Improved image Handling: Custom and Insanely Custom Widgets

Part of the work that we’ve been doing for the next platform release is standardizing how we deal with widgets that have been customized through the custom widgets menu and what we internally call “insanely custom widgets” which have very unique looks usually driven by branding requirements.  As shown below, many of these widgets won’t even have to be exactly square in design.

Another example unrelated to the above image: In September, we’ll be running a trial with a good sized wireless carrier in Latin America that has a user base in the few million user range. Not huge but likely “right sized” for our current server infrastructure.   The trial runs for two weeks to 10% of their user base.  If successful, the trial will morph into a full set of revenue producing services available to the entire user base.   The widget to be used for this test is essentially a reproduction of  the carrier’s brand – there is nothing similar to it in our existing widget library.   That said, the code for this wireless carrier’s widget and,  more importantly, the dynamically produced images of widgets that we use in Facebook and other places are now handled through the same new code as our regular and more conventionally custom widgets.

The bulk of this change will be invisible to most of you but was a huge leap forward for us.  It improves your experience as a user by providing new out-of-the-normal widget designs and reduces the possibility for errors since we only have a single set of code to maintain.  It also opens the door to a whole slew of widget customization opportunities that we hadn’t been able to do before through a web interface. We previously were having to manually build exceptions for every insanely custom widget that we built.  That is no longer the case.

Where Have We Been and What’s Next?

With the decline of the worldwide real estate market, we’ve been focused on developing our content management, mapping, and mobile capabilities for smaller cable companies and cell phone carriers in other parts of the world. We unwound all of our real estate related functionality from the core platform in the process.  We would not still be around had it not been for our services provided to these other industries.

That said, our next release is very real estate focused with the added benefit of our enterprise experience in other non-real estate markets.  The real estate functionality is now a module or add-on to the core platform that can be seamlessly added or removed for specific organizations and individuals with a single configuration setting within the platform.

Real estate technology is the perfect showcase for answering the question of “what’s of value and what’s around it?”.  Even if you aren’t involved in the real estate industry, you’ll still find all sorts of new functionality related to mapping, content management, and mobile capabilities.  The next release is just around the corner…

Improved Image Handling Across The Platform In Next Release

Our next platform release is just around the corner and one of the specific areas we’ve put a lot of work into is image handling.  There’s likely a whole set of readers that are letting out a sigh of relief…we’ve heard your feedback and really focused on getting image handling correct in this next release (July 2010).

from feedback we received, we’d done well in some image handling areas especially those that involved video previews upon upload.  As we sat down and really had some open discussion internally and with enterprise customers, we need that we needed to do better.

The first focus area was the automation around knowing exactly what type of image had been uploaded. While we likely have a state of the art implementation for automated detection on the video side, our image detection system wasn’t nearly as comprehensive or reliable.  We’ve made tangible improvements for the next release.

The second focus area was making ensuring that every image was efficiently sized and shaped for all potential uses throughout the platform and on partner sites.

Lastly, we spent a lot of time focusing on how members store, share, group, and syndicate  all of the available content within the platform but especially images.  I’ll cover the specific functionality and platform features in more detail as we get closer to release.  There are a number of surprises here that I cannot wait to share.

We’ve also done a lot of image specific work within the developer API.  Members with access to the dev API will receive a separate document detailing all of the changes just prior to the next release going live.  We done our best to incorporate as many improvements as possible into the existing API set so you *may* already be seeing benefits from the code changes for the next version.

This is just a taste, I know…but we still have two weeks or so until launch. Until then, know that we’ve heard you….

Video Processing Changes

Today, we are putting a number of logic changes in place to prepare for our next version of the platform and to provide better service for platform customers.

As background to better understand what we are changing, we have a small cluster of processing servers that convert videos, resize photos, and do all of tasks like create documents on the fly.  Currently, when a video is uploaded, it’s sent to the first open server. If there are no open processing servers, the requests  waits in line for an open server and then the request is processed. The first request is the first one processed.  Like most sites, we have more unregistered guests than members and more free members than paying members.  We are taking steps to make our platform “smarter” to improve our service for our members.

Processing Priority: We’ve introduced the ability for prioritizing processing requests based on membership status or affiliation based membership such as a company or organization.  During periods of high activity, paying members will be prioritized over members who in turn will have priority over unregistered guests. Note that this change will have no impact when there are plenty of open servers. It will only kick in when requests are waiting in line.  In order to ensure that even low priority requests don’t have abnormal waits, we have also added a way of gradually raising a request’s priority over time automatically.

Parallel Processing: When a member with privileges to process multiple files at once sends the files, they’ll be processed on different machines at the same time.  Of course, this will depend on the member’s processing priority and open server availability.  Our old logic reserves a single server for them and they were processed one after the other.

Members using this functionality should see marked improvements in the amount of time needed to wait on a group of large videos to be converted.

Customized Processing Options: Certain membership levels will have highly customizable processing options.  We’ll cover this in detail upon launch of the next platform version.

If video processing is a bit wonky today, it’s because we are putting these changes in place in the live environment.

These changes will result in reduced waiting and far more flexibility for our individual and corporate customers.  They’ll apply across the board to all of the different types of platform customers (real estate, mobile services, etc.)  This should lay the foundation for our continued growth in 2010 and beyond.

FSBL Content Platform 2010

FSBL Platform 2010 is finally getting close.  We finalized the overall platform feature list as well as the free demo behavior  yesterday.  This means that we know what the target is for release now after the new year and specifically what that release will look like.  Innovation and ease of use are the watchwords for this release – it’s really going to be special.

Exciting times…

We’ve been busy…

You haven’t heard much from us lately.  We’ve always been out on the forward edge of rich media functionality doing things that now are “cool” before they were cool.  We’re continuing that tradition.

A few months ago, we rolled out some experimental functionality to several non-real estate enterprise customers.  The response was awesome enough that we have been “heads-down” to roll out a production version of it with the changes that they came back with.  It’s a set of deep changes for us so we want to be sure that we have all of the bases covered and the functionality fully tested before rolling it out to other companies.

Note that some of the work that we are doing has occasionally impacted our live environment for reasons that we’ll discuss when we launch in December.  Though we haven’t seen any new impacts, we apologize in advance for any issues. 

The new functionality that we will be releasing aren’t small changes or simply putting a new shine on an old product…it will totally shift how you think about and deliver rich content to your customers via your web sites. See you soon….

Restating Our Vision: What We Are About

We’ve taken a different approach to syndicating online content than most vendors. Our method brings the viewer back to the broker’s site to consume content rather than our site(s) or a third party site. It can best be described as syndicating the opportunity and brand rather than the content. The content is consumed on your website where the chances for customer capture are most rewarding for you.

Incidentally, Google will also index the syndicated content on your site rather than ours which is quite a different approach than most vendors.

You’ll be seeing a lot more about this theme soon.

Broadening The Set Of Enterprise Content Platform Customers

In 2006, when we were just starting this project, we knew that a niche would be required to be able to successfully monetize our then video-only platform. We knew that the flash video player would eventually become commoditized just as it did and focused on platform capabilities rather than a feature rich player.  Turned out to be a smart move on our part – we’d otherwise be right where almost all of the other youtube clones are right now.

Because we were (and still are) a bootstrapped company, our approach was fairly simple: build out a section of our platform, get paying customers, rinse, and repeat. We made the decision to focus in real estate because it was and still is a great niche to demonstrate our platform capabilities with geolocation, content management, and language support.  Let’s face it, we were video before video was cool, local before local was cool, and international before international was cool.

That said, we always knew that the platform had broad applicability outside of real estate.  We are starting to market to those non-real estate markets and beginning to find success. We’re entering into a series of technical discussion with multinational players around a number of applications for our platform.  We’ll still have a team 100% focused on the real estate specific platform features but are also now hiring for broader expertise and executing against a plan well beyond real estate.

As a result, this blog will continue to focus on the broader aspects of the platform.

When we said that we would be an enterprise solution, we meant it. If our current technical discussions go well, we anticipate seeing rollouts of non-real estate versions of our platform within the next 2-3 months.  Stay tuned – we’ll release details as we are allowed.

Month Of Code – Adding A New Feature Or Major Fix Every Work Day In April

The Month of Code for our team is now is full swing.  What is the Month of Code?  It’s an attempt to release a fully tested new feature or major fix every working day during the month of April 2009.

The Month of Code concept came about as a result of a simple question – “what features do we have just laying around that aren’t live?”  because we are bootstrapped without external investors, we’ve sometimes get sidetracked by large projects that keep the lights on.  This means that development on a certain feature begins, gets interrupted for a period of time, and then isnt a priority when we get back to feature development – the iPhone conversion feature is a great example.

So we’ve decided to do our best to see if we can empty our library of unreleased changes.  Here is our progress so far:

—————————–

1 April – significantly boosted our platform’s video conversion quality without losing the instant start and buffer free video that we are famous for

2 April -Added capability for members to associate videos by language so alternative language version appear in the player

3 April –  Integrated profile videos with certain types of content (example is from our real estate video portal)

4 April – Saturday

5 April – Sunday

6 April – released by popul;ar demand Unbranded Video for MLS and the ability to group Community/Neighborhood videos (real estate specific fixes)

7 April – added iPhone friendly video conversion for most types of video uploads

8 April – added FLV processing for videos uploaded as video podcasts

9 April– Improved the handling of related videos that are shown underneath the main video

10 April– Rewrote the interface and logic for creating custom branded embeddable video widgets far more simple

11 April – Saturday

12 April – Sunday

13 April – Added the ability to Syndicate  Documents/Youtube/Real Estate Show Content From A Folder to a tab underneath the video that is playing

14 April – Made a bunch of virtual folder upgrades and externation syndication changes for partner sites  (added a video demonstration for virtual folders)

15 April – Made updated new feature and system information available in member area

16 April – Added Geo targeted local market reports and hyperlocal blog posts available in member area

17 April – Made it possible to Syndicate Member Blog Posts To Videos/Filter Blog Posts by Tag

18 April – Saturday

19 April – Sunday

20 April – made a member’s shared video content available on the member’s public profile page (didnt quite get all of the bugs out of the functionality for video podcasting that we hoped to release today)

——————————-

I’ll keep adding to the list from time to time as we go through the month.  Taking the platform route was a choice that cost us early on but now we feel that we have “crossed the hump” and can develop new features and implement them on customer sites rapidly.  Thanks for sharing the ride with us.

Improvements To Set The Stage For Better Location And Enterprise Services

We made a lot of interface and under the hood changes this past week.  The changes were for two reasons:

1) to improve the location based video, document, and external link services that the online content management provides.

Back story: we started building the bulk of our geolocation services in 2007.  It was initially slow and clumsy and limited to just video but we knew that it would be the future path to differentiating our video services and providing a set of valuable services to our customers. We kept geolocation in the middle of our priority list as we finished the huge work items that we had.  In short, we worked on different parts of location based services at different times and added new content types for these types of services throughout 2008.

The result was a system to geolocate properties, documents, images, and members  that worked well but didnt have a lot of standardization on the back end.  With most of this phases big work items out of the way now, we spent this past week updating and standardizing how geolocation across the entire platform so that we can really focus on a new set of local video and content services for both users and application developers of web and mobile applications.

2) To tee up broader organizational focused functionality associated with corporate memberships

We believe that enterprises should be able to pass unique organizational branding and syndication to their staff, members, and partners to provide a seamless experience for their users. We’ve been selling memberships to selected real estate, web portals, and other organizations for over a year now that allow these orgs to do just that. We haven’t been offering membership levels on our website in an automated way because much of what we have been offering involved a very hands-on, boutique style of professional services approach. Now that our first phase of functionality is built out, we are specifically focusing on automated set of value-add online services that we can help organizations, portals, and associations provide to their employees, members, and internal business online. In addition to an automated method to sign up for various membership levels (including corporate), you’ll also be seeing a whole set of functionality aimed at the enterprise customer in the coming weeks.

Imagine being able to provide different types of content branding and syndication for different business units, teams, or levels within your organization. Content can be accessed, shared among the organization, and syndicated externally to others without multiple uploads, cutting and pasting of embed codes, or constantly moving content into “networks”. Content can uploaded that remains private…as in “no one can see it but you”.  Your organization doesn’t have to be limited anymore to public facing channels.

We realized that we had a number of dependencies on the back end of the platform that were fine for individual members but not as flexible for broad organizational use as we think that organizations need. Those dependencies have since been cleaned up this past week.

The result is that we are now ready to start offering a broad set of content management and branding functionality that allows a member to be part of any number of organizations. apply that organization’s branding and use any unique sharing/syndication associated with that organization when it is needed, and still retain the capability to use their individual membership to the fullest.  Multimedia and content developers will also find the new capabilities appearing in both the member set of the platform developer application programming interfaces as well as the full set.