Scheduled Reports: Getting the Right Numbers to the Right People
The gap between a report that exists and a report that reaches the person who acts on it is most of the value. Scheduling, who to send to, and the discipline that stops scheduled reports becoming ignored mail.
There is a quiet asymmetry in reporting that most organizations never address. Enormous effort goes into building reports and almost none into getting them to people. The result is a system full of good reports that require somebody to remember to log in, navigate to them and run them — which happens dependably when the week is quiet and stops entirely when it is not.
That is exactly backwards. The weeks when nobody has time to check the numbers are the weeks the numbers matter most. Scheduling fixes this by removing the memory requirement, and in our experience it is the single highest-return change available in reporting — larger than any improvement to the reports themselves.
Pull versus push, concretely
Pull — they come to it
- Requires remembering, logging in and navigating
- Happens when things are calm; stops when they are busy
- The person who most needs it is the least likely to look
- A report's value depends on one person's habit
- Nobody notices when it stops being read
Push — it comes to them
- Arrives whether or not anyone remembers
- Survives a bad quarter, which is when it counts
- Reaches people who would never log in — a director, a branch manager
- Value depends on a schedule, not on a habit
- Absence is noticeable, so breakage gets reported
The person who most needs a number is usually the person least likely to log in and look for it. Push, always.
Send to roles, not just to people
A practical detail that saves a great deal of maintenance. If a report goes to three named individuals, it silently breaks the day one of them changes job — the report keeps arriving to the wrong inbox and stops arriving to the right one, and nobody notices until a decision is missed.
Sending to a role instead means the distribution list maintains itself as people move. AWRA supports recipients as named users, as roles, or as plain email addresses for people outside the system — and the role option is the one to prefer wherever the report is genuinely about a function rather than a person.
-
Decide the decision first, then the recipient
A report is for somebody who will do something. Start from the action, then work back to who takes it, and only then to how they should receive it.
-
Prefer a role over a name
Branch managers, finance, procurement leads. Roles survive staff changes; names do not, and a broken distribution list is invisible.
-
Match cadence to the decision
Weekly for anything operational, monthly for anything commercial. A monthly report supporting a weekly decision is history.
-
Choose a format people will actually open
Something readable on a phone for a quick scan; a spreadsheet only where the recipient genuinely works the data.
-
Include external recipients deliberately
Accountants, auditors, board members, sometimes a client. Sending directly beats somebody exporting and forwarding it by hand each month.
The failure mode: scheduled mail nobody opens
Scheduling has one characteristic failure, and it is worth designing against from the start. Reports get scheduled generously — to everybody who might conceivably want them, on the most frequent cadence that seems defensible — and within two months they are filtered into a folder nobody reads. The report is now worse off than before, because it is technically delivered and genuinely ignored.
Three rules prevent it, and all three are about restraint rather than technology.
How to keep a schedule alive
- Send only to people who act. Observers can pull. Every additional recipient who cannot do anything about the number dilutes the meaning of receiving it.
- Send as rarely as the decision allows. Daily is almost never right. Weekly for operations, monthly for commercial review, and stop there.
- One report per email, with the point in the subject line. A digest of six reports gets skimmed once and filtered thereafter.
- Include the threshold in the subject where you can. "Receivables over 60 days: 2.8m" is opened; "Weekly AR report" is filed.
- Review recipients quarterly. People move roles, and a list that only ever grows becomes a list nobody reads.
- Notice when a schedule breaks. If a report stops arriving and nobody mentions it for a month, that is your answer about whether it was needed.
The absence test
Turn off a scheduled report and see whether anybody asks for it. Uncomfortable, cheap, and definitive. Anything nobody misses within a month was not a report, it was a habit — and cancelling it makes room for the ones that matter to be noticed.
What to schedule to whom
An illustrative distribution, showing the shape rather than a template to copy.
| Report | To | Cadence |
|---|---|---|
| Stock cover on A-class items, with anything under two weeks flagged | Procurement role + branch managers role | Weekly |
| Receivables ageing with accounts over 60 days first | Finance role + the account owners | Weekly |
| Committed vs available by project line | Project leads role | Weekly |
| Tickets by category with month-on-month change | Support lead + IT | Monthly |
| Margin by customer, bottom ten | Sales lead + directors | Monthly |
| Management pack — the six numbers together | Directors, plus the external accountant | Monthly |
Note the last row. Scheduling the monthly pack to your external accountant directly is a small change that removes a recurring manual export-and-email task, and it means the person advising you is looking at the same figures as the board rather than a version somebody assembled by hand.
What we do and do not do
What AWRA OpsHub does today
- Schedules on saved report definitions and saved filters, so the scheduled version is the view you actually built.
- Cadence control — including day of week and day of month, so a report lands when the decision is made.
- Recipients as named users, as roles, or as plain email addresses for people outside the system.
- Multiple formats per schedule.
- Delivery records, so you can see what was sent and when rather than assuming.
- Exports for the cases where a person genuinely needs the data elsewhere.
What it does not do
- No conditional delivery. A schedule sends on its cadence; it does not send only when a threshold is breached.
- No per-recipient data filtering from one schedule — if a branch manager should see only their branch, that is a schedule per branch rather than one schedule that personalises itself.
- No in-email interactivity — recipients read the report; they do not drill into it from the message.
- No delivery to Slack or WhatsApp as report destinations. Slack notification exists for helpdesk tickets, which is a different thing.
The first two shape how you design schedules: expect to create one schedule per audience-and-filter combination rather than one clever schedule that adapts, and do not rely on silence meaning "nothing breached".
One schedule per audience
Because a schedule does not personalise its data per recipient, the pattern that works is one schedule per audience-and-filter pair: a stock report filtered to Mombasa going to the Mombasa manager, another filtered to Nakuru going to Nakuru. That feels repetitive and is actually preferable — each recipient gets a report about their own responsibility rather than a company-wide list they have to find themselves in.
A report that requires the recipient to locate their own rows will be skimmed. A report that contains only their rows gets read, and the small duplication in setup buys that. Saved filters make it a few minutes of work per branch rather than a rebuild.
Our take
Schedule everything you expect anyone to act on, send to roles rather than names, and send as rarely as the decision allows. Then apply the absence test once a quarter — switch a report off and see whether anybody notices. The reports that survive that test are your actual reporting, and there will be fewer of them than you think.
See reports arrive without being asked
Schedules on your saved definitions and filters, cadence down to the day, recipients as users, roles or external addresses, and a record of what was delivered.
Explore scheduled reportsFrequently asked questions
Why schedule rather than let people run reports themselves?
Because pulling a report requires remembering to do it, and that memory fails exactly when the week is busy — which is when the numbers matter most. Scheduling removes the memory requirement so the report arrives whether or not anyone thinks of it. It also reaches people who will never log in, such as directors and branch managers, who are frequently the people whose decisions the report was built to inform.
Should we send to individuals or to roles?
Prefer roles wherever the report is about a function rather than a person. A list of three named individuals silently breaks the day one of them changes job — the report keeps arriving in the wrong inbox and stops arriving in the right one, and nobody notices until a decision is missed. Roles maintain themselves as people move. Plain email addresses are available for external recipients like your accountant.
Can a schedule send only when something is wrong?
No — schedules send on their cadence rather than conditionally on a threshold being breached. That is worth knowing when you design your reporting, because it means silence does not mean "nothing breached"; it means the report has not been sent yet. Design your six numbers around a regular rhythm with a human reading them, rather than around exception-only alerting.
Can one schedule send each branch manager only their own branch?
Not from a single schedule — it does not filter per recipient. The working pattern is one schedule per audience-and-filter pair: a report filtered to Mombasa going to the Mombasa manager, another filtered to Nakuru going to Nakuru. It feels repetitive but is genuinely better, because each recipient receives a report about their own responsibility rather than a company-wide list they have to find themselves in.
How do we tell whether a scheduled report is being read?
Turn it off and see whether anybody asks for it. It is uncomfortable and definitive: anything nobody misses within a month was a habit rather than a report. Do this once a quarter, along with a review of recipients, because distribution lists only ever grow and a long list of ignored mail makes the reports that do matter harder to notice.
