Landlord and repair texts: building the record

What was wrong, when they were told, what they promised.

Last updated


Not romance

The record is the point. If a repair dispute escalates — to an inspector, a tenant union, a deposit adjudication or a court — every route depends on being able to show three things: what was wrong, when the landlord was told, and what they said they would do about it. Not a lie detector. Inconsistency is not proof of deception, and you do not need it to be: a landlord can be entirely sincere, genuinely waiting on a contractor who is not calling them back, and you can still be nine weeks without heating. Non-performance is the claim that matters and the claim a record can carry. Motive is neither provable from a text thread nor necessary.

This is a procedure rather than a detection method. It is written in order, and each step exists because it closes a specific gap that turns up when people bring a phone full of messages to somebody who can help and discover that the messages do not answer the questions being asked. Most of it takes seconds at the time and is impossible to reconstruct afterwards, which is the entire reason to do it now rather than when the dispute is already live.

What is the record for, and what has to be in it?

The record exists to answer three questions without you narrating it: what was wrong, when the landlord was told, and what they committed to — and anything in your thread that does not serve one of those three is optional.

Those are the questions every escalation route asks first. An environmental health or code enforcement officer wants the date of first report and the condition since. A deposit adjudicator wants what was raised and how it was answered. A small-claims filing wants dated notice and a failure to act after it. A tenant union caseworker wants to know whether there is a paper trail before advising on anything else. None of them opens with how unreasonable the landlord has been, and a thread organised around that is a thread that will need rebuilding under pressure.

Judge your existing thread against a simple test: hand it to a friend who knows nothing about your tenancy and see whether they can tell you the date each problem was first reported. If they cannot, the gap is in the record rather than in your case, and it is almost always fixable now with one or two dated messages. That is worth doing before anything else on this page, because every later step assumes a first report with a date attached.

What does not belong in it

Leave out your assessment of the landlord’s character, your speculation about why they are not acting, and the arguments about other subjects that live in the same thread. None of it strengthens a repair complaint and all of it gives the other side something to characterise instead of answering. This is not about being polite — it is that a record of six factual dated reports and four unanswered follow-ups is genuinely harder to respond to than the same thing wrapped in commentary, and the person eventually reading it has very little time.

What should the first repair message actually say?

The first message should contain one issue, the date you first noticed it, a description of the effect rather than your guess at the cause, a photograph, and a request for a date — in that order and in about four lines.

Describe the effect because you can be certain of it and it stays true if you are wrong about the mechanism. “No hot water from any tap since Tuesday 3rd” is a fact you can stand behind indefinitely. “The boiler’s pump has failed” is a diagnosis that may turn out to be incorrect, and an incorrect diagnosis in your first report is the easiest thing in the world for a landlord to answer months later by pointing out that the pump was fine. Say what you observe, attach the photograph, and let the tradesperson diagnose it.

Then ask for a date rather than for the repair, because a request for a repair can be answered with acknowledgement and a request for a date can only be answered with a date or a refusal. “Please let me know by Friday when someone can come” creates a checkable commitment on a specific day. It also makes your follow-up unarguable: on Saturday you are not nagging about a repair, you are asking about a date they were asked for and did not give, which is a much smaller and much more answerable question.

One issue per message, always

Send a separate message for each problem even when you notice three in one week, and resist the urge to write the comprehensive list. A message covering the damp, the front door lock and the extractor fan gets a reply about the extractor fan, and the reply now technically answers your message, which leaves the damp with no dated response attached to it. Separate messages produce separate threads of accountability, let you chase one item without reopening everything, and mean that when you count outstanding issues later the count is already there in the record rather than in your head.

Does a text count as written notice, or do you need something more formal?

A text is always evidence and is not always notice: it reliably proves the landlord was told and when, but whether it satisfies a legal notice requirement depends on your jurisdiction and often on a clause in your own tenancy agreement specifying how notice must be served.

Check the agreement itself first, because this is the one document that is specific to you and it takes two minutes. Look for a clause about notices, service of notices, or how to contact the landlord or agent — many specify an address, an email address, or post, and some explicitly exclude text messages. That clause does not usually stop a text being evidence of what was said, but it may well determine whether a legal clock started running on the day you sent it, and that clock is what most remedies hang from.

The practical resolution is to stop treating it as a choice. Send the request by text because that is what actually gets read, then send the identical wording by email or in a letter through the channel the agreement names, and add one line to the text saying you have also emailed it. That single line is worth more than it looks: it dates the formal copy inside the informal channel, so that even if the email is later disputed, your text thread records that it was sent. For anything affecting health, safety or heating, use recorded or certified post as well and keep the receipt, because at that point an independent delivery record is worth the cost of a stamp.

How do you turn a phone call or a doorstep promise into part of the record?

You send a short confirming text immediately afterwards that restates what was agreed in neutral terms, because an uncorrected confirmation is strong evidence and an unconfirmed memory of a call is close to worthless.

Keep it to one or two sentences and make it easy to correct: “Thanks for the call just now — to confirm, you said the plumber will come Thursday 12th between 9 and 12, and you’ll let me know if that changes.” Send it within minutes while the detail is exact. If any part is wrong, a landlord will say so, and their correction is itself a dated written record of the real commitment, which means you cannot lose. If it is right, the absence of a correction is consistent with agreement, and you have converted a conversation nobody can evidence into a line in your thread.

Do not record calls covertly as a substitute for this. Whether covert recording is lawful varies by jurisdiction and in several places recording a call without consent is a criminal offence, and a recording obtained improperly can damage an otherwise strong case rather than helping it. The confirming text achieves what people want from a recording — a fixed account of what was said — with none of that exposure, and it does it in a form that is easy to read, easy to forward, and easy for the other side to correct if you have genuinely misheard them.

How do you log the follow-ups without becoming the difficult tenant?

Chase on the date that was promised, not on a schedule of your own, and when no date was given, send the same neutral sentence once a week — because a repeated identical request reads as a record and an escalating one reads as a conflict.

Chasing on the promised date is the strongest position available and it is not confrontational: you are asking about something they said, on the day they said it. “Just checking on Thursday’s visit — nobody arrived and I haven’t heard anything. When can it be rebooked?” That message does three things at once. It records the failure on the day it happened, which is far more credible than reconstructing it later; it asks a question with only one reasonable answer; and it starts the next cycle with a fresh dated request rather than a grievance about the last one.

Where no date was ever given, keep the wording constant and the tone flat. Reusing the same sentence weekly feels strange to write and is the right call, because a sequence of five identical calm requests is much harder to answer than five increasingly frustrated ones — there is nothing in it to react to except the request. It also produces something quite specific for the record: a countable number of requests over a countable number of weeks with a countable number of replies. That sentence, when you eventually need it, writes itself from the thread.

What to log when someone does turn up

Log the visits as carefully as the promises, because a landlord’s account of a repair dispute is usually a list of visits. Note the date, roughly who came, what they said, what they did, and whether the problem persisted — in a text to the landlord rather than a private note, since a message saying “the plumber came this morning, replaced a valve, and there’s still no hot water this evening” is dated, shared and unarguable while a note in your phone is none of those. A visit that did not fix the problem is not a neutral event in the record; it is the point where a reported fault became a known and continuing one.

What does the record look like once you lay it out as a timeline?

Laid out as a timeline, the record becomes two columns of dates — what you reported and what they committed to — and the gaps between them are the case, because a repair dispute is fundamentally an argument about elapsed time.

Build it with one row per event and four fields: date, who, what was said, and what happened next. Include your reports, their replies, promised dates, visits, and the silences, and mark every promised date with what arrived on it. The most useful column is usually the last one, because it converts a scattered thread into a short set of statements — first reported on the 3rd, first reply on the 11th, first date promised for the 19th, nobody attended, chased on the 20th, next reply on the 30th. That reads in ten seconds and it does not require anyone to believe you about anything.

Two derived numbers are worth writing at the bottom, because they are what an inspector or an adviser will reach for. The elapsed days from first report to the present, and the count of promised dates that passed without attendance. Both are facts rather than characterisations, both survive being checked against the thread, and both are stronger than any adjective you could use instead. If you would rather not build the table by hand, the timeline tool below takes a pasted thread, orders it, and flags the gaps — it runs in your browser and nothing you paste is sent anywhere. The same two-column method applied to any stalled job is set out in when someone keeps saying tomorrow.

What do you do when the answer is “someone will come next week” for the fourth time?

At the fourth moving date, stop asking for a date and set one: send a single message that states the record, gives a specific deadline, and names the step you will take if the deadline passes — without heat, without adjectives, and without a threat you are not prepared to carry out.

The message has a fixed shape. What the problem is and when it was first reported. What has been promised and on which dates. What has happened. What you are asking for, with a date. And one sentence on what happens next: that you will contact the council’s environmental health team or the local code enforcement office, that you will raise it with the letting agency’s complaints process, that you will seek advice from a tenant service. Say it plainly, as information rather than as leverage. This message is simultaneously a genuine final opportunity for a landlord who has simply been disorganised, and the document you hand over when they were not.

Send it through both channels — the text and the formal one — and keep the copies. The reason this works when four polite chases did not is that it changes what the next silence means. Silence after a vague chase is ambiguous and easy to explain. Silence after a dated, specific, twice-delivered deadline is a discrete event with a date on it, and it is the event that most external routes need in order to open a file. Then follow through on whatever you named, because a deadline you let pass quietly teaches the other side exactly what the next one is worth.

What should you never do while building the record?

Never edit or crop a screenshot, never delete your own less-patient messages, never record a call covertly without knowing your local law, and never stop paying rent without advice — each of these turns a strong record into a contested one.

Edited evidence is the fastest way to lose a case you would otherwise have won. A cropped screenshot invites an argument about what was removed, and once one item in a record has been altered the whole record is treated differently by anybody assessing it. Keep the messages where you were short-tempered; a real dispute contains them, the other side has the same thread, and a record that includes your worst moments is more credible than one that has been curated. Export the whole conversation rather than assembling screenshots, so the ordering and the timestamps are carried by the file rather than by your captions.

Withholding rent deserves its own warning because it is where strong complaints go to die. Some jurisdictions provide for rent to be paid into escrow, or for repair-and-deduct, and every one of those routes has strict preconditions about notice, amounts, and documentation that must be satisfied before a penny is withheld. Others provide no such right, and unpaid rent is simply arrears that can support eviction regardless of the state of the boiler. Take the record you have built to a tenant advice service, a tenant union or a housing solicitor and ask before you change what you pay. The record is what makes that conversation short and useful, which is the whole argument for building it in the first place.

Where does the record actually get used?

The record gets used the moment you involve anyone outside the tenancy — a housing or environmental health inspector, a code enforcement office, a letting agency’s formal complaints process, a tenant union, a deposit adjudicator, or a court — and in every one of those the first question is the date of first report.

Which of those routes exists, what it is called, and what it can compel varies enormously between countries, states and cities, and any page that tells you otherwise is describing one jurisdiction and hoping. What does not vary is what they need from you. Every one of them works from dated reports, dated responses, and evidence of the condition over time, and every one of them is materially weaker when handed a recollection instead. That is why the procedure above is worth running even while you still expect the landlord to sort it out — the record costs almost nothing to build in real time and cannot be built retrospectively.

Find the local names before you need them rather than during a crisis: search for your city or council plus “environmental health”, “code enforcement”, or “housing standards”, and find your nearest tenant advice service or tenant union. Then look at what you already have on your phone and ask the question that decides whether any of it will work — for each problem in your home, what is the date it was first reported in writing, and can you point at the message? If you want the thread laid out as a dated sequence without doing it by hand, paste it into the tool below, or run the conversation through the analyser.

Check the neighbouring dispute the same way

Free tool · runs in your browser

Not a lie detector. Inconsistency is not proof of deception. This counts what is in your thread — timestamps, gaps, wording. It cannot tell you whether someone is lying.

Nothing you paste leaves this browser. No account, no upload, no model call.

Flag gaps over

This counted the gaps. It cannot read the messages. Run the actual conversation →

Common questions

Does a text message count as written notice to a landlord?
Sometimes, and it depends on your jurisdiction and on what your tenancy agreement says about how notice must be given — which is why a text should never be the only copy of an important repair request. A text is always good evidence that the landlord was told and when, because it is timestamped and neither side can alter it. It is not always sufficient notice to start a legal clock or to support a formal remedy. The practical answer is to send the same request twice, once by text so it is read and once by email or letter so it is formal, and to say in the text that you have also emailed it.
How should you word a repair request text to a landlord?
Word it as one issue, described by its effect rather than your diagnosis, with the date you first noticed it, a photograph, and a specific ask. For example: “Since Tuesday 3rd the boiler has produced no hot water — photo of the display attached. Please can you arrange a repair, and let me know by Friday when someone can come.” Naming the effect keeps the request accurate even if you are wrong about the cause, and asking for a date rather than for the repair gives you a checkable commitment rather than an acknowledgement.
How do you prove your landlord received a repair request?
You prove receipt by using a channel that records delivery and by making the landlord’s own reply do the work: any answer they send about the issue, even an unhelpful one, is an admission they received it. Where a text goes unanswered, send the same request by email and keep the sent copy, and for anything serious use recorded or certified post so there is an independent delivery record. Save the whole thread rather than individual screenshots, since a continuous export carries the timestamps and the ordering that separate screenshots lose.
What do you do when your landlord promises a repair over the phone?
Send a short text immediately afterwards restating what was agreed, because an unanswered confirming message is far stronger evidence than your memory of a call. Something like: “Thanks for the call just now — to confirm, you said the plumber will come on Thursday 12th between 9 and 12.” If that is wrong, a landlord will correct it, and the correction is also useful. If it is right, silence is consistent with agreement, and you now have a dated written record of a promise that was only ever spoken.
How often should you follow up on a repair without being a nuisance?
Follow up on the date that was promised rather than on a schedule of your own, and when no date was given, follow up once a week with the same neutral wording each time. Chasing on the promised date is unarguable — you are asking about a commitment they made — while chasing daily produces defensiveness and gives the other side something to characterise. Repeating identical, unemotional wording also builds a record that reads well to a third party, because a sequence of five calm identical requests is much harder to dismiss than an escalating argument.
Can a landlord’s delays prove they are acting in bad faith?
No. A record of delays proves non-performance, not intent, and it is worth being clear that this is not a lie detector and inconsistency is not proof of deception — landlords wait on contractors, insurers and parts exactly as tenants do, and a delay can be genuine and still leave you without heating. This is useful rather than limiting, because non-performance is what code enforcement, a housing inspector, a tenant union or a court is actually equipped to assess. Build the record to show what was reported, when, and what happened after, and leave the question of motive out of it entirely.
Should you withhold rent if repairs are not done?
Not without taking advice first, because withholding rent is the single most common way a tenant with a strong repair complaint ends up facing eviction proceedings instead. Some jurisdictions allow rent to be paid into escrow, or allow repair-and-deduct, and each of those has strict conditions about notice, amounts and documentation that must be met before you stop paying anything. Others provide no such right at all, and non-payment is simply arrears. Take the record you have built to a tenant advice service, a tenant union or a housing lawyer before changing what you pay.

A suspicion score, not a lie detector. It measures what does not line up — nothing more. Everything else is your call, which is the only place it can honestly sit.

Run your conversation through it · See pricing