Thoughts on Streaming Media East 2015
My high level takeaway from Streaming Media East was that the time for talking and garage-hacks is past, and everyone's getting down to serious business now. As we saw at CES, it's clear that the world is getting more connected, and there are more businesses and more applications capitalizing on that connectivity. What was once novel and geeky is now mainstream and profitable. However, this creates tremendous infrastructural challenges, ranging from high-capacity networks (so-called ECDNs) within the firewall to high-efficiency CDNs and content delivery strategies over the open Internet. AsDan Rayburn and Mukul Krishna have been saying for the past two years, MSOs and telcos need to stop viewing video as a traffic problem and start seeing it as a business opportunity. We finally saw those lights begin to flicker on during this year's SME program. Increasingly CSPs across the globe are looking at ways where they can leverage an existing infrastructure and customer base to deliver video. One key reason for this is they want to see higher margins as their voice and in some cases data components are increasingly commoditized.
Video seems to be a major way in which they hope to help their margin situation. We are seeing two ways in which they are typically looking at leveraging their infrastructure: a) Launch a multi-screen video service themselves and/or b) provide an end-to-end managed service to enable content owners to monetize in a multi-screen environment with a focus on QoE and a branded, differentiated, persistent experience - instead of just throwing stuff on to YouTube.
In terms of new technologies, HTML 5, DASH and HEVC are top of mind. We continued our series of real-world business-centric talks on HEVC (slides available upon request and also posted to the SME site; video coming shortly). As predicted, several silicon-centric implementations (e.g. from Advantech and Qualcomm) are being introduced, but the vast majority of production encoders continue to be built on Intel processors today. This will shift over time - Imagine and Ericsson are betting on designs using their own custom silicon - but it is still a position of strength. In terms of AVC transcoding, Intel is gaining strength as software-defined workflows and virtualized high-density transcoders are accepted as best-practice architectural choices. We presented a paper commissioned by Vantrix showing that the total cost of ownership of JITT (powered by Intel) was nearly half that of JITP. Vendors like Imagine Communications, Wowza and Ericsson independently corroborated those findings in private conversations. The impact of DASH (now embraced by all four leading DRM vendors), and the fall of proprietary Widevine and SmoothStreaming formats, is poised to be hugely disruptive. We'll be discussing the implications of this in our upcoming DRM study.
Monetization in general, and advertising in particular, continue to be open issues. While there is growing maturity in ad-insertion and ad-optimization solutions, there is still limited ability to provide ad inventory (except for the live sports application, which is thriving on sold-out ad inventories). Very few companies are able to actually enable content providers to fill their ad carousels in any meaningful way; this remains perhaps the most critical unmet need for the streaming industry today.
Security is also becoming a more important consideration - not from the DRM front (which is well understood already) but throughout the IP-based workflow. More, including a recent thought leadership paper we wrote on the subject, is in my blog post here: http://www.frost.com/reg/blog-index.do#.VV0TtaUq5yU.twitter
In terms of workflows, we saw a number of talks and instructional sessions aimed at building out very large scale networks, storage facilities and streaming facilities. Most of these talks were aimed at broadcasters, news gathering services and telcos. Education was also a key vertical - digital instruction methods (as analyzed in our Lecture Capture Systems study) are pushing education to become arguably the fastest growing source of professionally created and curated video, with soaring consumption volumes. We are seeing growing need to unify video applications and workflows that have traditionally been siloed: live v/s VOD, inside-firewall v/s OTT, in-browser v/s app-based streaming, IP v/s traditional networks... these are all converging, and solutions are being enhanced to enable such convergence and unification. Branding and potential disintermediation are becoming more and more business-critical conversations, even as a land grab war for big data gets fervently underway. We are actually seeing a resurgence in DRM market fragmentation with vendors like Ericsson revitalizing their own DRM systems (Azuki in Ericsson's case) in the quest to capture last-inch usage data and parlay that into analytics, personalization and prediction. We saw discussions ranging from best practices in newsgathering, through scalable IP workflows, all the way through high-volume, highly scalable broadcast and personalized delivery, we are beginning to see streaming video come of age as a fundamental, business-central mode of operation for digital media companies. It is clearly past the chasm, and is on the upswing towards becoming the mainstay of digital media businesses. We find that in a small yet significant way, the traditional NAB show audience is beginning to look for business critical solutions and forward-looking technical instruction under the Streaming Media umbrella.
As the bottom line, a seismic shift driven by online video being streamed to consumer devices is well underway. Everyone - technology vendors, system integrators, content marketing teams, content service providers and consumer device vendors - will need to transform themselves from product-centric, delivery-focused businesses to monetization-enabling solutions. This SME we released, for the first time ever, our vendor positioning map for the M&E industry which categorizes various market participants into quadrants and calls out recommended industry best practices according to Frost and Sullivan. (slides available upon request and also posted to the SME site; video coming shortly). Interestingly, our audience straw poll in the nearly full hall showed that half our audience was operators or broadcasters and the other half was vendors. There were very few (if any) attendees from companies related to user-generated content or enterprise-only content, again emphasizing how crucial OTT strategies and offerings are to mainstream content businesses today.
As always, we're happy to discuss these findings or any other questions you may have. Just reach out via phone, email, linkedin or twitter.