Task Records
When a specialist watches something over time — support tickets, sales leads, incoming orders — it keeps one record per item with a current status, so it always knows what it has already seen and what's genuinely new.
One record per item
Say a specialist is monitoring a queue of support tickets. The first time it sees a ticket, it opens a record for it and gives that record a status — say, "new." The next time the specialist runs and comes across that same ticket again, it recognizes it and updates the existing record instead of creating a second one for the same ticket.
This is what keeps counts honest: a ticket that's been looked at three times still shows up as one item, not three.
Status moves the item forward
Each record's status changes as the item moves through its life — a ticket might go from "new" to "in progress" to "resolved," a lead from "contacted" to "qualified" to "closed." A specialist reports on where things stand by reading these statuses, not by re-reading everything from scratch on every run.
A daily reconcile
Once a day, a monitoring specialist compares its records against what's actually there — the real list of open tickets, live leads, or pending orders. If a record still says something is open but it's actually been closed elsewhere, the reconcile step catches the mismatch and corrects the record. This keeps the records trustworthy even when something changes outside of a specialist's own runs.
Why this matters for your reports
A report only states a number when there's a record behind it. If a specialist tells you "12 tickets are open," that means 12 individual records currently carry an open status — not an estimate and not a re-count of raw messages. This is also why the same item never gets reported twice: it has exactly one record, no matter how many times the specialist has looked at it.
The same item, seen more than one way
An item can resurface through more than one path — a ticket that comes in by email and then gets mentioned again in a form submission, or a lead that shows up on a list you shared and again through a follow-up message. As long as a specialist can tell it's the same underlying item, it still gets matched back to the one record instead of starting a new count. This is also why an item can safely be checked more than once in a day without inflating anything you're told.
Not just tickets
The same approach applies anywhere a specialist is asked to keep an eye on a changing set of things over time — a queue of leads working through a sales process, orders moving from placed to shipped to delivered, or a list of open items on a project. In every case, the pattern is the same: one item, one record, one current status, updated in place rather than piled on top of the last one.
What this looks like day to day
You don't manage records directly — a specialist keeps them as part of doing its job. What you see is the result: steady counts that only move when something real changes, lists without duplicates, and reports that still hold up when you go check them against reality yourself. If a number in a report ever looks off, the first thing worth checking is whether the specialist has run its daily reconcile recently — that's usually what re-aligns the records with what's actually true.