Aggregating The Long Tail: Making Micro Useful

There is a lot of talk about micro-blogging in the real estate blogosphere but not a lot of discussion about the issues or how to deal with micro-content in the aggregate.  

To start, I like the idea of local content (hence the name ForSaleByLocals for the engine that drives our sites…created long before “local” was popular) – We talked about microsites almost a year ago and “local” was the idea that started the development of our software. However, local and micro content are not easily managed on a macro level and hence a problem still to solved. “Better late than later”, I always say…  

The biggest single issue for me in the current management of local content (local blogs, micro-content, multimedia content, and single property blogs) are the lack of spatial relationships among the data. Nothing describes the relationship among the data and using those relationships to add value. Said differently, people seem to be thinking about the long tail content for their own SEO advantages but losing sight of the fact that the overall body of varied long tail content in a specific geographic area is where the value is for the end user. There isn’t any site that I am aware of doing it quite right…yet.

Issue #1: Lack of spatial context for micro and local data

With vast amounts of local data brewing on the horizon, we need to start thinking about how this data will be aggregated to provide value. 

Let’s start out at the micro level.  I’m a buyer considering a house on Elm street in Town Y.  No microsite or single property site exists for properties on Elm street but does exist for Shaw street, only two blocks away (likely relevant information for such a search). Google won’t give me that information as it is current configured but shouldn’t I be able to know that there is such data available within a reasonable distance of my search?

Now, lets move up to the town or local level with a site such as Localism.com (I’m not picking on them – I love them, am an active rain member, and am just illustrating my point with a site that actually has a fair amount of local data).  Localism.com lets you select a town and look for content. Let’s take the town where I grew up as an example: Birmingham, Michigan (and also make Maureen from the mioaklandcounty blog happy 😛 ). 

Demonstrating the Problem

Using the localism.com homepage, Birmingham, Michigan has 7 entries in the dropdown box related to Birmingham, MI once you type “birmingham”  but choose the town itself (“Birmingham, Michigan”) and no data shows up and no redirect occurs.  When I select “Birmingham Place In Birmingham, Michigan”, I at least get redirected to a page that indicates no data but has a map with a push-pin in Birmingham.  This may be due either to the beta nature of the tool or because the data elements do not know internally where they are in relation to other.  Additionally, no data appears for the surrounding towns of Troy, Beverly Hills, or Southfield.  If there is data in surrounding towns, I’d suggest that users should be at least made aware that the supporting content exists. Alternatively, wouldn’t it be nice to know where the nearest data is instead of having to search area by area?

Possible Solution

A possible solution would be to aggregate long tail content and wrap some basic geolocation data around each item. This gives the data some internal context in terms of location. Logic can then use the location data to be aware of where content is in relation to any other content. In this way, long tail hyper local data can be displayed within the same UI currently in use without having to rely on a separate search, mapping interface, a 3-5 click drill down, consolidated pushpins or any other external interface. 

Example

We are currently using a similar approach with our real estate video interviews (http://vidlisting.com/interviews.asp).  Interviews have a consolidated interface that they share with the with thumbnails of the 8 closest interviews (by proximity) to the real estate video interview being shown.  You can see thmbnails of the related interviews just below the real estate video intervew being played in the screenshot above.  All of our real estate video interviews share this interface. We’ve demonstrated that this approach works with both US and international data. they are not constrained by city, province, or national borders. This approach could easily be extended to all types of long tail data – the presentation options are limitless.  Users could even choose how far they want to include as a configuration option of the UI (1/2 mile, 1 mile, 5 miles, etc.)

Advantages

This approach brings relevant information directly to the user as part of an existing interface.  No additional user intervention, clickthroughs, or zooming required  🙂

Thoughts?

(We’ll discuss a second key issue with micro content in an upcoming article….)