So the blogosphere has been lighting up with references to the recent launch of Roostsm. It has been quite interesting to read the reactions to the product on many levels.
Having played some small part in helping to bring this product to market, I though I would make an effort to give a description of the technical and business model.
Please understand that I am going to attempt to remain neutral on the topic, identifying both potential strengths and weaknesses in what I think is a fascinating, buy side enabling technology.
Roostsm; is essentially a virtual IDX directory that feels like a vertical search, but fortunately is not. That is important, so I need to take a moment to elaborate.
First, a bit broader of a statement. Roostsm is a very effective real estate search interface that is fast and quite intuitive. It is a well branded, well funded product designed to allow an interested broker to sponsor a property search which ultimately showcases properties on that local brokers site.
In a sense it allows a broker to directly control their marketing expenditure in a manner that closely resembles modern SEM models.
The site launched January 23rd, and is currently running with live content in several markets and I am told expanding rapidly.
A vertical search in this context is a bad thing. Vertical searches are a way to describe a search engine that delves straight into specific content gleaned from other websites (this is the traditional definition in my view, and can certainly be argued).
Yahoo Shopping is a very good example. When you search from this search engine, the results returned are a subset of what Yahoo sees on the web.
It is a vertical search that pulls the searcher straight down to content on that plane - in this case things to buy.
In the case of Yahoo, this sort of vertical search is great, because the websites that they glean this content from A) Want that information to be consumed by aggregators like Yahoo so that you can find it and purchase it, and B) because those sites are the definitive source of that information.
Since they are the definitive source, they are well incented to have ample quality control measures in place to ensure that the aggregated content is accurate.
In the case of real estate, this is not true. Property information exists in so many places, some ‘legitimate’ sources of data with at least an effort to be current and accurate, some not so much so.
‘Gleaning data’ from these sort of disparate, non-definitive sources is a losing proposition.
There IS of course a definitive source of this information, but often the MLS is divorced from the public display of the information, and I think this is a good thing. MLS applications are challenging enough in their own right to not have to be concerned with display, which is a venue that is changing frenetically.
Consider. Two weeks ago I got a call from broker asking why we (my MLS/Association) were sending bad/old data to a website. I won’t name them here, but we can just call them site x. I replied that of course we are not, and have no affiliation with site x. The ‘data’ displayed there was from his listing three years ago. Site x is a prominent ‘deep vertical’ search engine.
I have a theory that with enough examples I can provide a formula to accurately predict the degree of listing data error as it diverges from the source, specifically because of all the steps involved in acquiring and/or moving the content from place to place.
Even the best syndication will see an echo as details change at the source and flow downstream… if you don’t ask the source directly every time, it is wise to limit yourself to one step removed from that source.
So, Roostsm, again, feels like a vertical search but is not. ‘Feeling’ like a vertical search is a very good thing. Vertical searching is cool.
Many definitive content sources, one search result interface. I don’t have to go ford.com and toyota.com I can just go to cars.com and get them all. But in real estate it just doesn’t work.
Even large search engines like Yahoo and google have suffered from the vertical search dilemma I have described above. If they want to know the solution to that problem, I have access to the working prototype, but cannot divulge the details. Drop me a line and lets talk.
In the case of Roostsm, it feels like you are having a vertical search experience. The content is ‘market fresh,’ with rich listing content flowing in from many areas.
But this content is not gleaned from other sources. It is official, quality controlled content (not data). The search is therefore modeled on top of the source content - one step removed - in the care of a highly reputable and responsible vendor. That vendor IS in the business of displaying content well and rapidly adapting to the needs of the market for search solutions.
This makes the Roostsm search a plugin.
So lets examine that plugin to this ‘official’ content store. The integration is so seamless and essential that in the end Roostsm cannot really be called a search at all. Roostsm is really a hub of activity to relay a public consumer (lead) to an interested broker’s own IDX website.
The exchange from Roostsm down to that individual brokers site feels so much like a vertical search because of the richness and quality of the content AND because the experience is so seamless and smooth. This fluidity is due to the use of a virtual IDX site that then passes you directly on to that interested brokers own website when you click an individual property.
Roostsm takes the consumer, passes them to a cities worth of results via a plugin to an IDX virtual site, branded precisely like the interested brokers site. The next click is a seamless transition into that interested brokers site.
Let me give you an analogy from the consumers perspective. Picture a river. You are fishing for properties in your kayak. And of course there are lots of kayaks on the river. All by yourself, it is hard to ensure sufficient quality properties to make for the best fishing (searching). Some searches are all fished out, others results need to be thrown back.
Now picture a riverboat. Very Mark Twain’esque, with the revolving paddlewheel like a windmill on the back to propel it.
Instead of scooping water to propel the boat, it lifts listing content up to the group of well dressed, well informed consumers sitting on the riverboat enjoying a nice search outing. The interested broker simply places some modicum of her advertising budget into controlling how fast that wheel rotates to pull the listings content up to different consumers on the riverboat.
The more she wants to invest, the more likely her interest in representing a qualified buyer will be expressed by her position at the top of the search page and the fact that the consumers clicks will navigate to her broker idx webpage. If she is not represented to the consumers waiting for the content as the interested broker, then she is not charged.
Let me try it from the reverse (the brokers side). Picture a series of turnstyles (like you find at the transit station). Hundreds of them. And lined up looking at them are brokers eager to serve a consumer that they know is out there. To try to meet that consumer, they have to invest in a choice… which turnstyle to go through.
The only caveat is that some of the turnstyles don’t really work well, and some just don’t work at all. So you need to invest wisely in that choice.
What Roostsm attempts to do is provide a very well oiled, high performance, centralized turnstyle. This mechanism takes one interested broker at a time, and places them on a well funded marketing platform. If you don’t want that service, then you simply don’t opt in. You still get the benefit of additional traffic to the network of IDX sites powered by your MLS and its content, which in the end brings you buyers.
So lets take a look at the interested broker. Why is she interested?
Well, of course because she is trying to help you to buy a property, and she knows that you are looking online. She may be trying to get leads for her own listings as well as qualified buyers to work with, or she may just be a great farmer of territory and has a laserlike focus on the buyer. As a listing agent, you WANT her, you NEED her. This is her specialty. She can bring you the qualified buyer and get the deal done, but she has to be able to find the buyer. She has to be able to spend her dollars advertising her listings and soliciting qualified buyers intelligently.
The prospect of an effective search that feels like it is vertical, but has quality real estate content (and of course has a mix of the latest technology and blazing speed), appeals to her greatly. What appeals much more is the fact that this single portal allows her to both market her listings well, as well as enable her to express her interest in marketing her own services. The degree of her interest can be carefully controlled by her investment in sponsoring searches.
I hear agents all the time referring to trying to nail their business plan down in terms of their ideal volume of transactions. That may sound like a bad joke in this market, but regardless of how many are actually moving, the dedicated agent needs to keep marketing themselves and their content.
For many in this market, it is they - themselves and their ability to work to find an ideal buyer for your listing - that IS the content they offer. Being able to tightly and carefully throttle that interest level, and examine in detail the return on that investment, would seem to be desirable.
I have heard the critique “I am not sure listing agents are going to want to be having leads sent to another brokers site when someone clicks on details for THEIR listing.” I can understand the sentiment. In an ideal world, you could pocket all of your listings and still sell them yourself by generating the immense amount of traffic to your own resources required to serve your clients well.
Certainly, if that is your goal, you would want to opt your listings out of internet display and market them solely from your brokerage. There are some other suggestions that I could make, and possibly some development work that could be done for you; but in the end isn’t that just the point? It is easier and usually better done by those that focus on this as their core business.
However, knowing that you as the listing agent are always given proper attribution according to the rules for public display of MLS content, and that this is ensured by one of the most reputable vendors in the industry, would seem to be soothing.
Knowing that this model can bring the sort of leads that typify the vertical search experience, e.g. people genuinely interested in precisely and solely what you have to offer, as opposed to the stumbleupon sorts, would appear to be comforting.
Knowing that the investment in quality is matched by an investment in bringing the traffic to warrant interested brokers’ attention, would probably feel empowering.
If, in the end, the worst criticism that can be leveled at the application is that it creates a sort of ‘prisoners dilemma’ whereby brokers feel the need to shift some of their advertising dollars to this venue, I suspect that the authors will happily accept that.
Of course, this dilemma already exists across the spectrum of advertising choices that brokers make daily. The fact that this venue at least offers the ability to tightly control the expenditure and map it back to metrics and results, would seem to be a huge step in the right direction.
So that is my take on the Roostsm model. I heard it described by someone as magic. It is not magic, because of course magic doesn’t truly exist. It is a very well thought out, seamlessly executed prestidigitation to get the consumer to the brokers website.
The search is invoked from Roost.com, but actually served on the content backbone of a very reputable vendor (iHomefinder. Having personally inspected their content provisioning and compliance system, I can tell you that they are precisely the state of the art). The first click takes the search to the sponsoring brokers existing local IDX search page, and you have landed. The transition is smooth and effortless.
It is a fascinating way to bring quality content and targeted searching together, while appearing to honor the rigor and rules of the traditional real estate industry.
Now just to add my own little additional wrinkle… What I really think would be fascinating is if Matt and Alex over at Roostsm could give the sponsoring broker some criteria to choose from about what sort of properties they can most effectively help connect someone with. They have obviously already nailed the geographic localization part of this riddle. The ability to express your interest as a broker to represent buyers in hyper local focus but with national presence and SEO is quite a nice combination.
But what else can be done in this regard? If you gave brokers a list of twenty questions about the types of buyers they can most effectively represent (e.g. SRS designees would presumably identify servicing that generational cohort as a specialty), but only give them the choice of 5 or so specialties. Now that would be really interesting. At that point you are qualifying the lead somewhat by the search mechanism itself, and qualifying the broker to the consumer by affirmation.
The interested consumer and interested broker both walk away from the experience feeling like they just met their perfect match. 
And then the turnstyle turns for the next interested broker.
If this sort of capability can meet a changing market in 2008 as the best time to buy becomes ‘apparent,’ then a lot of folks will be coming home to Roostsm.
Love it or hate it, Roostsm seems to offer a next generation platform for buyers agents to market their services against the most powerful backdrop of all - compliant, ‘market fresh’ MLS content. I have to hand it to them; Roostsm gets an A in my book. If for nothing else than the elegance of the model.
Product review: A
Author: Michael Seguin
