Waiting Is Not a Status

1. Waiting Is Not a Status

Open your job list and count how many are sitting on something somebody else owes you. An engineer’s detail. A variation the client has not signed. A certifier who has had the paperwork for three weeks. A consultant whose report was due a fortnight ago and who has not been asked about it since.

Each of those jobs has a status, and the status is some version of “waiting”. That word is doing a lot of damage, because it describes a state rather than an action. Nothing about it says who is waiting, how long they have been waiting, who was asked last, when they were asked, or what happens if nobody asks again. A job can sit in “waiting” for two months and look, on every screen in the business, exactly the same on day sixty as it did on day two.

This is why the duty keeps appearing in job ads almost word for word. “Follow up outstanding information, approvals and consultant documentation.” “Maintain action registers, follow up outstanding items.” “Liaising with certifiers, trades and suppliers to obtain outstanding information and keep jobs progressing.” Businesses are hiring a person to be the thing the system does not do: remember what is open, and go and get it.

That person is usually good at it. They are also the single point of failure, and the moment they take annual leave the register stops existing, because the register was them.

2. What You Are Actually Waiting On

“Outstanding items” sounds like one thing. It is at least seven, and they behave differently enough that a single reminder rule handles none of them well.

They differ in who owes the item, what the consequence of silence is, and how hard you can reasonably push. A certifier and a client choosing tiles do not get the same message on the same schedule. Any system that sends both the same nudge will be ignored by both.

3. Why the Register Lives in Somebody’s Head

Most builders running this badly are not disorganised. They have project software, a job list, and probably a spreadsheet somebody started with real discipline. The register still ends up in somebody’s head for a structural reason: an outstanding item is created by an absence, and software records events, not absences.

When you send an RFI, something happens, and a system can log it. When the answer does not come back, nothing happens. There is no event to record, no field that changes, no row that appears. The item that matters most is the one that generates no data at all, so the only place it can live is in the memory of the person who sent the email and has been quietly counting days ever since.

The spreadsheet fails for the matching reason. It works for about five weeks, because keeping it accurate means someone typing in what did not happen, every day, for items spread across email, phone calls, site conversations and text messages. It is not a hard task. It is a task with no natural trigger, and those are the ones that stop getting done.

This is the same failure that Your Scheduler Has No Hands describes from the other direction: scheduling software keeps the record, but it does not keep the record true. Here the record does not exist to be kept.

4. This Is Not Debtor Chasing

It is worth being precise about this, because the two get sold as the same product and they are not.

Chasing an overdue invoice is a conversation about money somebody already owes. The amount is known, the terms are documented, the escalation path is well established, and the relationship can survive being firm. Everything about it is defined. That is why it automates so cleanly, and it is what Accounts Receivable Automation is about.

Chasing an outstanding document is a register of who owes you what, where the amount is not money, the due date is often informal, and the person you are chasing is frequently someone you need goodwill from next month. The certifier who signs off your next four jobs is not a debtor. Push them the way you would push a late payer and you will get the certificate slower, not faster.

5. What It Looks Like When It Runs Itself

Five moving parts. The first two are most of the value, and neither of them sends a single message.

1. 1

   The item gets opened without anyone opening it

   An email that asks a consultant for something, a variation sent for signature, a submission to a certifier: each of these is recognisable from the message itself. The register entry is created from the outbound message, because asking somebody to do the data entry is what broke the spreadsheet.

2. 2

   It closes itself when the thing arrives

   The reply with the report attached, the signed variation coming back, the certificate landing in the inbox. If closing an item is manual, the register drifts, and a register you do not trust is worse than none because you go back to checking by hand anyway.

3. 3

   Age and consequence, not just a list

   Sorted by how long it has been open and what it is blocking. Eleven days on a colour selection is a nuisance; eleven days on a certifier holding up a handover is the most expensive line in the business that week. The list has to say which is which.

4. 4

   Follow-ups that go out on their own, drafted for the recipient

   Different cadence and tone per item type, and a message that restates what was asked, when, and what it is holding up, so the recipient can act on the reminder itself rather than going to find the original email. Most people are not ignoring you; they lost the thread.

5. 5

   Escalation to a human, by exception

   After two or three automated nudges, the item stops being an automation problem. It surfaces on a short list of things that now need a phone call from someone with authority. The point of the system is to make that list short and correct, not to replace the call.

6. What Stays With a Person

Naming this is what makes the rest credible. The following do not automate, and a vendor who tells you otherwise has not done the work.

• Deciding how hard to push a certifier, a consultant or a client you need again next month. That is a relationship judgement and it is the whole skill of the role.
• The phone call that actually unblocks a stuck item. Automated follow-up gets you the two-week nudges; it does not get you the conversation on day twenty-one.
• Anything contested. The moment there is a disagreement about what was requested, what was agreed, or whose delay it is, a person owns it.
• Deciding to proceed without the item. Whether to start work on an unsigned variation is a commercial risk call with your name on it.
• Knowing which silence is normal. Some consultants are always ten days and always deliver. A register flags both the same way; a person knows the difference.

7. What It Is Worth

The admin hours are the smaller half. Somebody spending an hour a day rebuilding a mental list, re-reading email threads to work out who was asked what, and writing follow-ups from scratch is real money, and it is the part that shows up on a timesheet.

The larger half does not appear anywhere. It is the job that sat an extra three weeks because the RFI went unanswered and nobody noticed until the site called. It is the variation done at risk and argued about at final claim. It is the handover that slipped a month past practical completion while a certificate sat in somebody’s queue, unchased. None of these are recorded as costs, which is precisely why they persist.

8. Where to Start

In this order. The first two need no software at all.

1. 1

   Write the list by hand, once

   Every open item, who owes it, when you last asked, what it blocks. One afternoon. It tells you the volume, and the volume decides whether any of this is worth doing.

2. 2

   Make the requests go out from one place

   If RFIs, variations and document requests leave from four inboxes and two phones, nothing can see them. This is unglamorous and it is the prerequisite for step 4.

3. 3

   Pick the one item type that hurts most

   For most builders it is consultant documentation or certifier sign-off, because those have the longest silences and the hardest deadlines. Automate that one and leave the other six alone until it works.

4. 4

   Automate opening and closing before you automate chasing

   A register that maintains itself is useful even if every follow-up is still sent by a person. Automated nudges on top of a register nobody trusts will annoy your consultants and fix nothing.