Let's Talk

Location matters

An ongoing look into how location shapes where we live and what we do

[Product] Why Filtering by Location Characteristics is a Bad Idea

Posted by Aron Korenblit on Nov 12, 2018 4:33:25 PM

Blog images - 900px × 500px (1)-1


It’s a question I’m often asked:

“Hey Aron, can we give our customers the option to search for properties by location characteristics, like 8/10 on quiet or 5/10 on cafés?”

If you’ve already heard my monologue on why this is a bad idea, I apologize in advance. For those who haven’t, and are considering it, here are three reasons why we strongly recommend against it.

On your homepage, less is always more

The search page is the very first step of your funnel. It’s your introduction to the client, and definitely not the right time for information overload. Your objective at this point must be to simply move users down the funnel, giving them what they want: property listings. Asking too many questions for them to think about before getting a reward risks turning them off completely.

Like the old adage “shut up and sell”, the trend for selling online definitely follows the school of “less is more” … in this case, showing less on the homepage. Some portals now mask non-essential search options in a “more” category. Others boldly eliminate filters until the second page, with “location” as the only exception.

To see this in action, look no further than the leader of real estate search – Zillow. Visitors to their site must answer only one question to get engaged: location. All other filters are saved for a deeper, more engaging search on the results page.


Source: zillow.com

The Homepage is just not the right time to ask for so much commitment. That would be like assaulting someone who just walked into your open house with 20 questions.

Not a good idea.

Location characteristics are trade-offs, not filters

Remember, finding a home is essentially a game of weighted constraints. You start off with a budget, then a location, then proceed through a series of trade-offs on every possible characteristic to find a home you can live with that fits within your lifestyle criteria.

“Do we mind a little noise if it’s near a good school?”

“What if we had to drive for coffee but could walk to get groceries? Is that worth spending 15K less?”

It’s only after the search results have been filtered by location and price that the home buyer can best make those judgement calls.

Enter Local Search

Local Search allows users to select the lifestyle criteria they care about, once they’re ready to. When selected, each listing on the search results page is scored on a scale of 0-10 of how well its location fits the selected criteria.


Source: www.royallepage.ca

Consider: if we would have filtered properties based on these characteristics, we would have assumed (perhaps incorrectly) that amenities and neighbourhood feel are as important as the home buyer’s budget or the number of desired bedrooms. We would have stopped them from seeing everything that’s available to them within their actual hard requirements.

With Local Search, we don’t filter out low-scoring properties.

If a home buyer considers an amenity to be a hard requirement, they’d never click a bad match. They’d flock to the higher matches and leave the bad matches behind. Would you consider a two-bedroom property if you want five?

Here’s a graph which shows how often users clicked on a property for each possible score and how often that score appears:

Surprise! Users click on low scores. This means by filtering out “bad matches”, we’d deprive users the opportunity to make their own decision making trade-off between home prices, home characteristics and location.

If people click on good and bad matches, does location actually matter?

Of course it matters! The data actually shows us this. But if location amenities and feel didn’t also matter, there would be no differences in how often a score appears and how often users click on properties with that score.

The data shows that home buyers consider location preference by diverting their attention from bad matches to good matches. In other words, if a property ranks below a five out of ten match, a home buyer is less likely to click on the property.

If the property ranks above five out of ten, they are increasingly likely to click on the property. I’ve highlighted the difference in the graph below:


What began as a gut feeling, is now backed by data.

In summary

In summary, we don’t recommend filtering properties by location characteristics such as feel or neighbourhood amenities on the home page because they are not hard requirements for the home buyer.

We do, however, recommend you factor in amenities and neighbourhood feel into your search or discovery experience.

This will help guide users to the right property while allowing them to make their own trade-off decisions on location amenities to fit their budget.

Tags: Product
Aron Korenblit

Written by Aron Korenblit

Customer Success Manager at Local Logic. If I think it's interesting, I'll probably write something about it. If other people find that writing interesting, it'll end up here. Avid traveller & mediocre squash player despite all my best efforts.