Over the past month, there has been a great deal of discussion in the real estate industry regarding Zillow. The debate seems to center on the perceived inaccuracies of its home valuation functionality. This thinking reflects a rather narrow tactical perspective.
Zillow.com has rocketed into the fifth ranked spot in the ComScore Media Metrix rankings of real estate sites with 2.77 million unique site visitors. It’s time for a blinding glimpse of the obvious: the dogs like the dog food. Consumers want what Zillow offers.
Smart real estate brokers and franchise operators have already accepted this and are now moving to provide similar content and tools, including automated home valuations, sales comparables and property data. They know that attracting consumers to their sites means an opportunity to build brand awareness and establish relationships that can turn visitors into clients.
Brokers seeking to provide this type of content will need to answer several interrelated questions:
What will be the implementation method? There are three basic choices:
(a) Frame-in. The simplest method is to “frame” a partner site that already provides the relevant content and functionality. This involves effectively no software development by the broker. Since this means sending traffic to that partner, the broker will need to be comfortable that there is a clear alignment of business objectives.
(b) XML. This represents the middle choice in terms of development effort. Your tech team will need to build an interface for the site but the application ‘calls out’ to your vendor’s server to actually get the data. If you want to display a home valuation or property data for a specific address, your application just grabs the relevant data in real-time from the vendor.
(c) Bulk file license. This option is probably available to only the largest brokerage and franchise companies, as it requires, by any measure, a large development effort. It does offer the greatest flexibility in terms of designing the functionality and integrating multiple data sets such as for-sale listings, past sales and assessment data. The vendor will provide you with the data via an initial feed (think multi-gigabit file size) and weekly or monthly updates. Your tech team will need to develop the database and application since it will all be housed on your company’s servers.
Which vendor(s) will you use?
There are five primary vendors for public records property data: First American RES, Fidelity NDS, DataQuick, REIData (Stewart) and OnBoard LLC. Develop a detailed understanding of each company’s data. What is their geographic coverage? You should understand not just the total number of counties but specifically which counties. Collecting data from 10 counties in Arizona doesn’t mean much if the vendor doesn’t cover Maricopa (Phoenix). In each county, does the vendor collect both assessment and deed transfer data and, if so, how often? If you are interested in national coverage, there are roughly 2,000 counties so you’re talking about a sizable matrix necessary to compare vendors. This process can be simplified somewhat by comparing the population coverage of each vendor for the relevant geographic area.
Carefully define your requirements in terms of approved use, delivery format, data elements, update frequency and geographic coverage. Limit the licensed data to what you really need to serve a consumer audience. By managing a “request for proposal” process, you can structure licenses that don’t include unnecessary and expensive data elements.
For most brokers, hiring an army of statistics geeks to build an automated valuation model (AVM) engine makes no sense, particularly since there are many market-tested models already out there. Key questions to ask vendors are: is this a single AVM engine or a cascading platform that checks multiple models for the best valuation? This can also have implications for the “hit rate” or percentage of addresses that produce a valuation, which is also important to understand.
What is the scope of investment for this functionality?
Draft your product requirements document before you contact a vendor. The scope of investment changes markedly based on the content and functionality you want to provide on your Web site and the implementation method you chose. A simple frame-in requires almost no investment whereas the annual license cost for a bulk file license can run several hundred thousand dollars. However, a bulk file license generally has a fixed annual cost, whereas the XML option is often priced on a transaction basis. The more visitors that use the functionality you provide via XML, the higher your license expense climbs.
We would also suggest that brokers think about the broader context rather than follow a simple copycat strategy. The Internet automotive industry provides a good case study for determining where real estate is headed.
An Automotive Analogy
In the early days of the Internet, content on automotive portal sites consisted primarily of used vehicle listings aggregated from the inventories of local car dealers. The content of each listing consisted largely of a sale price, some advertising copy and dealer contact information.
Over time, it became clear that consumers wanted more information such as a consistent vehicle description and one or more photos of the vehicle. Research portals added stock photos, reviews and specifications data (moon roof, seat type, radio/CD, engine size, fuel economy, etc.). But what really fired up consumers was valuation information. For new vehicles, this meant “Invoice Pricing” not just the manufacturers’ sticker price. For used vehicles, in addition to the dealer’s sale price, the Blue Book value was displayed right on the same site as the vehicle listing. Fast forward 10 years and consumer research sites like Kelley Blue Book, Edmunds and NADA Guides are among the most popular automotive sites enjoying in excess of 10 million unique visitors per month.
As the auto industry demonstrates, the key content required for a good transactional research site consists of (1) listings content, (2) consistent specifications data and photos and (3) valuation information, not just sale/listing price.
So where is the real estate industry heading? Eventually it will reflect the same tightly integrated combination of for-sale listings, property data and valuation information. And just as automotive sites display a commingled search result of both dealer and private party listings when the consumer wants to search for a Ford F150, so too will consumers expect to see broker-listed properties, new homes, foreclosures and (gasp) FSBOs in the same search result set when they look for a home in Santa Monica. Home buyers don’t consider this notion heretical and would, quite understandably, just prefer to see all properties currently available for sale in their area that meet their needs regardless of the economic arrangements made by the seller.
Will this happen quickly? No, unlike NAR and the MLSs, the National Association of Auto Dealers does not exercise effective monopoly control of all used vehicle listings inventory. But brokers that want to get ahead of the curve can draw a simple lesson from the launch of Zillow: Internet consumers want comprehensive information presented objectively. Buying and selling residential property is a complex and high-value transaction that all evidence suggests is best handled with the advice of a trained professional. Leading brokers will have confidence in the value of their services and their ability to communicate that value to consumers. Give the people what they want and let’s stop playing catch up.
Stephen Bedikian is a partner with Los Angeles-based Real IQ Consulting, which provides marketing analysis and consulting services for the mortgage industry. He can be reached by e-mail at sbedikian@realiq.com.
***
What’s your opinion? Send your Letter to the Editor to opinion@inman.com.
