Daily glance, weekly review, monthly close

A working reporting cadence for a 5–50 person B2B firm has exactly three loops: a daily glance that takes under two minutes, a weekly review that takes under an hour, and a monthly close that settles the record for good. Each loop answers a different question — is anything on fire, are we on track, and what actually happened. If any of the three requires someone to compile numbers by hand, the cadence will not survive its first busy fortnight.

Why do most firms report too much or not at all?

Most firms sit at one of two extremes. At one end, no reporting: the MD asks how the pipeline looks and triggers three days of spreadsheet archaeology, ending in a number nobody quite trusts. At the other, reporting theatre: a weekly deck that takes an operations person half a day to assemble and is stale before anyone opens it.

Both extremes share a root cause — reporting treated as a task rather than a system. When a number has to be compiled, then it arrives late, costs labour, and passes through the judgement of whoever compiled it. The fix is not more discipline; it is a dashboard that reads live from the systems where the work happens, which is the whole argument of The MD Dashboard Blueprint. The cadence in this article assumes that dashboard exists. Once reading the numbers costs nothing, the only questions left are when to read them and what each reading is allowed to decide.

What belongs in the daily glance?

The daily glance is exceptions only, and it should take under two minutes. Nothing that reads as a trend belongs here. The test is simple: when a number demands action today, then it earns a place on the daily view; when it does not, it moves down to weekly or monthly.

For a typical B2B service firm, the daily list is short:

• New enquiries not yet contacted. Speed decides outcomes here — contact rates drop roughly 8x after five minutes, as the industry rule of thumb has it.
• Replies awaiting handling. In outbound, replies get handled daily or the campaign quietly dies; the mechanics are covered in The Complete UK B2B Outbound Playbook.
• Deals with an overdue next action. Not slow deals — deals where the agreed next step has passed its date.

If the daily glance is green, you close the tab. Its job is to make silence trustworthy: no news genuinely meaning no news.

What does the weekly review decide?

The weekly review is where decisions live. Thirty to sixty minutes, run against the pipeline view, same slot every week. The discipline that makes it work is a verdict rule: every deal that has sat in its stage longer than the agreed threshold gets one of three outcomes — advance it, recycle it into nurture, or close it lost. There is no fourth option, and "leave it another week" is not a verdict.

The weekly loop also checks activity against plan: did the sends go out, did the follow-ups happen, did booked calls take place. Most firms stop at two follow-up touches while deals typically need five or more; the weekly review is where that neglect becomes visible early enough to correct, rather than showing up months later as a soft quarter.

What does the monthly close settle?

Monthly is trend truth: win rate, pipeline coverage against target, average deal size, where lost deals actually died. These numbers are meaningless day to day and essential month to month. The monthly loop is also a close in the accounting sense — once it is done, the month's numbers stop moving, and any later correction is made visibly rather than silently.

The monthly slot carries the hygiene sweep too: duplicates merged, dead deals archived, ownerless records reassigned. Trends are only worth reading when the record underneath them is clean, and the record is only clean when every fact has exactly one home — the argument made in One system of record per fact.

How do you install the cadence?

The mechanism, step by step:

1. List the numbers you actually use, then sort each into a loop. When a number requires action today, then it is daily; when it informs a decision this week, then it is weekly; when it only means anything as a trend, then it is monthly.
2. Build one view per loop in the CRM or dashboard. When a view needs manual input before it renders truthfully, then either fix the data flow or delete the number — no compiling, ever.
3. Put the loops in the calendar with named owners: the daily glance first thing each morning, the weekly review in a fixed slot, the monthly close on the first working day of the month.
4. Attach the decision rules to each loop — the exception list, the three-verdict rule, the close — so each session produces verdicts rather than commentary.
5. Run it for a month, then prune. Any number nobody acted on gets demoted or dropped.

None of this depends on expensive software. We run our own cadence on an open-source CRM, and the trade-offs of that choice are written up honestly in Running an open-source CRM: the honest experience. The cadence is the system; the tool is just where it lives.

---

Next step: the Growth System Audit — £450, seven days, credited against any build — maps where your growth system leaks and what to build first.