[{"data":1,"prerenderedAt":199},["ShallowReactive",2],{"blog-\u002Fblog\u002Fexception-ageing":3,"blog-related-\u002Fblog\u002Fexception-ageing":175},{"id":4,"title":5,"audience":6,"body":10,"cluster":137,"conversionGoal":138,"cornerstone":139,"date":140,"description":141,"draft":139,"extension":142,"factCheckedAt":143,"faq":144,"featured":139,"language":143,"meta":154,"minRead":155,"navigation":156,"order":157,"originalAsset":143,"path":158,"pillar":159,"primaryKeyword":160,"publicationOrder":161,"relatedProject":143,"releaseScope":143,"reviewCycle":162,"reviewMethod":163,"reviewStatus":164,"reviewedAt":140,"reviewedBy":165,"searchIntent":166,"seo":167,"sources":143,"stem":168,"tags":169,"type":173,"updated":140,"__hash__":174},"blog\u002Fblog\u002Fexception-ageing.md","Exception Ageing: What Reconciliation Teams Don’t Report",[7,8,9],"transformation-lead","enterprise-architect","treasury-manager",{"type":11,"value":12,"toc":128},"minimark",[13,23,26,31,34,37,40,43,47,50,53,56,60,63,66,70,77,88,99,103,106,109,112,115],[14,15,16,17,22],"p",{},"Every ",[18,19,21],"a",{"href":20},"\u002Fblog\u002Fai-workflow-teardown-bank-reconciliation","reconciliation"," report I’ve seen leads with the same number: how many items are open. It’s an honest number and it tells you almost nothing. Three hundred open items can be a healthy process on a busy day or a quiet liability that has been growing for a year. The headline is identical.",[14,24,25],{},"The number that separates the two is age. How long has each item been open, and how is that distribution changing? This article is about that number: what it reveals, why teams don’t report it, and three changes a team can make in a week each.",[27,28,30],"h2",{"id":29},"what-age-reveals","What age reveals",[14,32,33],{},"An item open for two days is normal work. Someone will match it when the statement line arrives or the customer’s remittance comes in.",[14,35,36],{},"An item open for two weeks is cooling. The email that explained it is buried. The analyst who first looked at it has handled two hundred items since. Whatever context existed is being lost at a steady rate.",[14,38,39],{},"An item open for two months is no longer reconciliation. It’s an investigation, with a formal request to the bank, a search through someone’s inbox, and a growing chance that it ends as a write-off nobody fully understands. The five-minute match became a half-day search, and the half-day search becomes a line in the difference account.",[14,41,42],{},"So the same count means different things. Two hundred and ninety items from this week and ten from last quarter: a process that works, with a small tail to clean up. Ten from this week and two hundred and ninety from last quarter: a process that stopped working months ago and has been hiding behind a stable headline.",[27,44,46],{"id":45},"why-items-age","Why items age",[14,48,49],{},"Here is the uncomfortable part. Items don’t get old because they’re hard. Most old items were never particularly hard. They got old because nobody’s name was on them.",[14,51,52],{},"The queue is shared. Everyone clears what they recognise. What nobody recognises stays, and each day it stays it becomes slightly less likely that anyone will pick it up, because the context is colder and the easy items keep arriving. Ageing is an ownership problem wearing a reconciliation costume.",[14,54,55],{},"The second reason is tolerance that exists in practice but not in the system. Someone decided years ago that differences under a few euros aren’t worth chasing. The system still raises them. People park them. They age until month end, when they’re written off in a batch. The count stays high, the age climbs, and the decision that would fix both was made long ago and never written down.",[27,57,59],{"id":58},"why-teams-dont-report-it","Why teams don’t report it",[14,61,62],{},"Partly because the tools don’t make it easy: most systems show a list, not a distribution. Partly because count is comfortable. It moves every day, it responds to effort, and a good week shows. Age moves slowly and mostly in the wrong direction, and a growing tail is a statement about the process, not about this week’s work. Nobody enjoys reporting that.",[14,64,65],{},"But that is exactly why it’s the number worth reporting. It can’t be improved by clearing the easy items faster. It can only be improved by fixing something.",[27,67,69],{"id":68},"three-changes-one-week-each","Three changes, one week each",[14,71,72,76],{},[73,74,75],"strong",{},"Week one: report by age bucket, not by count."," Under seven days, seven to thirty, thirty to ninety, over ninety. Four numbers, and the oldest item’s age next to them. Most systems can export open items with a creation date; a pivot does the rest. Keep it next to the count for at least a month. The conversation changes by itself.",[14,78,79,82,83,87],{},[73,80,81],{},"Week two: a name and a date on everything past fourteen days."," Not a team. A person. If an item can’t be assigned, that is the finding: ",[18,84,86],{"href":85},"\u002Fblog\u002Ftreasury-exception-management","nobody owns this type of exception",", and the organisation should decide who does. The date is when the owner will next act, and the system, or the spreadsheet, reminds them.",[14,89,90,93,94,98],{},[73,91,92],{},"Week three: count the reasons."," For every item over thirty days, one line: why is this still open. Waiting for the bank. Waiting for the customer. Nobody knows what it is. Below tolerance, nobody wants to write it off. Cluster the lines. The largest cluster is the next rule: a tolerance to configure, a counterparty to call, a matching rule to add. The feedback step most ",[18,95,97],{"href":96},"\u002Fblog\u002Fdesigning-the-exception-path","exception processes"," never had.",[27,100,102],{"id":101},"the-metric-to-watch-after","The metric to watch after",[14,104,105],{},"Not the count. The age of the oldest open item, and the size of the over-ninety bucket. If both fall for three months while the count stays flat, and not because the tail was written off, the process is working. If the count falls and the tail grows, the team is clearing easy items faster and the problem is getting worse behind a better headline.",[14,107,108],{},"One number, reported honestly, is worth more than a dashboard. The age of what’s open is that number.",[110,111],"hr",{},[14,113,114],{},"A quick way to see your own distribution: export open items with their creation date, bucket them with a pivot table, and put the four buckets in one line in Monday’s report.",[14,116,117,118,122,123,127],{},"And if ",[18,119,121],{"href":120},"\u002Fblog\u002Fthe-spreadsheet-beside-the-system","the tracker you keep next to the system"," has grown a “Days open” column, an “Owner” column and a “Notes” column, paste its header row and a few rows into the ",[18,124,126],{"href":125},"\u002Ftools\u002Fspreadsheet-column-audit","Spreadsheet Column Audit",". It will tell you which of those columns are fields the system is missing, which are a decision log nobody can count, and which only repeat what the system already knows.",{"title":129,"searchDepth":130,"depth":130,"links":131},"",2,[132,133,134,135,136],{"id":29,"depth":130,"text":30},{"id":45,"depth":130,"text":46},{"id":58,"depth":130,"text":59},{"id":68,"depth":130,"text":69},{"id":101,"depth":130,"text":102},"method","spreadsheet-column-audit",false,"2026-11-12","In reconciliation, the open-item count tells you almost nothing. The age of open exceptions shows whether the process works. Three changes, one week each.","md",null,[145,148,151],{"question":146,"answer":147},"What is exception ageing in reconciliation?","Exception ageing is how long each open reconciliation item has been open, and how that distribution changes over time. The open-item count alone cannot tell a healthy busy day from a liability that has been growing for a year; the age distribution can. An item open for two days is normal work, one open for two weeks is losing its context, and one open for two months has become an investigation that may end as a write-off nobody fully understands.",{"question":149,"answer":150},"Why do reconciliation items get old?","Mostly not because they are hard. They get old because nobody’s name is on them: in a shared queue everyone clears what they recognise, and what nobody recognises stays. The second reason is a tolerance that exists in practice but not in the system, so small differences are parked until month end and written off in a batch.",{"question":152,"answer":153},"How should a reconciliation team report exception ageing?","By age bucket rather than by count: under seven days, seven to thirty, thirty to ninety and over ninety, with the age of the oldest open item next to them. An export of open items with their creation date and a pivot table are enough. After that, watch the age of the oldest item and the size of the over-ninety bucket: if both fall while the count stays flat, and not because the tail was written off, the process is working.",{},5,true,2.7,"\u002Fblog\u002Fexception-ageing","ai-workflow-design","exception ageing",263,"annual","editorial","reviewed","Tan Gravam","informational",{"title":5,"description":141},"blog\u002Fexception-ageing",[170,171,172],"workflow-design","process","controls","text","7o1ApECcilqVFrB0_a2Svy8LoqU5yJ6Gaij7ULlovBA",{"related":176,"bridge":143,"next":188,"guide":192,"place":195},[177,180,184],{"path":20,"title":178,"description":179},"AI Workflow Teardown: Bank Reconciliation","Bank reconciliation taken apart: why the matching engine was never the bottleneck, what AI does with the residue, and which breaks need a named human.",{"path":181,"title":182,"description":183},"\u002Fblog\u002Fhuman-decision-rights-in-ai-native-workflows","Human Decision Rights in AI-Native Workflows","The useful question is not what AI can automate. It is who is allowed to decide what — a rights matrix over every action a workflow contains.",{"path":185,"title":186,"description":187},"\u002Fblog\u002Fai-automation-vs-augmentation-vs-autonomy","AI Automation vs Augmentation vs Autonomy","AI automation vs augmentation vs autonomy are three different designs, not rungs of a ladder. The criteria that decide which one a piece of work deserves.",{"path":189,"title":190,"description":191},"\u002Fblog\u002Fai-workflow-teardown-cash-forecasting","AI Workflow Teardown: Cash Forecasting","A corporate cash forecast, taken apart: where the elapsed time really sits, what stays deterministic, what AI genuinely does, and what a named human decides.",{"slug":159,"name":193,"total":194},"AI Workflow Design",9,{"label":196,"position":197,"total":197,"hub":198},"Method",7,"\u002Ftopics\u002Fai-workflow-design#cluster-method",1794508944869]