Google’s new algorithm and real estate results

Last week, Google rolled out a new algorithm change. For the past couple years most changes to the algorithm — the bit of math and logic that determines who really is No. 1 for a given search term on Google — have been fairly subtle. This new one is not.

Google says that results for more than 10 percent of the search queries are significantly different with this new algorithm.

While most of the press around this change has focused on large advertising-based business models such as Demand Media and AOL, any business that gains value from "long tail" searches will be affected by this change.

Real estate is an industry that gains value from long-tail searches — those searches that are fairly specific (an example: "condos for sale in the old north end, burlington vt").

What this Google update is about

Over the past couple of years there has been growing concern over the quality of the search results on Google. Simultaneously, social media is driving more and more traffic to websites each day — raising the potential of advertisers abandoning search marketing for social marketing.

For Google to maintain its golden goose — advertisements sprinkled along the margins of search engine results pages — it needs to maintain the quality of results that the search-engine delivers.

If people think the search engine just delivers low-quality, spammy results, then Google will be dropped just like we all dropped AltaVista back in the day.

This new algorithm is meant to deliver higher-quality results from Google. This is important because it means Google has some logic/math system for determining the quality of content.

This means that the quality of your real estate content, something I hope you’ve been paying attention to all along anyway, is now being mathematically calculated.

How the new algorithm affects real estate search results

Google has chosen to roll this update out country by country, starting with the U.S. This gives us an opportunity to learn a little bit about how this algorithm works for real estate.

By conducting searches from non-U.S. locations on the non-U.S. versions of Google, we can see what the search results were before the algorithm change. Comparing those results with what we get in the U.S. lets us understand how Google is interpreting content quality in real estate websites.

I did a series of four searches on different U.S. markets for fairly generic real estate search terms. Then I performed some searches on non-U.S. Google via a non-U.S. proxy.

There were indeed differences. But before I present the data I do need to bring up some of the caveats — the different locations will have an impact on search results; this was just a one-time dipstick check; it’s only four markets; and the algorithm will probably adjust as it settles in. All the same, this is the data.

I chose four different markets to sample to get a selection of market sizes and geographical differences. The table below shows some of the aggregated data for the results of these four searches.

• Aggregator is any site showing real estate content that doesn’t make money from being involved in real estate transactions. This would include newspaper sites, Trulia, Yahoo, Realtor.com and so on. These sites typically make money from advertising.
• Broker/Agent is any site showing real estate content that does make money from being involved in real estate transactions but is not a national brand site.
• National Brokerage is any site showing real estate content that directs leads toward an agent/broker affiliated with the brand.
• Domains with MLS data include sites that show information that is available on any website that displays multiple listing service data.
• Domains with blogs include sites that have a blog on the domain (website.com/blog but not blog.website.com) and newspaper sites that have major content sections on the same domain as the search result.

Some thoughts on this initial data

(Reread the paragraph on the caveats and then continue.)

When I first heard about Google’s new algorithm being focused on removing links that they considered to be "low quality," I was curious how this would affect sites that show MLS data. By its nature, all MLS data is the same. It would be difficult to figure out which copy is the quality one.

Turns out there’s really no big difference in this sample between sites that have MLS data or not. Across all four markets, two fewer sites in the top five have MLS data available on them — not a big enough change for me to get overly excited (of course, if those two sites were mine then it would be different).

The biggest loser in this admittedly small sample appears to be national broker sites — sites that have a lot of MLS data but little local content.

In detail as well, many of those national brokerage site placements have gone from being within the top two slots to being nowhere on the first page at all.

There’s also the issue of what happened to the 39 million sites that aren’t listed as part of the approximate results. Some of that could well be the difference in geolocation but I’m not sure why a search for U.S. property originating in Australia would show so many more results.

I suspect some sites that are exceptionally heavy on duplicate content have been delisted entirely.

The biggest winner was the agent/broker category. In fact, the only winner was the agent/broker category — not only in terms of an increased number of broker or agent sites breaking into the top five, but also in terms of actual rank.

While not yet taking the top spots, agent/broker sites climbed at least as high as position No. 2 in three of the four markets.

Timing of this update may be challenging

Short forays into the data like this will likely raise more questions than answers. And seeing things in aggregate doesn’t necessarily help at the direct local level. The important thing here is that watching how the new algorithm ranks the content of your site is a good idea.

If your business derives a benefit from search traffic you will want to be aware of where you stand today, because where you stand today could well be different from where you stood last week at this time.

To make it a bit more complicated, in markets that have a distinct summer real estate season there isn’t really enough time to develop an effective search engine optimization effort to capture the search traffic between now and then. (If this applies to your market, you’ll need to plan alternative methods of reaching your audience this year while prepping your search efforts for the fall and next year.)