Gains:
- Compiling indicators such as regional price trend, supply-demand balance and rent multiplier with the support of artificial intelligence and turning them into a meaningful market summary
- Ability to see the currentness and source problems of the figures provided by artificial intelligence and verify them with official / current data
- Ability to convey market interpretation to the customer as a forecast, without guarantee and in accordance with the legislation
The most powerful trump card of a real estate consultant is knowing the market. Only an advisor who can read the market can tell the right price to the owner, the right timing to the buyer, and the right region to the investor. Market analysis; It is to bring together the price trend, supply-demand balance, rent multiplier, sales period and development potential in a region and turn it into a meaningful story. Artificial intelligence is very useful in this compilation and interpretation task: it organizes and compares scattered data, puts the trend into sentences, and creates a summary to be presented to the customer. But the biggest pitfall here is the problem of timeliness and resources: language models do not know the live market; The number given may be old or fake. So the principle is: you bring the data, let the AI interpret it; Do not ask the AI for the number.
Key indicators of market analysis
Let's clarify a few concepts. Price trend is the direction of increase/decrease of the m² price in a region over time. Supply-demand balance is the ratio of the number of listings for sale and buyer interest; If supply is abundant, the price is suppressed, if demand is abundant, it rises. Rental multiplier (rental yield period) is the property price divided by the annual rental income; It is the answer to the question "in how many years will it amortize itself?" (e.g. if the price is 6M and the annual rent is 300 thousand, the multiplier is 20 years). Average selling time is how many days it takes for an ad to find a buyer; It shows the speed of the market. The occupancy/vacancy rate is how much of the rental properties are rented.
Data for these indicators should come from official and current sources: your own sales records, the office's CRM data, official house price indices, portal statistics, title deed/sales numbers and field observation. The AI's job is to read, compare and interpret this data — not produce it.
Attention: Ask a language model "By what percent did prices increase in neighborhood X?" The most common and dangerous mistake is to ask and tell the customer the number that comes up. The model's information is old, not live, and can fit the number fluently. The number comes from you, the interpretation may come from the model.
Step by step: reliable market summary
- Collect data: Compile the last 6-12 months' comparable sales, rental samples, posting periods and supply numbers from reliable sources. Note the date and source.
- Give the raw data to the AI: Say "Only use this data, do not add" and have it calculate the trend, average and multiplier.
- Ask for comments, ask for guarantees: Have the direction of the trend, possible causes and uncertainties explained; Don't ask for a certain promise of the future.
- Translate into customer language: Put the technical summary into plain language for the owner/buyer/investor.
- Verify and tag: Include the source of each figure, note "this is an estimate", update any that are outdated.
Template 1 — Trend summary from raw data:
Role: You are a real estate market analyst. Below is VERIFIED yield for region X. ONLY use this, do not add/make new digits. Subtract: average m² price, 6-month change direction, average sales period, rent multiplier. Write down the uncertainties as well.Data: [paste your own data]
Template 2 — Rent multiplier and yield:
Calculate the rent multiplier (price / annual rent) and gross rental yield percentage with the following data; show transaction. Specify that vacancy rate and expenses are not included. Data: sales price [X], monthly rent [Y].
Template 3 — Market note to client:
Write a 120-word market note to be presented to an owner from the analysis below. Don't say "guarantee" or "it will definitely increase"; Note that this is an estimate based on historical data and the market may change. Analysis: [paste summary]
Template 4 — Source and currency check:
For each numerical claim in the market summary below, the questions "what is its source, what is its date, is it current" appear in a table. Mark those whose source/date is unknown. Summary: [paste text]
Comparison table: data sources
Source
Currentness
reliability
note
Your own sales/CRM record
high
high
Most valuable; actual data
Official house price index
Medium (delayed)
high
Good for trend, limited neighborhood detail
Portal ad statistics
high
medium
Listing price = asking, not selling
Field observation / tradesman information
high
medium
Qualitative insight, quantify
"Memory" of the language model
Low/unclear
low
USING for digit
three mini cases
Case 1 — Strong summary with own data. The consultant gives YZ the 22 sales closed by the office in the last 9 months. YZ extracts the average m² price, the 6% uptrend and the average selling time of 41 days. The consultant presents this to the owner and positions the price accordingly; The property is sold in 38 days. Lesson: actual data is the most powerful analysis.
Case 2 — Fake ratio. Another consultant asked the model, "By what percent has rent increased in this district?" asks; the model says "22%". The advisor tells this to the investor. The actual increase is closer to 40%; Once the investor verifies with another advisor, trust ends. Lesson: don't make the model ask the number, bring the data.
Case 3 — Rent multiplier fallacy. AI calculates the rent multiplier for a property for 18 years, but does not take into account the vacancy rate and expense. The advisor presents this as a "net return". When vacancy and expenses are added, the actual payback increases to 23 years. Lesson: don't present the gross multiplier as a net return; Be clear about your assumptions.
Weak prompt / Strong prompt
Weak prompt:
Summarize how the housing market is in Çekmeköy and how much the prices have increased.
The model fits without live data; no source, date, verification.
Powerful prompt:
Role: Market analyst. Use your own VERIFIED data below, do not add new figures. Take out the trend direction, average m² price and sales period, state the uncertainties and write that it is an "estimate". Data: [give your own]
Common mistakes
- Asking the language model for live market numbers. Risk of stale/fabricated data.
- Mistaking the advertised price for the sales price. The difference between what is desired and what happens.
- Presenting the gross rent multiplier as a net return. Vacancy/expense neglect.
- Turning the trend into a guarantee. "It will definitely increase" is a misleading statement.
- Sharing a summary without specifying the source and date.
Tip: Add a small "source and date" line at the bottom of each market note: "Data: office sales records, [period]. This is an estimate." This single line makes you both reliable and protected.
Microregion and seasonality: the lie of the average
The most common mistake in market analysis is applying a broad average to a narrow reality. The sentence "This much per square meter in Istanbul" says almost nothing about a property; There can be a 30% difference even between two streets of the same district. Facade, street, entrance to the site, distance to school/metro, noise, view — all of these create a micro-district difference. Instead of asking the AI about the average, give it peers in similar micro-locations and make a comparison. The second trap is seasonality: the housing and especially the rental market fluctuates throughout the year (e.g. rental demand increases before university openings). It would be misleading to extend the data of one period to the whole year; I asked AI "What period does this data belong to? Is there a seasonal effect?" Make them question. Third, there are breakout moments: events such as interest rates, exchange rates, legislation or urban transformation decisions can break the market from past data. In such periods, historical AI interpretation ceases to be guiding; Field observation and up-to-date process data come first. Look at the real micro-location of the property and the current period, not the average.
In summary
Market analysis is the advisor's most valuable asset, and AI accelerates that work — but only when you bring reliable data. Compiles the model, compares it, interprets it and translates it into the customer language; It does not produce the number. Your own sales record is the most powerful resource. Write the source and date of each number, maintain the gross/net distinction, make the trend an estimate, not a guarantee.
Application task
Compile (anonymous) sales and rental data for the last 6-12 months for a region you work in. Derive trend summary with Template 1, rent multiplier with Template 2. Do a source/date check of each figure with Template 4. Finally, write a market note labeled “forecast” that will be presented to the client with Template 3.
checklist
- [ ] I brought the figures from a reliable/up-to-date source; I did not have it fitted to the model.
- [ ] I separated the advertised price and the actual sales price.
- [ ] I have clearly written the gap/expense assumptions in the rent multiplier (gross/net distinction).
- [ ] I presented the trend as a "prediction"; There is no warranty statement.
- [ ] I added the source and date of each issue to the note.