A Spreadsheet Agrees With Whatever Is Typed Into It

Spreadsheets do not fail. That is the problem with them.

A property portfolio run on Excel does not produce an error when a renewal is missed, a cheque sits past its date, or an owner's balance is calculated from a formula somebody dragged one row too far. It produces a number. The number is presented, accepted and acted on, and the only signal that anything was wrong is that two people eventually quote different figures for the same building and both can show their working.

That is worth stating precisely, because it explains why the replacement decision gets deferred for years past the point it should have been taken. Businesses replace things that break. A spreadsheet never breaks. It agrees.

Why it survives so long, honestly

Give the spreadsheet its due, because pretending it is useless is the fastest way to lose an argument with the person who built it.

For a first portfolio — a handful of units, one owner, one person handling everything — a spreadsheet is genuinely the right tool. It costs nothing. It can be reshaped in an afternoon when the business changes. It requires no implementation, no training and no vendor. The person who built it can answer any question about it instantly, because the whole model is in their head.

Every one of those advantages is real, and every one of them is a property of one person operating a small set of facts. The thresholds below are the points at which that description stops being true. None of them is a unit count.

The four thresholds

One: you are holding money that is not yours

The day you manage property for somebody else, the business changes category. It is no longer bookkeeping, it is a trust relationship: rent collected belongs to an owner, less your fee, less what you spent with his authority. That balance is a liability you owe, and the person it is owed to is entitled to ask about it whenever he likes.

A spreadsheet can compute an owner statement. What it cannot do is prevent the four things that produce the arguments — paying from receipts that have not cleared, spending above the mandate, letting a unit whose management agreement has expired, and taking the fee on billed rather than collected rent. Those are refusals, and a spreadsheet cannot refuse. It is the most permissive tool in the building.

That whole subject has its own treatment in the money in your account is not your money, and it is the threshold that matters most, because it is the only one where the failure ends in a dispute with the person paying you.

Two: rent stops being a single payment

One tenancy, one annual payment, is a row. A tenancy with quarterly instalments, four post-dated cheques with their own dates and custody, a bounce, a replacement and a partial transfer is not a row. It is a small ledger.

The moment collection is a calendar rather than an event, the spreadsheet's core weakness becomes structural: it has no concept of a date arriving. Nothing in it will tell you on Thursday that a cheque was due on Sunday. Somebody has to look, and looking is a habit rather than a mechanism, and habits fail in exactly the weeks when there is too much on. The mechanics of that are in the cheque in the drawer is not an asset until somebody walks it to the bank.

Three: more than one person needs to write to it

This is the threshold most people cross without noticing.

A spreadsheet with one author is a model. A spreadsheet with three authors is a negotiation. There is no record of who changed a figure, no way to prevent two people editing different copies, and no mechanism by which the leasing team's version and the accounts team's version are ever forced to agree. They diverge immediately and silently, and the divergence is discovered in a meeting.

The version circulated by email is the worst case and the most common one. At that point the business does not have a portfolio record. It has several, and the authoritative one is whichever was attached to the most recent message.

Four: somebody outside the company can ask you to prove a number

An owner asking how his balance was reached. A tenant's counsel asking why the instalments do not add up to the contract value. An auditor asking which supply a credit note reversed. A tribunal asking when a demand was served.

Every one of those is a request to reconstruct a figure, and a spreadsheet cannot be reconstructed. It shows its current state. It does not show what it said in March, who changed it, or on what basis. The formula is visible; the history is not.

The question is never whether the number is right. It is whether you can show how it was reached, to somebody who is not obliged to believe you.

The unit count question, answered

People want a threshold in units, and the honest answer is that unit count is a weak predictor.

Fifty units on a single-owner building, one instalment a year, one person handling everything, is comfortably a spreadsheet business. Twelve units managed for six different owners, with mixed residential and commercial tax treatment, quarterly cheques and a maintenance obligation, is not — and it will not become manageable by trying harder.

The better measure is manual steps per renewal cycle. Count them for one tenancy, end to end: the reminder that a renewal is due, the index check, the offer, the new contract, the registration, the schedule, the cheque collection, the invoice run, the accounting entries, the owner statement. However many of those require somebody to remember, retype or reconcile, multiply by your renewals per year. That figure is the real size of the operation, and it is the one that decides whether the business is scalable, which is why scale in this trade is less about record count than about manual steps.

What a spreadsheet cannot do at any size

Independent of volume, four capabilities are simply absent, and no amount of skill in building the file supplies them.

It cannot refuse. Every gate that makes property numbers trustworthy is a refusal: a schedule that does not balance, a tenancy the cheques do not cover, a hand-typed invoice against a live contract, a payout larger than cleared funds. A spreadsheet accepts all four without comment.

It cannot notice a date passing. Renewals, cheque presentations, mandate expiries and notice periods are all events that happen when nobody is looking at the file.

It cannot hold one fact for two audiences. The leasing view of a unit and the accounting view of the same unit are different shapes of the same record. In a spreadsheet they are two files.

It has no trail. Which is only an inconvenience until the first time it is not.

The honest case for not buying yet

Some businesses reading this should stay where they are for now, and it is worth saying so.

If you own what you manage, so there is no owner balance and no mandate. If rent arrives once a year per tenancy. If one person does the whole cycle and that person is not planning to leave. If nothing is jointly owned and no external party audits your figures. Then the spreadsheet is doing a job it is well suited to, and a property system would be an implementation cost against a problem you do not currently have.

Buying early is not harmless. A property system that goes in before the business needs it gets configured around assumptions nobody has tested, half-adopted, and then quietly worked around — and a stalled property system is worse than the spreadsheets it replaced, because now nobody knows which of the two is right.

The trigger to act is not ambition. It is the first of the four thresholds you have actually crossed.

What it is costing while you decide

We would rather not quote an industry statistic here, because an average is a number nobody can check and this site does not run on them.

The mechanism is more useful anyway, and it is arithmetic you can do with figures you already hold. A renewal handled late is either a vacancy or a weaker negotiating position. A cheque presented eleven days after its date is eleven days of credit you did not decide to extend. An unresolved maintenance request is the most common reason a tenant declines to renew, which converts a repair cost into a vacancy cost an order of magnitude larger. None of those produces an entry in any report you currently receive.

If you want your own number rather than somebody else's, the cost of chaos calculator multiplies your figures by your figures and prints the multiplication next to the answer, which is a more defensible thing to put in front of a board than a benchmark from a vendor's brochure.

The short version

The spreadsheet is not failing. It is agreeing, which looks the same from the outside and is worse.

Four thresholds change the answer: holding somebody else's money, collection becoming a calendar, more than one author, and an outsider entitled to ask how a figure was reached. Cross one and the tool is no longer suited to the work, whatever the unit count says.

When that point arrives, what replaces it has to hold the unit, the contract, the instrument and the mandate as records that relate to each other — which is the argument in a tenancy is a schedule of things that have not happened yet, and the reason Odoo's Rental application is the wrong base for a tenancy even though it looks like the right one.

Ours is Proptec, live in six companies. Darna, a developer in Egypt, moved onto it in 2019 from exactly this position: units, contracts and collections on Excel, which agrees with whatever is typed into it. The rest are in the case studies.