The location of a property is one of the most important criteria for apartment hunters. The distance to the next supermarket or the connection to public transport can decide whether someone decides for or against a property.

The German real estate portal Immowelt now offers a location check that generates ratings for every listing. The new feature is based on Targomo’s Location Scoring API and allows visitors to Immowelt to assess public transit quality and “daily needs” access.

“The Immowelt location check is a real and tangible added value for searchers on immowelt.de,” says Felix Kusch, Immowelt Country Managing Director. “Anyone who can assess the location of a property at a glance saves valuable time in the search for a suitable apartment or house. In the test phase, users were already enthusiastic about this practical function.”

The Immowelt Location Check is a benchmark for the attractiveness of a property’s location compared to all other property locations in the region. It is calculated based on the number and accessibility of relevant points. The “Daily Needs” score takes into account the extent to which grocery stores, drug stores, ATMs, pharmacies, etc. are reachable within the area. The score “Public Transit” takes into account the proximity and number of transit stops, the amount of transit lines and frequency of their services as well as the effectiveness of the transit network itself – how good is the network at connecting people across areas.

The result is presented on an easy-to-read, five-point scale: Super, Good, Okay, Moderate and Poor. For the future, Immowelt plans to expand the location check with two additional scores: Family-friendliness and leisure opportunities. These are intended to display information about kindergartens, playgrounds or sports facilities, among other things.

Targomo’s Location Scoring API enables the evaluation of locations by analyzing sociodemographic data and the accessibility of relevant places (points of interest) in the vicinity. The criteria according to which locations are evaluated can be set individually. “The high adaptability of the Location Scoring API was one reason, along with the speed and precision of the calculations, why we worked with Targomo to develop the Location Check,” says Felix Kusch.

“We enable real estate portals to provide their users with additional information that they would otherwise research on other websites,” says Henning Hollburg, Managing Director of Targomo. “This allows them to keep users on their own platform. They can customize the scores and tailor them to different countries, cultures or user types, for example. Even an individual weighting of the criteria by the user herself is theoretically possible – for example, a user could specify whether there should be many or few bars or restaurants in a neighborhood.”

About immowelt:

immowelt is part of AVIV Group, one of the largest digital real estate tech companies in the world.

immowelt’s mission is to digitize all steps of the real estate transaction in the future to make it as uncomplicated and easy as possible for all parties involved. The basis is provided by the wide-reaching immowelt portals, which are among the leading real estate platforms in Germany and Austria. They already successfully bring together owners, real estate professionals and seekers. immowelt supports the uncomplicated search for a rental property, the effective marketing of a property and customized financing of one’s home with data-supported services. Thanks to decades of experience and broad real estate expertise, immowelt thus creates the perfect success story for tenants and landlords, real estate professionals, property owners and buyers.

In addition to immowelt, other leading real estate online marketplaces in France, Belgium and Israel belong to the AVIV Group, which is part of Axel Springer SE.

Here we can add the link to Ubilabs post

The scoring model is a quantitative analysis procedure for putting several key figures, which are difficult to compare with each other, in relation to locations, regions or areas. The aim is to derive decision-supporting measures for each location. As a starting point for the scoring analysis, the evaluation criteria are determined according to the specific use case. In the example for private real estate search, we selected the accessibility and reachability of Points of Interests (POIs) and mobility as evaluation criteria.

The POIs were clustered in three categories:

Another score represents the quality of mobility in the category Transit, bringing the total number of criteria to four. To calculate a score for mobility, it was taken into account that the pure availability of a stop is not very informative. For this reason, demographic data was used to calculate the number of people who can be reached by public transport within 30 minutes.

Depending on the use case, a scoring could also be enriched with more key figures. For example, accessible playgrounds, kindergartens and parks could say something about family friendliness or recreational value. All key figures are converted into a single score value using mathematical-statistical calculations.

The next step is to determine how locations will be scored against the evaluation criteria. In our example of private property search, the calculation of the POI scores is based on the number of POIs that can be reached on foot within 15 minutes. To give less relevance to POIs located further away, they are weighted according to their distance in travel time, with relevance decreasing quadratically with increasing distance.

Calculation:

This means that POIs at 0 minutes distance are weighted 1, at 5 minutes with ¼, at 10 minutes with 1/9, at 15 minutes with 1/16 . 

For the criteria Local Supply, the calculation takes into account that several POIs of the same type don’t provide additional benefit for a resident, as for example five pharmacies. Rather, what counts here is that there is at least one POI of the following types: supermarket, bakery, pharmacy, drugstore, post office, bank, ATM, kindergarten, school, and bus stop.

Since in reality not every user evaluates the respective location factors in the same way, the scoring gives the user the possibility to set the weighting of the single evaluation criteria individually. In this tool, it is possible using sliders, allowing the user to set the importance of each. As a result, all weighted score points are aggregated, resulting in a scoring measure for each location.

A major challenge in location scoring is that the scores are usually distributed very unevenly. This means that most locations have a low score and only a few locations have a very high score. At the same time the aim is to ensure that the ratings are distributed in such a way that the available spectrum is covered (as completely as possible) and a meaningful range of ratings is created

The following example shows a very simple score using the example of gastronomy where the unequal distribution of values is clearly visible. Most Places have a very low score and very few places have a very high score.

To counteract this unequal distribution, the rating is based on percentiles, so that it always adapts to the distribution in the country, district or local area. In the location scoring demo by Targomo and Ubilabs, the user can also toggle between two different area type modes: municipality, or borough. By changing the area type, the total score is calculated whether in relation to the entire municipality, or the borough the location belongs to.

The following diagram shows the distribution of gastronomy scores and the related rating:

Calculation example: A location selected by the user has the score 50 (red curve, right Y-axis). This corresponds approximately to the percentile 0.967 (X-axis). The corresponding rating is then 0.967³ = 0.9 (blue curve, left Y-axis). Values above the maximum get the rating 1.0.

We distinguish here between “score” and “rating”. “Score” means the absolute values whereas the “rating” is a value between 0 and 1 or 0 and 100%.

In general we can distinguish between three different types of scores and their calculations:

Calculations:

While scoring models are originally known primarily from the banking industry when it comes to credit scores, they are also very useful in the evaluation of locations. Integration of a location scoring in maps allows users to quickly gain an overview of which properties or location are relevant for them.

Benefits of location scoring at a glance: