October 21, 2026 · 7 min read
A 90-Day Content Calendar in One Sitting: Interview, Theme Map, Month One
Interview first, map second, write third. That order is why the 90-Day Content Calendar Prompt produces a quarter of specific content instead of a list of ideas. The three artifacts and how each is checked.
A 90-day real estate content calendar is built in three passes: a brief from an interview, a twelve-week theme map with one theme per week, and month one expanded into twenty dated pieces with a hook and talking points each. Built in that order, in one chat, it takes about an hour and produces content only that agent could post.
Most calendars fail before the first post because they start at the wrong end. An agent opens a blank spreadsheet, or a blank AI chat, and asks for ninety days of post ideas. What comes back is ninety generic ideas, because nothing specific went in. The spreadsheet gets abandoned by week three, and the broker hears "I tried a content calendar and it did not work."
The free 90-Day Content Calendar Prompt runs the passes in the right order and refuses to skip ahead. The order is the method. Here is what each pass does and why it is checked before the next one starts.
Pass one: the interview that writes the brief
The prompt asks seven questions, one at a time, and waits for each answer:
The towns or neighborhoods the agent actually sells in.
The price band where most deals land.
Three client situations the agent works with.
Three words for how the agent sounds.
Proof a client could check: years in the business, homes closed, neighborhoods sold in.
The listings or neighborhoods the agent wants to be known for this quarter.
The date the ninety days start.
One question at a time is what forces specifics. Asked all seven at once, an agent writes "Metro Atlanta, all price points, buyers and sellers, friendly and professional." Asked one at a time with a follow-up on any vague answer, the same agent writes "Decatur and Avondale Estates, $450,000 to $750,000, sellers trading a first home for a larger one." If the agent is stuck, the prompt asks about the last three closed deals instead. If there is no proof number, the line stays blank. Nothing gets invented.
Then a check before the brief is final. If a stranger read only this, what would they think the agent sells and where? Which line is the most generic? Is there any word that describes who a person is instead of what they are doing? The brief is six lines in a fixed format: market, price band, situations, voice, proof, known for. The agent copies it into a note labeled BRIEF and everything after is built from it.
Pass two: the twelve-week theme map
One theme per week, twelve weeks, four pillars: Market, Listings and neighborhoods, Process and answers, You. The rules keep it honest. No more than one "You" week in every five, because a feed that is mostly about the agent is a feed nobody follows. Every theme has to be specific to the brief's market and situations. Week one starts on the date the agent gave.
The map is shown as a table, and then the agent grades every row: K for keep, C for cut, R for rewrite. Keep means the theme is specific, sits in the right pillar, and the agent could talk about it for two minutes without notes. Cut means the agent would never say it, it names a group of people, it leans on a number the agent cannot prove, or it repeats another week. Rewrite means the idea is right and the wording is generic. Eight keeps is the target. Every cut gets replaced and every rewrite gets fixed before the final table goes into a note labeled THEME MAP.
An agent left alone keeps everything. An agent told to find two cuts finds them, and the map gets sharper.
Pass three: month one, written out
Weeks one to four expand into five pieces each: a Monday post, a Wednesday email, a Thursday post, a Friday story, and one 60-second video topic. Each piece gets a day, a format, a one-line hook, and two sentences of talking points. Twenty rows, five columns. Same rules as the brief: situations not people, no statistics outside the brief, plain language.
Then every hook runs through a five-point quality pass, and the failures get flagged:
Generic: a competitor could post it with their name swapped in.
Steering: it says who lives somewhere or who a home is "perfect for."
Unproven: it uses a number or claim not in the brief.
Not me: it uses words the agent would never say.
Duplicate: it says the same thing as another hook.
A flagged hook gets fixed the same way every time: add one place name or one situation. "Three things to know before you list" is generic. "Three things Decatur sellers ask me before they list a house they have been in since 2005" is not, and it took one place and one situation to get there. The final twenty rows go into a note labeled MONTH ONE.
What this looks like in practice
An agent runs the prompt on a Sunday afternoon. The interview takes twenty minutes because the follow-ups push back on two vague answers. The theme map comes back with twelve rows; she keeps seven, cuts two that named groups of people, and rewrites three. Month one comes back as twenty pieces, and the quality pass flags four hooks as generic. She adds a street name to two and a situation to two. By dinner she has a quarter she can execute: a brief, a map, and four weeks written, all in a note on her phone.
Pass four and the four reminders
Months two and three run the same expansion and the same quality pass, one month at a time, when the agent is ready. When all three are done, the prompt ends with four reminders that keep the calendar alive: save the three notes; book one repeating 30-minute weekly appointment to check what is scheduled and swap in anything timely; track three numbers on write day (pieces published, replies and messages, and conversations where someone mentions the content); and next quarter, update the brief, change the start date, and run the same steps.
The weekly 30 minutes is the one most agents skip and the one that decides the outcome. A calendar is a plan; the weekly check is where it meets a new listing, a rate change, or a client question worth answering in public.
By the numbers
The formats in month one are chosen on purpose. Metricool's 2026 Instagram study found Reels generate about four times the interactions of single-image posts, and carousels generate nine times the saves, while single-image posts lost 21.96% of their reach year over year. That is why the calendar carries a video topic every week and why the Monday and Thursday posts are written as carousels and Reels, not photos with captions.
The email is there for a different reason. Real estate email open rates typically land between 25% and 40%, with top performers at 40 to 50%, and a reply rate of 1 to 3% is a high-trust signal. NAR's 2025 Profile of Home Buyers and Sellers found 66% of sellers chose their agent through a referral or a past relationship. The Wednesday email is how a past client remembers the agent exists the week their neighbor mentions selling.
How Urban Marketing Edge handles this
This is the shape of every month Urban Marketing Edge runs for a client, at office scale. On Broker Growth, the monthly theme is set first, the calendar is built from it with the office's own listings, agents and neighborhoods in the slots, every piece is written and designed, approved once in the dashboard, and scheduled. The weekly meeting is where timely swaps happen. Plain-language monthly reporting shows what published and what it did. The difference between the prompt and the plan is who does the work and how many agents it covers.
The free 90-Day Content Calendar Prompt is for the agent who wants the plan tonight. Run the Find Your Niche Prompt first and the interview goes twice as fast, because the brief is already half written.
Questions brokers ask
How long does the 90-day calendar prompt take?
About 30 to 45 minutes for the interview, the brief and the theme map, and another 20 minutes for month one. Months two and three can wait for another sitting; the prompt picks up from the saved notes.
Why interview first instead of asking for ideas?
Because ideas from a blank prompt are the average of every agent's ideas. The interview forces the towns, the price band and three real situations into the chat before anything is written, so every theme and hook is built from them.
What are the four content pillars?
Market, Listings and neighborhoods, Process and answers, and You. Every week gets one theme from one pillar, and no more than one "You" week in any five, so the feed stays useful to the reader rather than a diary.
How does the quality pass work?
Every hook is checked for five failures: generic, steering, unproven, not me, and duplicate. A flagged hook is fixed by adding one place name or one situation. The final twenty rows have passed all five.
Will the calendar invent statistics?
No. The prompt uses only numbers in the brief, and if the agent has no proof number it leaves the line blank rather than suggesting one. Every figure still gets checked before it publishes.
Which AI tool works best for it?
ChatGPT, Claude or Gemini all run it. The tool page opens ChatGPT or Claude with the prompt already pasted; Gemini opens a fresh chat and the prompt copies to the clipboard. The prompt is written to behave the same in each.
Book a 30-minute strategy call
If the office needs this every month for every agent, that is a system, not a prompt. Bring a finished THEME MAP, or bring nothing, and Urban Marketing Edge will map the plan that runs the calendar for the whole office: monthly themes, written and designed pieces, one approval, a weekly meeting and a plain-language report, month to month, billed in advance, no setup fees. Book a 30-minute strategy call, and see how Rocks Realty and Clareo Group run it in the case studies.
