It can be a daunting task for consumers to slog through daily MLS prospecting search results and even consumer-oriented web sites to find the listings they are interested in. With increasing property inventories, consumers will need to expend even more effort to find the properties that appeal to them in search results that will likely grow even larger. We can do better for them.
We need to ask ourselves, “When paging through search results, why do consumers click on this property or that one?” Usually, they’ve already set their search criteria and are only looking at properties in their desired geography, price range and (using residential as the primary example hereafter) the right number of bedrooms, bathrooms, required square footage, etc. Of course they desire a bargain, and are looking for properties that are the best balance of price and their other required criteria. But consumers are also highly visual, so they look at thumbnails and click on homes that match the style, exterior, and colors that appeal to them. That’s obvious, right? Well, why are we making consumers page through dozens or even hundreds of properties every day to hunt out the bargains and to find the properties that appeal to them in other ways? We should stop continually sorting on a single arbitrary criteria, most typically price, and start presenting first and foremost the homes that meet the consumers’ desires.
How can this be done? Other industries have already shown us the way. Once you look at a few items for sale on Amazon.com, they start showing you other items you might like to buy. Once you rent and/or rate some movies on Netflix.com, they suggest other movies you might wish to put in your queue. My favorite example might be Pandora internet radio, which lets you set some initial criteria for music you want to hear, then fine tunes your playlist as you rate the songs you hear or move to skip the rest of the song they are playing.
We can use similar methods. We can see what properties users click into to see details. With some application changes, we could probably collect information on and analyze how long they stayed on each detail page. We could collect information on what properties they email to others or request more information on. This can be a subtle task, termed “establishing intent from gesture”, but we don’t need to be subtle. We could also, similar to the Netflix and Pandora approaches, have them actively rate properties as part of their search or even in a separate “getting to know you” activity. The rating can be as easy as “thumbs up, thumbs down”, could be a more sophisticated five-star rating, and we could even ask what aspect of the property was the primary basis of their rating or have them rate different aspects of each property. The more information we have, the more accurately the system should be able to order the properties shown to consumer to meet their desires. On sites where the consumer is identified via custom link tied to their identity or login, we can track more information over time, but even on anonymous usage sites we can collect some information. At any rate, if the consumer always clicks on two-story houses, on colonial houses, on houses with brick exteriors – we have the information (especially in the MLS system) to sort on and show them similar houses first.
Can this method ever be perfect? Of course not – especially since there are various qualitative aspects of property selection that we don’t currently track data for at the current time, and therefore we can’t use it in any type of automated process. When consumers are looking at photos and making those split-second judgments, they may look at landscaping, general conditions/curb appeal, and even house color (the trickiest of any criteria to use). Of course, if we start collecting sophisticated ratings (not just ‘thumbs up/thumbs down’) we can start increasing the amount of data we have an properties and use that information as well. For example, if 43% of 150 consumers rating a specific property poorly did so specifically because of property condition / landscaping, we know that consumers looking for homes in good condition and with good landscaping will probably not like that property. Yes, we have to answer the question of “What happens if the homeowner subsequently improves the condition” … then there should be a new photo and statistics need to be re-set. But what if agents keep uploading new photos on properties to try to “game” the system? And so on. This isn’t simple, by any means, and again, it won’t be perfect – but our MLS prospecting results and public site search results could be a lot better than they are right now, and we owe it to ourselves and the consumer to try to improve the property search experience.
I should note that experimenting with this approach could even benefit the real estate professionals, providing them with business intelligence on properties they have listed or even giving them more insight into the buyers they represent. It may provide support to tell the seller that it’s time to fix up the front yard or make other property improvements. And we’ve all heard the phrase, “Buyers are liars,” that they can tell you that they must have one thing in a home, then go for something completely different. We could start collecting the type of information needed to more fully understand their needs and provide them with better service.
This blog entry was a continuation of Improving Prospecting and is complementary to Future of MLS Features – 2008.
Share this post: