We Calculate the Value of Your Real Estate with the Power of AI
How We Calculate Value?
1. Monitoring
Our database is updated instantly with the actual sales closing values obtained from the real estate agents, the data collected as a result of the field studies carried out by the experts and the researches made on the internet.
2. Analysis
Data entered into our databases is subjected to internal consistency and accuracy checks and data that do not reach the required reliability level is eliminated.
3. Value Calculation
Endeksa provides, data-oriented, reliable real estate value by using machine learning techniques.
What are the Data Sources of Endeksa?
Endeksa monitors all sources that may affect real estate prices. In addition to resources on the internet, closing values reported by agents, valuation reports made, sales values received from our business partners and all data collected are evaluated both in terms of quantity and quality and included in the analysis.
| Usage Area | Data Source and Supplier | Update Frequency | Update Time | Last Update |
|---|---|---|---|---|
| Real Estate Value | Agents, Appraisers, Web - Endeksa | Instant | Instant | Nov 12, 2025 |
| Demographics, Census*** | Turkiye: TurkSat - Adres Harita Spain: Instituto Nacional de Estadística Portugal: Instituto Nacional de Estatística | Annual | Q2 | Jun 30, 2025 |
| Estimated Income | Turkiye: TurkSat - Adres Harita, Endeksa Spain: Instituto Nacional de Estadística Portugal: Instituto Nacional de Estatística | Biannualy | Q2/Q4 | Jun 30, 2025 |
| Socio Economic Status | Turkiye: Adres Harita, Endeksa Spain: Instituto Nacional de Estadística Portugal: Instituto Nacional de Estatística | Biannualy | Q2/Q4 | Jun 30, 2025 |
| Election Results | YSK - Adres Harita | Election | Following Month | Aug 14, 2024 |
| Citizenship | TurkSat - Adres Harita | Annual | Q2 | Sep 15, 2025 |
| Administrative Borders, Streets | Adres Harita | Annual | Q4 | Jun 30, 2025 |
| Address Information | Google, Yandex | Instant | Instant | Nov 12, 2025 |
| POI Information | Web, Yandex, OSM | Instant | Instant | Nov 12, 2025 |
What are the Methods Used in Value Calculation?
The calculation methods produced by interpreting the idea that a machine can think like a human which was put forward in the 1980s, according to the real estate market, are Endeksa's unique approaches. Based on the concept of thinking like a Real Estate Professional instead of thinking like a human, it focuses on the basic approaches of these people and international standards. With the developments in technology, Endeksa uses machine learning methods that are only possible today.
While calculating, the following analyzes are applied;
1- Narrow Market Analysis
It prioritizes the similarity of the comparable. The similarity threshold to the real estate whose value is to be estimated is kept high. It achieves the most similar comparable by providing a homogeneous distribution.
2- Deep Market Analysis
Prioritizes close distance. The closest equivalents to the real estate whose value is to be estimated are reached by going back in time.
3- Wide Market Analysis
Comparable are prioritized to be close in time. Similar to the real estate whose value is to be estimated and the most recent equivalents are found in the distance by gradually opening them.
4- Endeksa AI®
AI models dynamically find value-influencing parameters and use much more comparable. It produces sharp results because it can analyze according to similar regions in unique situations around it.
Why Transparency in e-Valuations Matters?
In Endeksa, values are calculated simultaneously with different analyzes. Each method is checked for consistency within itself and is scored on the following basic criteria. While Endeksa shows the result with the highest score in the first stage, it also provides access to other results and provides transparent explanations about the calculations. In this way, you can check the reliability of the calculated value yourself.
80+/100 👍🏻
'Higher score' = 'Better result' = 'Narrow estimation range'
We are confident in the calculations having 80 points and higher.
The following criteria are taken into consideration in scoring;
1- Standard Deviation
The closeness of the comparable forming the model to the mean is being controlled. If the standard deviation is small, it is concluded that the market is homogeneous and the model's score is higher.
2- Distance
The distance of each comparable forming the model to the real estate is considered as a criterion. The closer the comparable used, the higher the model's distance score.
3- Time
The creation dates of model-forming comparable are considered as an other important factor. As the creation date of the used comparable decreases, its relevancy score increases.
4- Feature Matching
The dates of the comparable forming the model are accepted as criteria. The newer the date of the comparable used, the higher the score.
It All Starts with The Value
Learn The Real Value of Your Property
TRUSTED BY
Industry leaders
ENDEKSA IN PRESS
As appeared in
