How to turn a real estate website into a market research

Written by Fernando Maciá

Managing a real estate portal opens up great opportunities to gain first-hand knowledge of changing market trends in real time. The fact that an increasing number of people looking for housing first turn to the Internet to obtain information, compare prices and find out about availabilities means that many portals and websites with real estate offers register hundreds of thousands of visits per month.

By applying the appropriate registration procedures and analysis tools, these visits can become a very valid sample that will allow us to gauge the preferences of the demand in aspects as diverse as location, types of housing, price ranges, etc., even segmenting these preferences by countries, areas of origin, etc. For the purpose of this article, we are going to focus essentially on three key data of the traffic that a website or a real estate portal registers:

1. Origin of visits and referrals.
2. Use of the internal search engine.
3. Conversion analysis.

Each of these three types of records will produce a series of data from which, properly analyzed, we will be able to extrapolate interesting conclusions.

1. Origin of visits

Practically all the software packages used for web traffic analysis allow us to know the origin of the visits that come to our website. This localization is done by the IP number of the connection. It is not the purpose of this article to go into a technical explanation of the IP number. Suffice it to say that every time we connect to the Internet, we do so through a unique number called an IP address. These numbers are distributed worldwide by zones, so that by knowing the IP number it is possible, in many cases, to know approximately the origin of a visit. Anyone who is curious can check it out by visiting the website www.hostip.info. When visiting this website, any visitor can see the IP number of his own connection and this website will show him, through a Google Maps map, the approximate location of this number with an accuracy at a neighborhood level.

In the previous paragraph we said that location by IP number works in many cases. Unfortunately, it does not always do so. In Spain, for example, people surfing the Internet using Telefónica’s ADSL connection do so through special servers called “proxy-caché” which, although they generally improve the speed of surfing the Net, prevent the exact location of a visit by IP, since all surfers accessing the Internet through this type of server share a single public IP, so that all their visits would be assigned to the location of the proxy-caché server and not to the location of each user.

Localization by IP

On the other hand, for many European countries and the United States, IP geolocation can tell us from which neighborhood, city and country a certain visit has taken place. To get the most out of this data, our web traffic statistics should be able to show separately the access data for each language version of our website. In this way, we will have clearly separated the origin data of the visits that accessed our website in English, German, Dutch or Spanish and it will be much easier to identify which geographical areas concentrate the greatest interest in our products. This information could help us when deciding on off-line promotion strategies (where are we more interested in contracting advertising in the press, in word ads or in organizing a showroom in a hotel?) as well as to check the return on investment of the different versions of our website in each language (in which language is it profitable to develop new content?), etc.

Knowing the geographic origin of the visitors to our website will also help us to segment each of the following data in order to profiling demand and positioning the website for each country (are the British looking for more villas on the Costa Brava or townhouses on the Costa Blanca?), as we will see below.

Once the visits have been located on the map, the second piece of information that will provide us with a wealth of information is the referent that originated the visit. To understand, visits to a website can come directly, either because we have the corresponding URL address stored in the favorites, or because we type it directly into the URL address field of the browser. But most visits come indirectly: the user arrived at the site after clicking on a link on a website other than ours: this is what we call referrer. Internet search engines such as Google, Live or Yahoo! tend to generate the most traffic, followed by directories and other websites.

Referrer information is found in virtually all traffic analysis packages. In some of them, the referrers will be separated into visits coming from search engines and visits coming from any other website. In the visits coming from search engines we will also find the section “keywords” or some similar term. This section contains the search terms that users entered in search engines to find our website. We can learn a lot from the analysis of these search terms. Let’s see how.

How do our visitors formulate their searches?

Looking at the list of keywords used by visitors, we can immediately see that search formulas are repeated, such as “type of property” + “location”. This gives us clues on how we should write the titles and texts of our product sheets to get closer to the way users search, as this will increase our chances of being at the top of the search engines.

Extracting each word separately from the complete search phrases

In this way, we will also be able to extract trends about where users’ preferences are heading: did “Madrid”, “Barcelona” or “Bilbao” appear more in the last month, did “bungalow”, “townhouse”, “penthouse”, “studio” or “duplex” appear more, did “sale” or “rent” appear more…?

Cross-referencing data

And, obviously, by cross-referencing this data with the geographical origin of the visits, we can begin to draw demand profiles for each country: did more visitors come from Great Britain looking for “villas” or “semi-detached”?

At this point, some may object that the number of visits must be very high for the sample to be sufficiently valid to support this type of extrapolation. However, any owner of a small or medium-sized real estate company that has a website with an adequate traffic registration system will immediately see that the number of visits registered by his website easily exceeds the number of visits he is able to attend to daily in his physical office. In addition, visitors to your site are under no commercial pressure and have no need to appear as they are guaranteed anonymity. So we can presume that the data we collect from our website can have a level of validity at least equal to, if not greater than, that which we collect on a daily basis from our face-to-face or telephone sales interviews.

2. Use of the internal search engine

The second moment where our visitors provide us with an enormous amount of information about what they want is when they use our internal search engine. Almost all real estate websites have an internal search engine where visitors can filter, from the properties that make up the available portfolio, those that meet a series of search criteria such as, for example:

• New or used housing.
• Rental or purchase.
• Type of property: townhouse, penthouse, apartment, apartment, bungalow…
• Location: city, neighborhood, beach or golf course, coast or inland…
• Price range.
• Etc.

Given the level of visits to a real estate website, recording each of these searches and making a statistical analysis of this data is equivalent to having a sales team continuously taking note of the preferences of the thousands of potential buyers who visit the website. In short, each of these searches is telling us the type of property the user is looking for, in which area they would prefer to live, whether they prefer it new or used and even the price they would be willing to pay for the property. All collected anonymously and without any commercial pressure. Crossing this data, again, with the geographic origin of our visitors allows us to draw a new demand profile where by countries, regions or cities we can establish the types of housing preferred, the areas that are growing or decreasing in interest for demand and the price levels that potential buyers could accept as reasonable.

The main advantages of a statistical analysis of this type is that it will work as a market survey in real time, which will take advantage of the thousands of monthly visits to our real estate website, to draw a profile of what the demand wants at all times at a very low cost, in real time and with availability of permanent consultation. The data from this analysis can better focus our efforts to attract properties for sale to those areas and types of properties most in demand at any given time, optimizing the performance of our website by bringing our offer closer to what the demand requires.

3. Conversion analysis

Finally, we must also consider the information that the client gives us about himself and the satisfaction he has obtained by visiting our real estate website. This could include aspects such as repeated visits to the website, number of pages viewed in each session, etc. But for the purpose of this article we will look exclusively at the measure of conversion. We say that a website visitor becomes a customer of a website when he/she fulfills a certain objective that we have set for the website. In the case of an e-commerce website, it is clear that the clearest measure of conversion occurs when a visit culminates in a sale. We could then say that if a website receives 100,000 visits and from these 1,000 sales, it has a visitor-to-customer conversion rate of 1%. In other words, out of every 100 visits, 1 ends in a sale.

In the case of real estate websites, this measure is not so clear, since the sale is not completed online. The objective of the real estate web is, in most cases, to provoke the first contact of the potential client with the real estate, either in the form of a phone call, an e-mail, etc. But there are other points on the website that indicate that the information about a particular property may have at least provoked some interest in the user. So we can distinguish two types of conversion:

• Off-line conversion: this will occur when, after a visit to the website, the user makes a phone call or a personal visit to the real estate or sales office. It is very important to establish methods of measuring this off-line conversion in order to fairly measure the contribution of the web to the creation of value for the company. The appropriate way to measure offline conversion can be, among others:
  • Collection of data in a sales questionnaire that captures the origin of the visit (whether it was triggered by a billboard, a radio spot, a mailing or a web visit).
  • Publication of a unique information telephone number for the web, different from any other means of diffusion: if we include a unique telephone number on the web, we will be able to measure that all the calls received on this number were originated by the presence of the real estate on the Web.
  • Inclusion of some kind of incentive so that the user does not forget to mention the origin of the web of his visit in his contact with the real estate.
• On-line conversion: the clearest form of on-line conversion would be the completion of a contact form by the visitor interested in learning more about a property. However, there are a number of other actions that also indicate a higher level of user interest in a particular property, development, housing model, etc. These actions may be:
  • Move from the property file page to the specifications page or floor plan page.
  • Download the promotion brochure in PDF
  • Use the mortgage installment calculator on the price of a specific property.
  • Fill in a “Call me” form so that the real estate agency contacts the user.
  • Fill in a description form of the property you are looking for.

The analysis of the conversion pages clearly shows us which properties, developments, housing models, areas, etc. aroused the greatest interest in visitors to our website. Statistical monitoring of this data and cross-referencing it with the origin of the visits will also enable us to establish demand profiles by country to gain a more detailed understanding of the preferences of potential Spanish, English, German, etc. buyers.

Your real estate website, much more than a showcase

As you can see, a well planned and managed real estate website is much more than a powerful showcase (very soon, the main one in this sector). The multiple tools that we can use in the analysis of recorded traffic enable us to extract a very clear picture of demand trends and to know where the market is heading in order to act before our competitors. All this at tremendously reduced costs and in an automated way, in real time, with full accessibility to data and information, and with infinite possibilities of personalization in the crossing of data, criteria, segmentation, etc.