A leasing team can use AI to draft a renewal sequence: an offer letter, two or three reminders and a short rent-change explanation, all written from placeholders and merged with real tenant data outside the AI. A person must check the lease's notice period, the approved rent figure and the dates before anything is sent.
The AI saves the writing, not the thinking. The risky parts of a renewal are all facts: the lease end date, the notice the lease or local rules require, and the rent you are entitled to ask for. A letter with confident wording and a wrong date is worse than no letter. So the routine below separates the pieces: policy and figures are decided by people, wording is drafted by AI, merge is mechanical, and a short review step catches errors.
The renewal timeline with a notice check built in
Start from the lease expiry date and count backwards. The exact numbers depend on the lease and any rules that apply to the property, so replace the example days below with your own. The point is that every step has a date, a letter and a check.
| Days before expiry | Step | Check before sending |
|---|---|---|
| 150 | Internal list of expiring leases; rent decision made | Lease end dates verified against the lease, not just the system |
| 120 | Renewal offer letter | Notice period in the lease met; rent figure approved; any limit on increases checked |
| 90 | First reminder if no reply | Reply log checked, so tenants who already answered are not chased |
| 60 | Second reminder, with the deadline to decide | Deadline consistent with the lease and the offer |
| 45 | Phone call or visit for non-responders | Notes logged |
| 30 | Final reminder; start marketing the unit if no decision | Marketing timed to the tenant's notice rights |
The notice check is the one people skip. Some leases require notice of a rent change or of non-renewal a set number of days before the end; some places add their own minimums on top. The AI does not know either, and a model's idea of a typical notice period is not a source. The check is a person reading the lease clause and writing the date next to the entry. A wider method for tracking dates is in how to track contract renewals and deadlines with AI.
Why a renewal is worth the effort: the sum
Take an illustrative portfolio of 240 apartments with 20 leases expiring next month, each at a current rent of $1,450 a month. A tenant who leaves costs you vacancy, preparation and re-letting. A reasonable illustration for one unit:
- 35 days empty at $1,450 a month: about $1,690 of lost rent (1,450 ÷ 30 × 35, treating a month as 30 days).
- Cleaning and repairs to make the unit ready: $900.
- Advertising and viewings: $300 of cost and staff time.
- Total: about $2,890 for one turnover.
Offer a 4% increase instead: $1,450 × 1.04 = $1,508, an extra $58 a month or $696 over a year. Even if the tenant would have stayed at a flat rent, the increase brings in $696 against a $2,890 risk of turnover. But if the 4% increase pushes one in five tenants to leave, the numbers change: four of the 20 leave, costing 4 × $2,890 = $11,560, while the extra income from the other 16 at 4% is 16 × $696 = $11,136. Those are the stakes the pricing decision should look at. They are one reason the rent figure is for the manager, not the AI, and why the letter should be the easy part.
The Complete Excel Course with AI Integration
From formulas and pivot tables to Power Query, DAX, VBA and AI with Copilot, ChatGPT and Claude — hands-on from the first lesson.
By Sagnik Bhattacharya and Virginia Thorn, Primetime Emmy Award winner
Buy nowThe offer letter: a prompt and what comes back
Write the prompt so the AI produces a template with merge fields, and does not see tenant data. A workable version:
Write a lease renewal offer letter for a residential tenant. Use placeholders only:
[First name], [Unit], [Current end date], [Current rent], [Option A term], [Option A rent],
[Option B term], [Option B rent], [Reply by date], [Contact name], [Contact phone].
Tone: warm, brief, plain English. Length: under 220 words.
Include: thanks for living with us; the two renewal options in a short table; what happens
if they do not reply; how to accept; a line saying we are happy to talk it through.
Do not state any legal rights, notice periods or deadlines other than the placeholders.
Do not promise future rent levels.A reasonable draft comes back along these lines (illustrative):
Dear [First name],
Thank you for making [Unit] your home. Your current lease ends on [Current end date],
and we would love you to stay.
Here are your renewal options:
- [Option A term] at [Option A rent] a month
- [Option B term] at [Option B rent] a month
To accept, reply to this letter or call [Contact name] on [Contact phone] by
[Reply by date]. If we do not hear from you, we will contact you again, and we will
need to plan for the unit after your lease ends.
If you would like to talk anything through, please call. We are glad to help.
Kind regards,
[Contact name]Two edits are typical. The sentence "we will need to plan for the unit after your lease ends" can read as a threat, and a manager might replace it with "we will check in again before then". And a reviewer will compare it against the lease for anything that must be said formally, such as how the offer is delivered. The AI draft is the first 90 per cent; the last 10 per cent is local rules and house style.
Writing the rent-change paragraph
The rent paragraph is the part tenants read twice. Keep it short and factual, without apologetic padding or sales language. Give the AI the approved figures and ask for options at different tones, then pick one. For the example $1,450 to $1,508 change:
Option 1 (plain): "For a 12-month renewal, the monthly rent will be $1,508, an increase
of $58 from your current $1,450."
Option 2 (context): "For a 12-month renewal, the monthly rent will be $1,508, an increase
of $58. This reflects increases in our running costs; we have kept the increase below
[X]% of your current rent."
Option 3 (for a long-staying tenant): "Because you have lived with us for [N] years, we have
kept your 12-month increase to $58 a month, to $1,508."Only use a reason such as "running costs" if it is true and you can explain it. Do not let the AI invent a justification, a market statistic or a comparison with neighbouring rents. And if rules limit how or by how much a rent can change where the property is, the wording must follow them; the AI cannot check that, and you should ask your own adviser if you are unsure. A letter that states a wrong legal position can create problems that outlast the renewal.
Reminders that sound like a person
The reminder letters are where AI helps most, because writing the same follow-up forty ways is boring. Ask for three short variants of each reminder, in different tones, and pick one that matches your brand. Example prompt for the 60-day reminder:
Write a short email reminder (under 90 words) to a tenant who has not replied to our
renewal offer. Placeholders: [First name], [Unit], [Reply by date], [Contact name].
Friendly, no pressure, no mention of legal rights. Give me three versions: one very
brief, one warm, one that asks a question to prompt a reply.The "question" version is worth testing first, for example, "Is there anything about the renewal options we can clear up for you?", because it invites a reply rather than a decision. Keep each reminder a different angle, not the same text with a new date, so a tenant who has seen the first does not feel nagged. Send by the channel the tenant prefers, as recorded in the file.
Merging real data without feeding it to the AI
Keep the AI away from tenant records. The process is:
- Export expiring leases from your property system to a spreadsheet: unit, tenant first name, current rent, end date, preferred channel, approved offer rent.
- Add the notice-check column, filled in by a person after reading the lease clause.
- Use mail merge in your word processor or email tool to combine the spreadsheet with the AI-written template.
- Spot-check ten merged letters, against the system, for rent, dates and names.
For cleaning the export, particularly inconsistent date formats, how to clean messy data is a good companion. If a team does want AI to see the data, for example to draft personal notes, use a business plan that does not train on your content and keep the extract to the fields needed; the broader rule is in stopping AI tools training on your business data.
Four situations that break a standard sequence
A tenant who replies with a question about repairs
The tenant answers the offer by asking when a leaking tap will be fixed. A reminder arriving the next day is tone-deaf. The reply log needs a status for "replied, open issue" that pauses reminders and flags it to the property manager. How AI helps triage such requests is covered in AI for tenant messages and repairs.
A lease ending mid-month on a Friday
An expiry that falls at a weekend can shift your own deadlines: an office closed on the date a reply is due, or a reminder landing on a public holiday. Generate the schedule from a spreadsheet formula that moves reminder dates to the previous working day, and have the reviewer glance at any date that fell near a holiday.
Two tenants on one lease
The letter addresses both, and either can accept. The merge field needs "[First name] and [First name]", and the system needs one answer recorded for both. A template for joint leases, written once, is easier than editing letters one at a time.
A tenant who wants a different term
Someone asks for a nine-month renewal instead of 6 or 12. Do not let the AI promise it. Log the request, price it using your approved rule (a shorter term often carries a higher monthly rent), and send a bespoke reply approved by the manager. A good prompt helps with wording; it does not replace the policy.
One month of expiring leases, filled in
This is how the working sheet might look for the first six of the 20 leases expiring next month. All data is illustrative, and the offer rents follow a simple written rule: 4% for a 12-month term, 6% for a 6-month term.
| Unit | Ends | Rent now | 12-month offer | 6-month offer | Notice checked | Flag |
|---|---|---|---|---|---|---|
| 2A | 30 Jun | $1,450 | $1,508 | $1,537 | Yes, clause 8 | None |
| 2C | 30 Jun | $1,390 | $1,446 | $1,473 | Yes, clause 8 | Open repair ticket |
| 3B | 7 Jul | $1,525 | $1,586 | $1,617 | Yes, clause 8 | Two occupants |
| 4D | 12 Jul | $1,610 | $1,674 | $1,707 | Not yet | Hold: lease is non-standard |
| 5A | 15 Jul | $1,450 | $1,508 | $1,537 | Yes, clause 8 | Long-stay: use option 3 wording |
| 6C | 31 Jul | $1,390 | $1,446 | $1,473 | Yes, clause 8 | None |
The checks the reviewer makes from this sheet: 2C is paused until the repair ticket is closed or scheduled; 3B uses the joint-tenant template; 4D is held because its lease is not the standard form and its notice clause has not been read; 5A uses the long-stay rent wording. The arithmetic is quick to verify: 1,450 × 1.04 = 1,508, and 1,450 × 1.06 = 1,537. Only two of the six letters are plain "send as drafted" in this illustration, which is a good reason to leave the review step in.
A mistake that happened, and how the review caught it
Here is a plausible failure in an illustrative team. A coordinator exports the lease list on Monday, merges the letters on Tuesday, and sends on Wednesday. Over the weekend, a tenant in unit 3B had already signed a renewal by email at the previous rent, and the system had not been updated. Without a review, the tenant receives an offer at a higher rent for a lease they have already renewed, and trust drops. With the review list above, the person checking removes anyone with a reply since the export. The fix is a rule rather than a tool: refresh the reply log on the morning of sending, and re-run the filter. It costs two minutes and avoids a very annoying phone call.
Letters beyond the offer
A renewal sequence is rarely only the offer and its reminders. Three more letters are worth having as templates, each short enough for AI to draft and a manager to approve.
The acceptance confirmation
When a tenant accepts, send a short confirmation that restates the term, the new rent, the start date of the new rent and what happens next: usually a lease addendum to sign. Keep it to 100 words. The AI is good at this, and the review is a quick match against the signed document.
The month-to-month option
Some tenants want flexibility and some leases convert to a rolling arrangement by default. If you offer month-to-month, state the rent clearly (usually higher than a fixed term), how much notice either side must give, and any consequences for deposits. Do not let the AI describe what the law says about rolling tenancies; write that sentence yourself or take it from your adviser, and keep the AI to the surrounding explanation.
The acknowledgement of a decision to leave
When a tenant gives notice, the best letter is short and kind: thanks for their time, the date the lease ends, how move-out works, when the deposit process starts and a contact name. Here is an illustrative version:
Dear [First name],
Thank you for letting us know. We have noted that you will be leaving [Unit] on
[End date]. We will send you our move-out checklist this week, and [Contact name] can
arrange a pre-move-out visit if you would like one. Your deposit will be handled as
set out in your lease. We are sorry to see you go, and wish you well.
Kind regards,
[Contact name]This letter is also a trigger for your own process: remove the unit from the reminder sequence, start marketing, and book the inspection. Tying those steps to the letter's sending means nothing falls between the cracks.
Handling tenants differently without treating them unfairly
A written rule keeps pricing and tone consistent while allowing sensible exceptions. A simple structure: a standard increase for everyone, a reduced increase for tenants with a defined length of stay, and a hold for any unit with an open maintenance issue until it is resolved. What it must not do is vary by anything that relates to who the tenant is as a person. Write the rule down, apply it from the spreadsheet rather than from memory, and keep the log. If a tenant later asks why their rent changed, the answer is on file, and the AI never had a say in it.
A tracker you can copy
Renewal tracker (one row per unit)
Unit | Tenant first name | Lease end | Notice clause read (Y/N, date, initials)
Rent now | Approved offer rent(s) | Offer sent (date, channel)
Reply logged (date, outcome) | Reminders sent (90 / 60 / 30) | Open issues
Decision: renewed (term, rent) / month-to-month / notice given (date) / no reply
Next action | Owner | Due dateThe tracker is the real product here. Letters are cheap to draft; the tracker tells you which letter to send to whom and when. The notice column is the one to fill in without fail, because it is the check no tool can do for you.
The review step: five minutes per batch
Before each batch leaves, one person who did not write it checks a fixed list:
- Lease end dates match the lease for the units sampled, not only the system.
- The notice requirement is met for each unit, with the date recorded.
- Rents in the letter equal the approved figures; a sum is done for ten letters.
- No letter contains a promise about future rent, a legal statement or a justification not on the approved list.
- Names, units and channels merged correctly; no placeholder text left in square brackets.
- Tenants who have already replied, given notice or have an open issue are removed.
A neat check for leftover placeholders is to search the merged file for an opening square bracket. It takes ten seconds and catches the commonest embarrassing error. If your team already has a consistency rule for how tenants are treated, the points in how property managers use AI to screen tenant applications fairly carry over: apply the same criteria to everyone and be able to explain them.
One last habit helps the first cycle: send the first batch to yourself and a colleague as if you were tenants. Read the email on a phone, because plenty of tenants will open it there. Check that the table of options is legible on a small screen, that the reply instructions are the first thing visible below the greeting, and that the sender name is a person rather than a no-reply address. A buried reply instruction or a no-reply sender can lose a reply as surely as a poor tone.
Measuring whether the sequence works
After two renewal cycles, count four things: the proportion of tenants who renewed, the average days from first offer to a reply, the number of letters that needed correction after sending, and the time your team spent per batch. If the renewal rate improves and corrections stay near zero, the sequence is working. If corrections rise, the review step is being rushed. If the response time does not change, the reminder wording or the channel is the problem, and you can test two variants on half the tenants each. Keep the winning variant, retire the other, and write the rule into the template so the next team member inherits it.
For a view of where renewals sit in a property manager's week, see how much of a property manager's week AI can take over. If you would like to build the timeline and the review step with someone, that is the kind of job my AI implementation consultation covers, and the monthly routine is small enough for your own team to run afterwards.
Questions leasing teams ask about AI renewal letters
Can AI decide how much to raise the rent?
It can lay out options and the arithmetic, but the figure is a management decision, based on your market evidence and any rules that limit increases. Give the AI the approved figure and let it write the wording. Never ask it to choose the number.
How early should a renewal offer go out?
Early enough to meet any notice requirement in the lease or local rules, plus time for the tenant to decide and for you to market the unit if they leave. The illustrative timeline here sends the offer 120 days before expiry, but the lease terms and the rules that apply to your property decide the minimum.
Should the same letter go to every tenant?
The structure and policy should be identical, so people in similar positions are treated alike. The details change: name, unit, dates, figures and any history worth acknowledging. Keep a written rule for how increases are set so the pattern is explainable.
Is it safe to paste tenant details into an AI tool?
Avoid it. Draft the letter with placeholders in the AI tool, then merge names, units and figures in your own system or a spreadsheet. If you must use real data, use a business plan with no training on your content and put only what the letter needs.
What if the tenant says they want to leave?
Stop the renewal sequence for that unit, log the notice, and move to your move-out process. Do not keep sending offer reminders. A short, courteous acknowledgement letter and a check of the notice date against the lease protect both sides.
Further reads
- AI for Letting Agents: Handle Tenant Queries Without Extra Staff — Handle tenant questions that follow a renewal letter.
- How Estate Agents Use AI to Prepare Valuation Appointments — Preparing the market evidence behind a rent decision.
- Compliance Calendar: Track Licences, Insurance and Certificates — A calendar method for dates that must not be missed.
- Can ChatGPT Draft a House Rental Lease? What Landlords Must Check — What a landlord must check when AI drafts lease text.
- How Property Managers Use AI to Triage Maintenance Requests — Repairs handled well make renewals easier to win.
- How to Automate Insurance Renewals With AI and Keep More Clients — A renewal sequence in another trade, worth comparing.
- Sales Progression for Estate Agents: Chain Chasers With AI — Chaser sequences in a neighbouring property job.
- Storage Unit Rental Agreement Template: Plain-English Draft Using AI — How a self-storage operator can draft plain-English rental terms with AI, with filled-in clauses and the list to hand a solicitor before customers sign.
- Alarm Maintenance Contracts: Renewal Letters and Reminders With AI — Renewal letters built from locked, lawyer-approved clauses plus AI-filled lines, with a reminder timeline, price-change wording and opt-outs.
- AI Leasing Agent for Small Apartment Owners: Chat, Voice or Both? — Chat, voice or both for leasing enquiries: when each pays off, where staff take over, and the fair housing guardrails to set first.
- AI for Apartment Leasing: A Day Plan for Tours, Follow-Ups, Renewals — An hour-by-hour leasing routine with tour preparation, message templates, renewal checks and human decision points.
- Can ChatGPT Write a Lease Agreement? What to Check Before You Use It — A landlord's clause-by-clause review of AI lease drafts, official forms and the points that need a solicitor.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: general lease-administration practice and the example's own arithmetic. Notice periods and limits on rent changes depend on the lease and the rules that apply to the property; check them for each unit.