27.998 identifiable set of records and documented transformations. 00:02:28.668 --> 00:02:34.485 Choose transactions around the reporting boundary and follow the evidence chain. Read the relevant contract 00:02:34.485 --> 00:02:40.691 terms, identify what was promised and inspect evidence that the promised delivery occurred in the reported 00:02:40.691 --> 00:02:41.079 period. 00:02:41.199 --> 00:02:47.279 Invoice date alone is insufficient. A subsequent credit note may reveal a cancellation or correction, 00:02:47.279 --> 00:02:50.521 but it needs interpretation rather than automatic subtraction. 00:02:50.641 --> 00:02:57.993 For bundled services, establish whether the recorded allocation and timing reflect the actual obligations; 00:02:57.993 --> 00:03:02.193 escalate technical accounting judgments to the responsible professional. 00:03:02.313 --> 00:03:08.382 Your research role is to surface and organise the evidence, quantify clearly supported exceptions and 00:03:08.382 --> 00:03:10.405 distinguish them from open questions. 00:03:10.525 --> 00:03:16.651 Preserve the original amount, proposed adjustment, reason, supporting documents and reviewer status for 00:03:16.651 --> 00:03:18.536 every exception you report. 00:03:19.206 --> 00:03:25.900 Our hypothetical review identifies six hundred thousand cedis recorded in the wrong period and a 00:03:25.900 --> 00:03:32.146 separate four hundred thousand that must be reversed. Corrected recognised revenue is therefore eleven 00:03:32.146 --> 00:03:32.593 million. 00:03:32.713 --> 00:03:39.176 Within that corrected amount, one million comes from a valid one-off project. Excluding it produces a 00:03:39.176 --> 00:03:42.407 ten-million recurring analytical base. Notice the two bridges. 00:03:42.527 --> 00:03:49.087 The first corrects recognition; the second changes the lens for forecasting. The one-off project remains 00:03:49.087 --> 00:03:54.334 valid recognised revenue, and removing it does not create an accounting correction. 00:03:54.454 --> 00:04:00.046 State the evidence supporting its non-recurrence and check whether similar projects regularly replace one 00:04:00.046 --> 00:04:00.445 another. 00:04:00.565 --> 00:04:06.459 A business built on successive projects may have repeatable commercial activity even when individual 00:04:06.459 --> 00:04:08.564 contracts are not recurring subscriptions. 00:04:09.234 --> 00:04:15.855 Retention asks what happened to the customers you already had. Freeze the opening cohort and its 00:04:15.855 --> 00:04:19.579 revenue, then classify each customer's change using consistent rules. 00:04:19.699 --> 00:04:26.534 Our monthly opening revenue is one hundred thousand cedis. Existing customers lose ten thousand through 00:04:26.534 --> 00:04:32.458 complete churn and five thousand through reduced spending, while expansion adds twenty thousand. 00:04:32.578 --> 00:04:37.809 New customers contribute a separate thirty thousand. Keep new customers outside the retention numerator 00:04:37.809 --> 00:04:40.798 because they were absent from the opening denominator. 00:04:40.918 --> 00:04:46.585 Track reactivations and mergers under a documented policy so that identifiers cannot manufacture 00:04:46.585 --> 00:04:49.637 retention. Also align currency and billing periods. 00:04:49.757 --> 00:04:56.135 A price increase or exchange-rate movement can lift revenue retention even when usage or customer counts 00:04:56.135 --> 00:04:59.324 weaken, so investigate the mechanism behind the movement. 00:04:59.994 --> 00:05:06.243 Pause and calculate three measures from the monthly cohort. First find gross revenue retention, which 00:05:06.243 --> 00:05:10.826 deducts churn and contraction from the opening base while excluding expansion. 00:05:10.946 --> 00:05:16.910 Then find net revenue retention, which includes expansion by those same opening customers. Finally 00:05:16.910 --> 00:05:22.022 calculate the total closing monthly revenue run-rate, now including the new customers. 00:05:22.142 --> 00:05:28.729 Write the population and period beside each answer. Explain why a closing run-rate should not be described 00:05:28.729 --> 00:05:31.441 as revenue earned throughout the entire year. 00:05:31.561 --> 00:05:37.487 A useful way to check your calculation is to build a bridge from the opening amount to the closing amount 00:05:37.487 --> 00:05:38.376 before dividing anything. 00:05:38.496 --> 00:05:44.658 That bridge exposes missing categories and helps distinguish a retention improvement from growth 00:05:44.658 --> 00:05:46.554 purchased through new-customer acquisition. 00:05:47.224 --> 00:05:53.547 Gross retention is eighty-five thousand divided by one hundred thousand, or eighty-five percent. Add twenty 00:05:53.547 --> 00:05:59.449 thousand expansion from existing customers and net retention becomes one hundred and five percent. 00:05:59.569 --> 00:06:04.025 Including thirty thousand from new customers gives a closing monthly run-rate of one hundred and 00:06:04.025 --> 00:06:04.619 thirty-five thousand. 00:06:04.739 --> 00:06:10.230 Net retention above one hundred percent is compatible with losing customers and revenue elsewhere in the 00:06:10.230 --> 00:06:10.916 opening cohort. 00:06:11.036 --> 00:06:16.895 That is why you should inspect gross retention and the distribution of expansion, not celebrate the net 00:06:16.895 --> 00:06:17.584 figure alone. 00:06:17.704 --> 00:06:23.242 If one large account supplies nearly all expansion, the apparent strength may be concentrated. 00:06:23.362 --> 00:06:29.161 And multiplying the closing run-rate by twelve creates an annualised snapshot, not a reconstruction of what 00:06:29.161 --> 00:06:31.698 the business actually earned during the year. 00:06:32.368 --> 00:06:38.649 To interpret growth, decompose it into customer counts, usage and realised price wherever the business 00:06:38.649 --> 00:06:44.512 model permits. A price increase can temporarily improve revenue while future renewal risk rises. 00:06:44.632 --> 00:06:51.778 Compare acquisition cohorts at the same age, because a mature cohort and a newly acquired cohort have had 00:06:51.778 --> 00:06:56.541 different opportunities to churn. Include lost customers rather than analysing only surviving 00:06:56.541 --> 00:06:56.938 accounts. 00:06:57.058 --> 00:07:02.177 Then examine concentration by ultimate commercial group, not only by billing identifier. Ten 00:07:02.177 --> 00:07:06.509 subsidiaries under one purchasing decision may behave like a single exposure. 00:07:06.629 --> 00:07:13.087 Interview selected retained and lost customers using a documented selection process. Those interviews can 00:07:13.087 --> 00:07:19.545 explain mechanisms, but a small convenient sample does not estimate population-wide churn reasons with 00:07:19.545 --> 00:07:20.468 statistical precision. 00:07:21.138 --> 00:07:27.776 Assume the eleven million corrected revenue consists entirely of credit sales and uses the same scope as 00:07:27.776 --> 00:07:32.852 receivables. Opening receivables are one point two million and closing receivables two million. 00:07:32.972 --> 00:07:39.373 With no other movements, collections equal opening receivables plus credit revenue minus closing 00:07:39.373 --> 00:07:43.805 receivables, giving ten point two million. The assumption matters. 00:07:43.925 --> 00:07:48.569 Taxes, foreign exchange, write-offs, reclassifications, unbilled balances or factoring can 00:07:48.569 --> 00:07:55.070 require separate bridge lines. Do not insert a balancing plug and call it cash. 00:07:55.190 --> 00:08:00.647 Reconcile the inferred collection amount to matched bank receipts and investigate differences. This 00:08:00.647 --> 00:08:07.783 exercise tests consistency among records; it does not prove every sale was genuine or every remaining debtor 00:08:07.783 --> 00:08:08.623 will pay. 00:08:08.743 --> 00:08:11.924 Those claims require their own evidence and review. 00:08:12.594 --> 00:08:18.788 An ageing schedule becomes useful when you understand what starts its clock. Age invoices from the 00:08:18.788 --> 00:08:23.821 contractual due date when assessing lateness, and retain invoice age separately if needed. 00:08:23.941 --> 00:08:29.775 Trace large overdue items to correspondence, disputes, credit notes and subsequent payments. Check that the 00:08:29.775 --> 00:08:35.609 cash was applied to the same invoice rather than merely received from the same customer. 00:08:35.729 --> 00:08:42.114 An advance for a new contract does not settle an old disputed invoice unless the allocation is valid. 00:08:42.234 --> 00:08:48.152 Analyse ageing by customer and cohort as well as in aggregate; new growth can dilute an old delinquency 00:08:48.152 --> 00:08:48.481 problem. 00:08:48.601 --> 00:08:54.573 Record the review date so that subsequent collections are not accidentally presented as cash available at 00:08:54.573 --> 00:08:56.066 the historical reporting date. 00:08:56.736 --> 00:09:02.887 A focused review should cover individually material accounts and transactions with unusual timing, terms 00:09:02.887 --> 00:09:09.917 or adjustments. Add a documented representative sample if you need evidence about the wider population. These 00:09:09.917 --> 00:09:10.795 purposes differ. 00:09:10.915 --> 00:09:16.546 A deliberately high-risk sample may discover problems efficiently, but its exception rate cannot be 00:09:16.546 --> 00:09:20.970 mechanically extrapolated to all revenue as though the sample were random. 00:09:21.090 --> 00:09:27.920 Retain the sampling frame, selection method and tested amount. Distinguish confirmed adjustments, plausible 00:09:27.920 --> 00:09:35.802 exposure ranges and matters awaiting evidence. Where several tests identify the same invoice, count the 00:09:35.802 --> 00:09:37.378 underlying exposure once. 00:09:37.498 --> 00:09:45.330 The purpose is an interpretable risk assessment, not a dramatic sum of overlapping flags. Ask what additional 00:09:45.330 --> 00:09:50.397 evidence could resolve each material uncertainty before recommending a transaction condition. 00:09:51.067 --> 00:09:57.210 Pause and assess a colleague's claim: receivables increased by eight hundred thousand cedis, therefore 00:09:57.210 --> 00:10:03.353 eight hundred thousand of revenue is fraudulent. First calculate collections under the clean bridge 00:10:03.353 --> 00:10:03.792 assumptions. 00:10:03.912 --> 00:10:10.584 Then explain why the claim goes beyond the available evidence. Propose three follow-up tests that could 00:10:10.584 --> 00:10:14.337 distinguish normal commercial timing from collection or recognition problems. 00:10:14.457 --> 00:10:20.639 Your answer should connect each test to a competing explanation, such as longer contractual terms, a 00:10:20.639 --> 00:10:25.661 late-period sales increase or a disputed delivery. Do not merely request more documents. 00:10:25.781 --> 00:10:30.750 Specify which records would change your conclusion and how you would reconcile them. 00:10:30.870 --> 00:10:36.376 This is the central discipline of diligence: move from a signal to a testable explanation before moving from 00:10:36.376 --> 00:10:37.906 an explanation to a finding. 00:10:38.576 --> 00:10:44.635 The clean bridge gives ten point two million cedis of collections. An eight-hundred-thousand increase in 00:10:44.635 --> 00:10:50.290 receivables says that recognised credit sales exceeded collections by that amount under our assumptions. 00:10:50.410 --> 00:10:57.779 It does not identify the cause. Compare agreed payment terms and sales timing, inspect overdue ageing, and 00:10:57.779 --> 00:11:00.813 trace subsequent receipts to the actual invoices. 00:11:00.933 --> 00:11:06.835 Review disputed deliveries and period-end credits where recognition may be affected. Finally confirm 00:11:06.835 --> 00:11:13.644 that tax, currency and other balance movements have been treated consistently. Report what each test 00:11:13.644 --> 00:11:14.098 establishes. 00:11:14.218 --> 00:11:19.455 Subsequent payment supports collectibility of the matched invoice, but it does not automatically resolve 00:11:19.455 --> 00:11:22.821 whether the sale belonged in the original reporting period. 00:11:22.941 --> 00:11:28.247 Keep those conclusions separate even when the same document informs both investigations. 00:11:28.917 --> 00:11:35.739 Build the conclusion around the decision, not around a catalogue of documents reviewed. Present the 00:11:35.739 --> 00:11:39.378 reported-to-corrected revenue bridge, then the separate recurring-base bridge. 00:11:39.498 --> 00:11:46.753 Show gross and net retention for consistent cohorts, highlight concentration, and reconcile collections 00:11:46.753 --> 00:11:54.566 with clearly stated limitations. For each material finding, explain the amount, mechanism, evidence and 00:11:54.566 --> 00:11:55.682 commercial consequence. 00:11:55.802 --> 00:12:02.698 An unresolved customer renewal may affect a forecast scenario; a confirmed cutoff error affects historical 00:12:02.698 --> 00:12:04.997 recognition. They need different responses. 00:12:05.117 --> 00:12:11.235 Identify the professional responsible for technical sign-off and retain the actual review status rather 00:12:11.235 --> 00:12:16.042 than implying approval. Attach owners and deadlines to outstanding evidence requests. 00:12:16.162 --> 00:12:22.477 A concise conditional conclusion is more useful than a confident headline that hides an unfinished 00:12:22.477 --> 00:12:24.161 investigation inside an appendix. 00:12:24.831 --> 00:12:29.208 For your next assignment, begin with a small reproducible slice of the data. 00:12:29.328 --> 00:12:35.754 Select one reporting month, reconcile it to the ledger, trace a few transactions through delivery and 00:12:35.754 --> 00:12:38.164 payment, and rebuild one customer cohort. 00:12:38.284 --> 00:12:45.378 This pilot exposes definition and join problems before they contaminate a much larger analysis. Then expand 00:12:45.378 --> 00:12:52.030 using documented rules and risk-based review. Complete the assessment below and the two applied tasks. 00:12:52.150 --> 00:12:57.902 Your work should show that a recurring label, a strong growth rate and a cash receipt answer different 00:12:57.902 --> 00:12:58.222 questions. 00:12:58.342 --> 00:13:04.488 Assess yourself on whether another practitioner can follow the calculation and identify exactly what 00:13:04.488 --> 00:13:05.366 remains unproven. 00:13:05.486 --> 00:13:10.902 The practical outcome is a decision supported by evidence, with commercial uncertainty expressed 00:13:10.902 --> 00:13:12.568 clearly enough to manage.