Why do internal reports outlive the reason they were written?
Nearly every recurring internal report began as an answer to one question. A partner wanted to know why new instructions had dropped off, so a weekly pipeline sheet appeared. A lender asked about work in progress during a facility review, so a monthly WIP summary was added. A job went wrong, so a quality pack was introduced to make sure it never happened again.
The question gets answered. The review ends. The person who worried about that job moves on. The report carries on, because nobody ever cancels a report. Cancelling carries a small reputational risk and no reward. Producing it carries a real cost, but that cost falls on whoever assembles it, usually a junior fee earner or a practice manager, and it is spread thinly enough across the month that it never reaches an agenda.
So the stack grows. A firm of forty people can be producing a dozen recurring internal documents, and the people reading them are rarely the people who commissioned them. It is the quiet version of the pattern in the admin tax on growth: each year adds reporting, and nothing is ever taken away.
How can you tell which reports are actually read?
Opinion is useless here. Ask whether a report is useful and almost everyone says yes, because saying no sounds like criticism of whoever introduced it. You need evidence, and there are three cheap sources of it.
The first is the silence test. Stop circulating the report for a month. Do not announce it. Note who asks where it is. If nobody asks, you have your answer. If one person asks, you have found your reader, and you can now ask them what they do with it.
The second is the decision trail. Go back through six months of management meeting notes and look for a decision that turned on a figure from the report. Not a mention, a decision. If the pack is discussed every month but has never changed what the firm did, it is a ritual rather than a tool.
The third is duplication. Lay the reports side by side and mark every figure appearing in more than one. Fee income, utilisation, WIP and debtor days tend to turn up three or four times in slightly different cuts, each assembled separately, each capable of disagreeing with the others. Disagreement is worse than absence, because somebody then spends an afternoon reconciling two versions of a number nobody intended to act on.
What questions expose a report nobody needs?
Take everything the firm produces internally on a repeating cycle and put six questions to each one.
- Who is the named reader, as a person rather than a role or a committee?
- What decision does it inform, and when did it last change one?
- Does any figure in it appear elsewhere, prepared from the same source?
- Who assembles it, for how long, and at what level of seniority?
- If it arrived a week late, would anything go wrong?
- Does a client, lender, insurer or professional body expect it, or did we invent it ourselves?
That last question sorts the list. Anything the firm invented for itself can be stopped by the firm itself, today, without permission from anyone outside the building.
Why is automating a dead report worse than leaving it alone?
This is the part firms get wrong. Internal reporting looks like an ideal candidate for rebuilding: repetitive, structured, and loathed by whoever does it. So it gets automated, the saved hours are celebrated, and a pointless document becomes permanent.
Three things follow. The report becomes cheap, so nobody questions it again. It becomes a dependency, wired into the practice management and finance systems, so every later change has to account for it. And it consumes the slot you had. A firm rebuilding one process at a time has a limited number of attempts in a year, and spending one on a document nobody reads is an expensive way to be efficient. That is why sequence matters, and why choosing what to automate first should start from what the firm has decided to keep.
Deletion is the only improvement that is free, instant and permanent. Do it before anything else.
What do you do with the reports that survive?
Reduce before you rebuild. There are three moves, in order.
- Merge. If four documents carry overlapping figures, one pack with four sections will do, built from one set of definitions. Agree what utilisation means once, in writing, and the arguments stop.
- Slow down. Much weekly reporting exists because weekly felt diligent, not because anyone acts weekly. Monthly figures read properly beat weekly figures skimmed. Keep the short cycle only where someone genuinely intervenes within the week.
- Cut the circulation. Send it to the people who decide. Everyone else can ask. A list of fifteen recipients usually means nobody wanted to leave anyone off.
What remains is a handful of documents with a named reader and a real purpose, and those are worth rebuilding properly, so the pack assembles itself from the systems the numbers already sit in and a named person reviews and approves it. That is the work covered by board and management reporting. The same logic applies to anything leaving the firm: client reporting earns the effort precisely because a client is waiting on it.
A useful rule for the survivors: if the firm would not pay an outside supplier to produce this document, it should not be paying fee earners to produce it either.
What about reporting you are not free to stop?
Some of it is not yours to cancel. Evidence kept for a professional body such as the SRA, ICAEW, RICS or the FCA, records a lender has asked for, and reporting written into an engagement letter all stay, as does the trail proving work was supervised and checked. There the question changes from whether to produce it to how little human time it should take, which is where compliance evidence and audit trails come in. Capture it as the work happens rather than reassembling it afterwards, and the cost falls without anything being lost.
How do you stop the list growing back?
A cull is only worth doing once if you change the rule that created the pile. Make every new recurring report carry three things from the day it is proposed: a named owner, the decision it informs, and a review date. At the review date it stops unless the owner argues for it. Sunset by default, renewal by exception.
Then do the arithmetic on what you removed. Multiply the hours each stopped report took by the people involved, by their charge-out rate, by forty-six working weeks. The unbillable hours calculator does it for you and the guide sets out the method. Put that figure to the partners before asking them to approve any rebuild. It is the cheapest return the firm will see all year, and it makes the case for fixing what is left. If you are unsure which survivor to tackle first, the audit will tell you where the hours sit.