Notes tagged
records
31 notes on records. All notes.
Reported progress and real progress
The percentage in a weekly report is a judgement made by the person who will be blamed if it is low. Quantities and binary milestones survive that pressure.
A snag list that actually closes
Most snag lists never close because a snag has no identity, no owner and no closure test. What a list needs if it is to shrink instead of being reissued.
What belongs in a handover file
The handover file is assembled at the end, which is why it is always incomplete. What belongs in it, why each item is needed later, and the habit that fixes it.
The paper trail a query demands
Years later somebody asks you to substantiate a decision. What survives is a dated record, a named author, an unedited document and proof it was sent.
One builder, many companies, one set of records
A developer runs several legal entities for sound reasons. The cost is operational, and it lands on staff time, shared material and money between accounts.
Leads that die in an Excel sheet
Every project starts with a sales sheet, and it fails for mechanical reasons — no history, no two people at once, and a date in a cell that never rings.
What a lead record should actually hold
Field by field, what belongs on an enquiry record in a primary sales office, what does not, and the one test that settles every argument about a new field.
From booking to agreement: the document ladder
The paper between a verbal yes and a registered agreement, rung by rung, and the places a booking file quietly goes missing while everyone assumes it is fine.
After the booking comes the longer job
The years between booking and possession are where a builder's reputation is actually made, and most sales offices are neither staffed nor organised for them.
Getting your data back out
Every system ends. A report is not an export and an export is not a backup, and the difference decides what you are holding on the day you leave.
Ask to see the audit trail before you sign
Every vendor says the history is kept. The way to find out is to make a record, change it, cancel it, and ask to see that one record's history on screen.
The challan that protects the supplier
A delivery challan is usually described from the receiving side. Written from the supplier's side it is evidence, and six months later it settles the argument.
What a work order should say
Read it before you sign it. The clauses that decide whether a job is profitable are rarely the ones anyone reads, and three lines are usually missing.
Signing a document from a phone
Three different things are called a signature on a phone, and they prove very different amounts. What to check before your thumb touches the screen.
Keep your own copy
At final account the argument is settled by whoever can produce a dated document the other side cannot contradict. Usually that is the larger party.
What a delivery challan is, and what it proves
A delivery challan travels with the goods and says what left the seller's gate. It is not a bill, and on its own it is not proof that anything arrived.
The concrete cube test, and the four weeks you wait
Cubes are cast on site, cured, crushed and recorded. Most of what goes wrong with the test happens in the first ten minutes, long before the crushing.
Set-off, and what it costs you to use it
Netting what you owe against what you are owed is quick and tempting. It also collapses two records into one and removes the evidence for both.
Paying less than agreed is a statement, not a typo
A short payment carries a decision inside it. If the system accepts the smaller number silently, the decision is lost and only the discrepancy survives.
Work order or purchase order: which one you are issuing
One buys a thing, the other buys an outcome. The paperwork looks similar and the risk, measurement and payment mechanics are completely different.
Why one person should never move the company's money
Two different people on every payment is not distrust. It is the cheapest control there is, and every convenience that erodes it erodes the whole thing.
Writing off money that is never coming
A write-off is a decision, not a correction. The amount matters less than the reason, and the reason is the first thing most systems throw away.
A gate register somebody will actually keep
The gate is the only place every lorry passes. Most gate registers fail for design reasons, not discipline ones — they were built for a desk, not a barrier.
The daily progress report nobody reads
Most DPRs are written by somebody who gains nothing from writing them, for somebody who does not read them. There is a version that survives that fact.
A site photograph is a record, if anything reads it
Thousands of site photographs are taken every month and almost none become records. The bottleneck is not capture. It is that nothing ever looks at them.
Who is allowed to decide what
Most permission systems answer "what may this person see". The harder question is what they may decide, and organisations usually cannot state their own answer.
What an audit trail is actually for
Not catching thieves. An audit trail exists so that a decision made a year ago can be reconstructed, including by the person who made it and now has to defend it.
Append-only: why a correction should never overwrite
A record that can be edited answers only today's question. Adding a correcting entry instead costs one row and keeps the history that makes the number defensible.
Tamper-evident, not tamper-proof
You cannot stop somebody altering a copy of a document. You can make any alteration detectable, and that is a much more useful property than it sounds.
One outage should be one ticket
Systems that raise an alert per symptom flood at exactly the moment attention is scarcest, and the flood is usually one cause wearing many names.
An import you cannot reverse is an import you should not run
Bringing history into a live system is a one-way door unless you build the way back first. Two properties make it safe, and both are cheap before the write.