Jeremy Allaire's Station

Pure SaaS Identity Management

Posted by: Jeremy Allaire, July 29, 2009

Somewhat unrelated to Brightcove, I wanted to quickly note an announcement yesterday from rapidly growing software company PingIdentity (disclosure:  I am a board member of PingID). 

The company announced a partnership with Google wherein Google Apps can now be used as a primary identity to authenticate into dozens of other SaaS applications, all through yet another SaaS application, PingConnect.

For decades, we've thought of our primary work identity as being tied to our corporate networks and PCs.  But as email, file sharing, collaboration, and nearly every facet of business automation and online publishing move into the cloud, it turns out that cloud/saas applications are better at hosting and managing many of these systems.  Identity management is next.

How does this work?

Basically, using industry standards such as SAML, PingConnect operates a sort of secure, hosted, identity-switch that connects an Identity Provider (such as Google Apps) with service providers (SaaS applications like Salesforce.com and SuccessFactors), who then rely on the identities authenticated by the identity provider.  All of this happens in the cloud.

If a new employee starts to work at a company using such a system, they are automatically provisioned and authenticated into all of the 'trusted applications'.  Likewise, if they leave, the company just shuts off their Google Apps account, and they can no longer access any of these external hosted applications.

It's cool and exciting to see PingID bring forward what was a very early vision for the company.  Congrats Andre and team.

Jeremy

 

Comments (0), Tags:

Brightcove 3.3 coolness

Posted by: Jeremy Allaire, June 27, 2009

About a week ago we shipped our latest update to Brightcove, BC 3.3.  I'm not going to go through everything in the update, but thought it would be good to highlight several things that I'm very excited about.

For full details on the release, you can check out this blog post from Emily Glass, or the detailed Release Notes here.

Distribution and Syndication 'stuff'

There are several new features that can work together to help drive unique distribution strategies using Brightcove.  We often find that our customers want to leverage their Brightcove accounts to manage distinct distribution and syndication relationships, whether that be a partner site that wants to integrate content, or a new iPhone application or other device connecting into their content library.

- Custom Fields.  We've added self-service tools that allow publishers to arbitrarily extend their content library data model to include additional meta-data and custom fields.  This has many powerful uses, including surfacing new related content inside of video players or video pages -- for example, a product video or music video could have related SKUs and URLs that are used for contextual commerce.  Custom fields can be used at run-time via Media APIs, Player APIs and via data-binding in BEML.  From a 'Distribution' perspective, custom fields are vital to mapping your content library to the diverse set of meta-data requirements that your distribution partners may have.  For example, the meta-data required by Yahoo is different than YouTube and still different than a Cable VOD system.  Using custom fields, you can store all of this unique meta-data and then use our Media APIs to output the right feed for the right partner.

- Custom Transcode Settings and Profiles.  With Brightcove 3, we introduced new cloud-based transcoding services as part of our platform.  These services will automatically generate different renditions of your content for different levels of user experiences and devices -- e.g. from iPhone up to HD h.264.  We've had a lot of feedback that our publishers want to establish their own unique transcoding settings and profiles, and so with BC 3.3. we've enabled this (though still requires working with Support, self-service UI will be forthcoming).  With this feature, publishers can upload a high-quality (3-6Mbps h264) master file, configure their transcode settings, and automatically generate a wide range of different versions of their content.  Using our Media APIs, a publisher can generate feeds or direct file access to these source files, which is useful for device distribution and syndication.

- Dual-Delivery Beta.  While not talked about in our release notes, we also have a beta program running with BC 3.3 for a new feature called 'Dual Delivery' -- dual-delivery makes it possible to make your content available through both secure, FMS-based streaming for your web properties and players, and at the same time available via PD/HTTP access through secure use of our Media APIs.  The idea is that many publishers want best-in-class quality and security for their web streaming, but also have a range of distribution scenarios for their content that requires more open and HTTP-based access.  If you are interested in this program, please contact support with your use cases, etc.

All of these features can work together to enable powerful distribution scenarios, and provide developers with a ton of flexibility in the kinds of video applications they are building.

There are some other cool new user experience and productivity enhancing things in BC 3.3, including Custom Player Skins.  Basicallly, every graphical element in Brightcove templates now can be skinned using JPG, PNG and SWF assets without requiring any programming.  A big part of what we've tried to do is to provide a platform that allowed for deep customization along a spectrum of skill sets, from basically colors and graphics, to UI component skinning, to custom template authoring using an HTML/CSS level skill-set, to full-on Flash or AJAX UI programming against our APIs.  Custom Skins are another step along the way to providing this rich customization.

In the advertising arena, we added a number of great new things.  First, we added a new visual cue point editor for managing mid-roll advertising -- whether it be overlays used in short-form video or ad breaks used in long-form content.  Ad Ops staff can now easily manage the insertion of these types and formats of advertising.  We also introduced a new Ad Policy API, essentially opening up our Ad Policy run-time to developer and 3rd party innovation.  This will be valuable for custom business logic, but also for plugging in flexible ad serving and inventory systems such as Freewheel.  Finally, we introduced the industry's first support for VAST, a new IAB standard that we helped to author which aims to simplify and streamline how video players and experiences interact with ad servers used by publishers and advertisers.  We're trying to make as much of these protocols open and inter-operable vs. the constant hacking that people have had to do to make all of this video advertising stuff work well.

There's a lot of other nice stuff in BC 3.3, including a range of back-end infrastructure that you can't see but that helps to drive quality, scale and reliability in our service, and bunches of other features.

We have a lot of product work underway for 2H 2009 and will be excited to share as we introduce.  We're also continuing to hire into our product and R&D organizations, so if you know truly great software people, let us know!

Jeremy

 

Comments (0), Tags:

Newspaper industry and online video growth -- any hope?

Posted by: Jeremy Allaire, April 7, 2009

Earlier today we posted a brief piece that talked about our aggregated analysis of growth in online video from the US newspaper industry.  Overall, the data are encouraging across a number of vectors including the volume of content published, the ways in which video is being integrated into the editorial and user experience mix, and the growth in video views.  This is an encouraging glimpse of hope in a business that is struggling in a many other areas.


Business Insider had a brief commentary based on our post, where the essentially message from them was -- it doesn't matter, it's too late, they're all going down in flames (that's my paraphrase).

It's hard to argue that the core local, regional or even national newspaper brands are going to remotely resemble what they have in the past.  Clearly, digital media broadly (and video specifically) are critical to the future of local, regional and national media, and what I would expect we would continue to see are companies who restructure consistently and shed weaker assets and components of the business while re-enforcing and building those parts that will actually grow.

Jeremy

 

Comments (2), Tags:

3D video navigation

Posted by: Jeremy Allaire, March 21, 2009

A fun example of FP10 and BC3 player APIs to create a 3D video navigation and consumption experience.


 

Comments (0), Tags:

The first 'Internet Phone' circa 1996

Posted by: Jeremy Allaire, March 16, 2009

While I was organizing my office recently, i stumbled across an important modern Internet industry relic -- the AT&T PocketNet Phone.  There is a great story behind this device and my explorations with it, and it's fun to think back to this just as the mass market is finally embracing the mobile Web through the iPhone, Android and the like.

First, here are a picture of the PocketNet phone.

WebPhone

This phone is actually THE FIRST web-enabled mobile phone in the world.  It is actually a prototype phone, it was never shipped to the general market.  It was designed in collaboration with AT&T and Unwired Planet, who invented the mobile web with a simple text markup language and browser for low-end mobile IP devices.  UP's Handheld Device Markup Language (HDML) later became Wireless Application Protocol (WAP), which is now ubiquitous in mobile devices around the world.

Back in the fall of 1996, we were busy getting ready to ship ColdFusion 2.0.  I was really passionate about the idea that the Internet as an application platform would mean that you could write applications for any device that spoke Internet formats and protocols.  I was convinced that the proliferation of IP-connected devices was soon among us and wanted to showcase how new server-based runtime platforms could support this vision.  I stumbled across Unwired Planet through some news article, went and met the founders, we hit it off, and soon thereafter ColdFusion was the first commercial web application server to natively support HDML application authoring and delivery.  You can see our announcement from back then here.

In addition to creating some sample apps that would showcase using CF and HDML doing database app stuff, I wanted to explore the 'converged device' vision a little further, so in a bit of uber-geekdom I created a home automation application application that tied together a mobile web app controllable via the PocketNet device, a ColdFusion server app in my home, connected to an X-10 controller system that could automate devices in my home over the electrical network in my house.  I convinced a developer in Florida to build an X10 'custom tag' extension for ColdFusion (you can still find the extension on the Adobe site), picked up some X10 gear at RadioShack, and connected it to lighting and my coffee maker.

From my mobile phone, i was able to pull up from anywhere in the world whether the lights were on or off and switch them from my phone.  For shits and giggles, i also connected it to the coffee maker just so that I could say that I had brewed coffee from my phone from across town.

It's 13 years later now, and the landscape of the mobile web and the converged Internet is really truly here, but I still feel like we're just getting started in many respects.  It's been humbling and awe inspiring to see the incredibly range of innovation coming out of new open mobile platforms such as the iPhone.  Just for comparisons sake, here a snapshot of the latest mobile video browser build for the iPhone using the Brightcove Media APIs.  

 

Comments (0), Tags:

Pick Your Own Ad Proposal

Posted by: Jeremy Allaire, February 28, 2009

Bill Gurley has a short post on his idea for online video ads -- let the user decide what commercial to watch.  My comments have less to do with the proposal, and more with some of his assertions in general about monetization fundamentals for premium online video.


First, the idea itself is a good one, and one that we POC'd as a new ad format in Brightcove almost 3 years ago.  Hulu has already adopted a variant of this -- chose to watch a few ads during your show or one long one up front, and I believe also a format with choosing which type of car ad you want to watch.  This kind of end-user empowerment is great for everyone, as Bill noted.

However, the real challenge for this kind of ad format TODAY is that the vast majority of online video publishers simply don't have enough scale to make it viable.  With the exception of Hulu and a few others, any given publisher's online video inventory simply does not offer up enough usage volume to make it viable to have multiple advertisers to choose from.  This is especially the case in a world of publisher-controlled ad sales (vs. networked sales), where they could easily sell-out all of their inventory with one or two publishers.  As a result, in fact, today we still face the "creative fatigue" issue with online video ads -- not enough inventory = poor frequency capping = crappy user experience.  

To make this concept viable today, more publishers would need to elect to use 3rd party sales network whom in aggregate having enough diverse inventory to meet the scale requirements for this level of consumer choice.  There are also challenges on the buy-side with this, with TV/video media buyers in general looking for certainty in their buys, and not willing or interested to buy video on a performance basis, which in some respects this 'choose your own ad' approach would require.

Anyway, I do love the idea, and I do think it will happen over the long-term.

However, I also wanted to challenge or at least engage with Bill against some of his other "core assumptions" on monetization in premium video on the Web.

"1) No one will ever monetize commodity content well.  If the same video is on YouTube, Veoh, and Metacafe, you won’t have the “right” to ask the consumer to wait for an advertisement.  Only by”controlling” unique/premium content can you ask the user to participate with advertisement."


I don't think it is true that no one will ever monetize commodity content well (i guess in part it depends on how you define 'well').  Certainly in the context of largely UGC and 'Torso' based experiences such as YT, Veoh and Metacafe, the advertising community has clearly devalued inventory given risks/context, but in 'premium' video contexts (MSN, AOL, Yahoo, Hulu, etc.) even commodity content is monetized reasonably well -- many of them carry CNN feeds (arguably a commodity), or AP feeds, and these still garner high CPMs for video ads.

2) Some networks, like ABC, are already doing a great job with online video advertisement.  I have heard numbers as high as 80% of the revenue per viewer hour compared with regular network television.  That said, I think the model below will dramatically enhance this number.


No argument here, and have heard the same numbers.  Clearly, the combination of premium content, premium user experience, and engagement-based selling proposition through interactive Flash ads has differentiated this product.  It is totally shocking to me that more publishers have not used interactive Flash take-over interstitial ads given the clear superiority of the format and its ability to drive higher CPMs.  We've had it in our product since late 2005.


3) Having multiple sales forces selling the same premium video advertisement is counter productive. It drives down pricing.


Not much of an argument here, but a couple of thoughts.  Commodities that are distributed through multiple premium outlets and sold by multiple sales forces are OKAY, since for the most part the media buyer is buying a defined audience, genre or segment from the publisher, vs. specifically buying that content (e.g. They're buying 'News' on AOL Video, not CNN, etc., and AOL is providing a distinct demographic product as well).  But the fact that these are commodities does have an impact on pricing, without a doubt.  More importantly, though, I think this is one of the very strong reasons why more and more premium content owners are engaging in syndication and distribution relationships with major sites where they maintain control over advertising sales.  Until the Big 4 (AOL, Yahoo, MSN and MySpace) gave in to the networks (Hulu and CBS) on control of advertising, things were pretty broken, and one of the key reasons was that it is/was complicated to have '24' sold by 5 sales forces on 5+ different sites.

4) “Distribution” is a confused word online.  Everything can be one-click away.  As such, there is no real reason to have someone “distribute” your video in the old classic sense - especially if it is important for you to control the advertising.


On this is I largely disagree.  Reality is that most online video consumption is serendipitous, not intentional, and so the paradigm of quickly searching and finding the program you want from the primary destination site doesn't necessarily hold water.  Furthermore, major 'distributors' (e.g. portals and large video sites) can bring unique marketing against specific content, and are a home to an audience that are often repeat visitors to those sites.  On the flip side, the 'chromeless player syndication' model that we and others have been promoting opens up a great approach where the distributors maintains control of the user experience and site/brand integration, the content owner maintains control of the programming, advertising and analytics, but bears the operating expenses for their online video and rev shares back to the distributor.  The number of premium brands that we work with who are pursuing this strategy continues to grow, largely because these publishers are in search of more and more inventory.  This isn't going away, it is really only getting started.

Jeremy



 

Comments (1), Tags:

5 Key Costs of an Online Video Business

Posted by: Jeremy Allaire, February 20, 2009

5 Key Costs of an Online Video Business

An earlier post generated some comments from a reader, where he looked to summarize the 5 key operating costs and considerations for an online business.  It’s a good summary of my points, and here are some additional points of clarification. 

Preetam’s comments are italicized.

+ sales agents and sales costs (could you shed some more light on this?)

Online video advertising sales can have many component parts, all of which impact the ultimate net revenue to a publisher.  If you are able to do direct sales of your inventory to a media buyer (which means you likely have 1M+ streams / month, since only every other stream is monetizable, that’s 500,000 streams /month), you are likely deal with CPMs that are $20-40.  Let’s say $25, which is the average for premium inventory.  The $25 is the Gross CPM, which is not the true CPM.  Direct sales to a media buyer is actually direct sales to an ad buying agency who is buying on behalf of their client.  They are typically taking a % of the sale as their agency fee, often 5-10%, so your actual Gross CPM might be $22.5.  Additionally, you are likely employing your own high-cost ad sales team that are selling your advertising, and they are paid on a commission basis, typically 10% of gross so you are now down to sub $20 effective CPM.

If you don’t have high-quality premium inventory and brands, and don’t have enough scale to sell directly, you are employing an ad sales network.  They are selling directly to media buyers and have their own full time staff of ad sales professionals, as well as the broader operating costs for their business.  A top-tier video ad network can generate a gross CPM of around $20, but average a good bit lower.  They will in turn negotiate a deal with you that will range between 40 and 60% of their net CPM, so you are likely looking at $8-10 best case, $1-3 worst case.

Also, if you have a sufficiently compelling content product, and a well defined and desirable demographic for the content, your best bet is likely to try and sell an integrated sponsorship that includes product placement and digital entitlements, as in a pure sponsorship it is less CPM driven and more ‘value-based’, which can mean that you can effectively garner a higher rate / value for your property.  When you have very limited inventory, this is probably one of the only ways to gain advertising dollars in online video.


+ marketing, advertising, pr (seo, ppc, affiliate, ad networks, social networks, syndication, one-click publishing to top eyeball destinations, etc.)


While these are clearly the right list of online marketing tactics, my point about it still coming down to great marketing has less to do with employing tactics and more to the more ephemeral dimensions of great marketing and branding, which distinguish the great products, from the good, from the poor.  Great marketing execution can make an enormous difference for any content property, but at the end of the day, also, if the content product is compelling, it will generate word of mouth promotion.

+ cost of content creation (- variable, obviously - )

And these costs can vary GREATLY based on what you are trying to produce.  If it is news and entertainment, production values matter, and quickly you are getting into the thousands or tens of thousands of dollars per minute to produce.  While there are examples of lower budget (thousands per minute is considered ‘low-budget’ in entertainment), you can see how pure play production for the Web is a loosing proposition right now.  Companies that have strong assets they can re-use, or where they have very strong existing online brands and audiences stand the best chance for building profits at this stage.

On top of raw costs of production and post-production, there are also typically copyright licenses one needs to acquire, including music performance rights, and any other copyrighted material you are using in your production.


+ technology and service operating costs (video platform, cdn, storage, scaling, enhancements)


I pretty much group all of this into the bucket of ‘Online Video Platform’, at least in the sense that Brightcove offers.  Our goal has always been to provide all of the technology and service operations needed to run an online video business, inclusive of the list you have and more.  Fortunately, it is quite cost effective to license access to such a platform with very reasonable annual and variable costs, and gain access to years of R&D and a continuous innovation cycle on new software and platform services.  Hopefully, too, platforms such as ours can also help facilitate ad sales, operations, and ad network integration, online marketing tactics, and with overall efficient service / site operations. 

+ service operations(uploading, asset & community management, administration, metatagging content, etc..is that what you meant?)

Yes, basically, the human labor involved in running a content and community based website.  Both the visual and web design, as well as the content management, editorial and community management needed for a decent sized property.  This can range for a couple of people to dozens of people.  Depending on the labor pool, demographics and geographic location, this cost can be in a pretty wide range as well.

Jeremy

 

 

Comments (0), Tags:

Developer action for Brightcove -- new SDKs for PHP and JavaScript

Posted by: Jeremy Allaire, February 20, 2009

Some great activity happening in our developer community, with new open SDKs released for Brightcove's APIs for PHP and JavaScript developers.  It's exciting to see online video moving into being a real set of developer platforms that web developers can use for integrating video and rich media into their sites in easy ways.


These SDKs cover everything from ingesting and transcoding, to meta-data management, to searching, filtering and displaying content, to managing content security, scheduling and other programming rules, to automating discovery of video through popularity, recency and relevance.  Powerful stuff that we hope a lot of web developers jump into.  These complement our Player APIs.

PHP and JavaScript BrightcoveSDKs are now live. http://echove.net and http://kudos-js.com

Jeremy

 

Comments (0), Tags:

Dan Rayburn - Online Video is a cost center....maybe....

Posted by: Jeremy Allaire, February 19, 2009

Dan Rayburn from Streaming Media and the Business of Video has a new post that attempts to put into perspective the reality of the online video business for most companies.


"for the vast majority of content owners, online video is a line item cost, not a revenue generator"


While it is absolutely true that many companies who are building ad-supported online video businesses are doing it at a loss today, it is not entirely true, and also, I believe, requires a more nuanced understanding of how media organizations and non-media organizations evaluate their online investments.

First, explicitly for media companies, we are absolutely seeing many companies who are generating positive contribution margins from online video.  These tend to be premium brands who have meaningful aggregated traffic and who have strong existing advertiser relationships because of their existing online and offline ad sales.  These range from music companies, to news companies, to magazine companies to TV networks.  We have seen CPMs remain stable, though down some walking into 2009, but for most of these the cost of production is either low or zero (re-used content), and the continued commoditization in CDN costs combined with the ease of launching these initiatives using online video platform services, has made it reasonably cost-effective.

For 'Torso' or 'Long-Tail' publishers of video, it is an entirely different story, and something that i have recently written about here.

However, I believe whether media companies are generating a net profit from video streams or not misses a larger point and context, which is that most media companies don't think about whether 'Video' is profitable or not, they are looking at the overall value of their online media properties.  Media companies operate their online media properties as holistic P&Ls, and are looking for overall profitability and growth from those overall businesses.  Over the past few years, most media companies have taken approach of investing ahead of revenue -- ie. deficit financing the growth of their properties through increased investment in content/editorial, new end-user features, new technology, etc. -- understanding that like any new media property that there is an investment phase of x$ and y years to get a media business to a certain scale and profitability.  There are well understood financial assumptions about this in any media market -- starting a news paper, a TV network, a cable network, a radio station, a magazine -- all require several years of deficit financing until they reach sufficient scale to extract either repeatable ad sales or subscription dollars.

The salient point here is that to be competitive as an online media property in the broadband era you need to have compelling video content -- it is a cost of doing business.  Having video keeps users in your sites longer, and creates additional opportunities for re-circulating users in a site.  Video has very high pass-along rates relative to text content, so is a strong organic traffic driver.  Video also happens to be the best medium to support brand advertising with the highest CPMs, so it can also be profitable (again, depending on some of the criteria above).  

Not every web publisher uses video as effectively as the next -- we see huge variances in the quality of execution with video, ranging from how it is integrated into a site, the kinds of content/editorial focus, their application of video SEO techniques, and the advertising execution with video.  You can find one publisher with identical content (e.g. licensed AP or Reuters video feeds) who is failing to monetize and drive usage, and the identical content with another publisher who is surfacing it in creative ways and executing well on ad sales.

The final point I'd like to make is that for non-media companies, video is, of course, by definition a cost center, typically a cost of marketing, sales or education/training.  But increasingly, organizations and corporations that are embracing online video for marketing and communications will become more focused on the ROI of the use of video, especially in the current economic climate.

Thanks,
Jeremy



 

Comments (0), Tags:

BEML + Flex Search Widget for Brightcove

Posted by: Jeremy Allaire, February 18, 2009

Great article/overview of how to create custom video search experiences using Brightcove APIs and BEML + Flex.  Includes source code and demo.  


 

Comments (0), Tags: