Giraffic – Enabling Smooth Internet Video on Demand

by Sharon Goldschlager

(Note: See Sharon’s BIO on the Contributing Authors’ Page)

Every day, over 10,000 new users join the Giraffic Fully Distributed Video Acceleration Cloud, while Giraffic has proven its ability to save video publishers over 50% of their video bandwidth costs.

Introduction

One of the fastest growing trends in Internet use today - and expected to grow in the coming years - is the consumption of video content. For example, on YouTube alone, over 2 billion videos are viewed daily - and the popular online movie service Netflix now accounts for over 30% of total US internet bandwidth at peak hours. 

Giraffic, a technology company led by a team with a proven track record of innovation and commercial success, has the solution for both the cost and the video congestion challenges caused by the massive adoption of video. 

With a growth of over 10,000 new users joining Giraffic’s Fully Distributed Video Acceleration Cloud daily, Giraffic has proven its ability to save its clients over 50% of their bandwidth costs, delivering video content with a smooth playback and virtually eliminating the annoying re-buffering pauses.

Though costs of video delivery in 2010 have declined by 20-25%, traffic volume on average grew by 50% and is forecasted to explode.  

Consequently, the total cost for content providers is rising and will continue to grow; while Internet congestion is threatening end-user streaming experience to the extent that the EBU (European Broadcaster Union) has concluded that video could collapse the Internet.

Content providers use the services of CDN (Content Delivery Networks) and local ISPs (Internet Service Providers) to deliver their Video content. CDN is a dedicated network of dispersed servers holding data copies, serving end-users by dynamically and intelligently determining where to serve content from, based on expected latency, which depends on geographical proximity and other parameters.

The Video Delivery Market 

A recent Frost & Sullivan survey states that "The biggest trend affecting the market over the past 12-18 months is the emergence of broadcast quality content being available through over-the-top services. Subscription-based offerings from Netflix and Hulu Plus along with videos from most major broadcasters have created a tremendous amount of new traffic for the CDNs to deliver." 

The major CDN providers have their own networks scattered throughout the world, offering their service for a fee. The cost of maintaining this vast array of servers and cost of bandwidth is eventually inflicted on the content providers who hire their services.

Giraffic - The Company 

While working as head of sales and business development in one of the RAD group companies, Yoel Zanger was exposed to the enormous potential of video delivery via the Internet. In 2008, Giraffic was co-founded by startup veteran Zanger – CEO --- and physics and mathematics algorithm expert Dr. Gil Gat - CTO.  

Zanger, former executive at Phonetic Systems - a startup which was sold for $70M – has held a number of executive positions in leading technology firms. Dr. Gat gained his unmatched expertise in encoding, streaming protocols and storage algorithms while working for numerous high-tech companies and in academia.

Giraffic is backed by an experienced Board of Directors and advisors. These include some of the most accomplished high tech and media veterans - such as former CEO of Warner Music International, Former General Manager of CNN.com, and founder of Sling Media.

The Product - Giraffic’s Fully Distributed VoD Acceleration Technology

Giraffic stands out in its dramatic cost reduction efficiency and top video delivery quality by significantly reducing re-buffering pauses and increasing delivery speed, using proprietary algorithms.

Giraffic's technology guarantees high availability and smooth viewing experience for niche, “long-tail” content (the audience consumes a more diverse set of content), as well as highly popular blockbusters.

Based on the proven and reliable fully distributed cloud technology, where content is delivered from multiple cloud peers (other network members), Giraffic has developed a superior solution. What makes Giraffic’s solution unique are the encoded format of its data as well as the asynchronicity of its streaming protocol. This architecture fosters benefits for both the end-user and the content provider. 

Content provider benefits include:

• Higher quality video streaming increases end-user stickiness and reduces churn (over 80% of users who encounter re-buffering pauses abandon the video playback)

• Dramatically reduced cost – massive infrastructure savings, with almost no reliance on traditional CDN (Giraffic’s field results prove over 50% bottom line cost savings)

• Infinite scalability – as the number of video consumers grows, the network capacity grows in parallel, with unlimited size potential

• Seamless deployment and seamless fallback to classical CDN or origin server is fully automatic and guarantees that the end-user experience is never compromised

Benefits for the end-user include:

• Significantly improved streaming reliability and end-user experience - reducing re-buffering events by 50-70%

• Advanced VoD-focused capabilities - media acceleration, seek forward and adaptive streaming

• End-user video acceleration, optimized for HD!

• Virtually non-disruptive - only uses a small fraction of the end-user upload bandwidth, typically about 10%, yielding to other activities on the user’s device

The reader is encouraged to watch an online demonstration of the technology. After installation of the Giraffic Video Accelerator, select a demo. On subsequent playbacks, viewers are advised to disable local caching (link below play screen) and check the "Show Detailed Streaming Information" box to get statistics of the amount of data retrieved from P2P (cloud peers), server and cache. 

Note:  The demo’s statistics display refreshes itself constantly, so playback quality may be affected. This is for demonstration purposes only; when the detailed streaming information checkbox is turned off, video playback is smooth.

Business

Giraffic is focusing on delivery of Video-on-Demand (VoD) streaming content, having the highest profit growth potential. Giraffic's business model is based on splitting cost savings, insuring a mutual interest with the client, and on enabling creation of new monetization opportunities for online video publishers.

Customers include, among others, Veoh.com – one of the world’s largest video entertainment sites (with over 12M unique users and 40M video views per month); Craze Digital, which owns over 5,000 full length features with over two million views daily; AudioTube music , with integrations and trials with leading content networks. 

Giraffic offers solutions for three major market segments.

• Media Companies and Content Distributors

Giraffic’s platform supports any type of media – video, games, software updates or data transfer.

• Social Networks and Video Sharing Sites

Only the first or second view is sourced from a service provider’s infrastructure or CDN. All the following views are delivered exclusively from the Giraffic fully distributed cloud, at high quality and with no buffering.

• Device Manufactures

Display devices and recorders can pull streaming content from other devices connected in a network, holding copies of the desired content.

Lessons for the First-Time Entrepreneur

Zanger agreed to share his experience from when he had founded his first startup.

• Seek guidance and help from anyone who will give it to you, not just techies and investors.

• Recognize your own limitations - Assemble an advisory board of successful and experienced personas.

• Seek Market Validation - Dialoguing with clients in the very early stages is key for designing a successful value proposition and the right company business plan, but mostly – is a must in order to raise seed capital.

• Protect your IP, using patents or other ways, to a limit. Sometimes the risk of not sharing your technology with potential investors and customers is greater than the risk of disclosure.

• Setup a team of not only talented, but also extremely committed people (offer options), having 'chemistry' with one another.

• Build a demo or a prototype as a proof of concept, but don’t be fanatic about getting it – the business is much more important.

• Delegate administrative tasks, adjust the team, and be focused mainly on business results. Remember: efforts in a startup count for very little. Results are all that matters. 

• Don’t be afraid to share everything with your Board of Directors, and sometimes with all of your investors and shareholders. 

Attracting investment

• Investors want to invest in the right trends, not only the right technologies and teams. Choose a hot trend to ride on, like Zinga (FarmVille) did on the Facebook trend. 

• Find a few dozen customers to support and recommend your idea – market validation; in any case, investors are more interested in the business side than the technology.

• Approach investors one by one, improving your pitch on the way.

Vision 

Zanger's vision is to incorporate Giraffic's technology into multiple consumer electronic devices - PCs, set-top-boxes, game consoles, mobile tablets, etc. 

As far as the future is concerned, Giraffic has also identified the opportunity to utilize its technology in order to increase revenues for its customers by generating new monetization opportunities for them. 

Can INEZE Innovate on Product Search?

by Ryan Clarke

(Note: See Ryan’s Bio on the Contributing Authors' Page)

For many years now, search has been the way we Internet users find information on products we want. As consumers, there are many sites for finding specific information about products, or comparison shopping, before we spend our hard-earned money: Google Products, Amazon.com, Pricegrabber, Shopzilla, and Shopping.com are some of the popular ones. These sites do a great job of finding and comparing products based on price, but can they really understand what we are looking for without us having to sift through a sea of product data? Well, Gil Rozenberg, Co-Founder & CEO, and Koby Berkovich, Co-Founder & CTO, of INEZE (pronounced “een-eh-zay” means “here it is” in Hebrew), think it is possible and are building their vision of an intelligent shopping decision engine for consumers. 

Gil Rozenberg, Co-Founder and CEO of INEZE

Many of the top shopping comparison engines are built on the premise that consumers want to compare product prices first and foremost, then product data. As a result, if we do a sample search for “cell phones”, a sea of products related to cell phones is returned: including accessories. Then, we use a set of filters to help us sift through the search results till we find credible choices. After that, we can compare the product specs to choose the right one for us. For uninitiated consumers comparing these products based on this information is a daunting task. “The results from these sites lead to a paradox of choice…” says Gil, “…and there is a real opportunity to help consumers turn this information into knowledge.”

According to Rozenberg, many of these businesses are built on revenues based on conversions of users to buyers. Since most of these sites look similar, and offer the same functionality, there is very little to differentiate between them, so there is not much customer loyalty. This also means that they are also suffering from an erosion of their conversion rates (click-throughs) which, in its current state, means that they will become unsustainable as businesses in the near future.

INEZE’s approach is decidedly different. For one thing, they started out as just another player in the aggregator/comparison space. They innovated on the filters a bit and made a compelling offering, but not wholly differentiated. Sometime after winning Mini Seed Camp Tel Aviv in February of 2010, then Seed Camp London later that year, and getting lots of attention from potential investors, the team profoundly changed their approach. “We had offers for investment and I was asked if I’d be willing to move to Silicon Valley to start doing business development, sales and marketing - while Koby built a technical team in Israel. We thought long and hard about both offers and decided against them, as attractive as they were.” Gil related. “We believe in the Lean Startup methodology of building a product, getting it to users, collecting feedback, then iterating till we have something of value.” 

Koby Berkovich, Co-Founder and CTO of INEZE

What also drove the change to the interface was the team’s commitment to solving the problem of turning information into knowledge - for users to more easily make purchasing decisions about products. A commitment that meant turning down investor cash in a frothy market, living off savings and freelance jobs. A commitment that meant doing every job for the site except graphic design. A commitment that has led to INEZE’s new user experience which discard’s the earlier filter boxes of a typical aggregator and, instead, offers semantic search with curated results with suggestions for equally capable phones (INEZE searches return one “recommended” phone and up to 15 “runners up”). Another innovation is the graphic display of comparison results which show, through colored arrows, how each “runner up” product measures up to the “recommended” one (see image below). 

Early responses from users to the beta, launched in mid-April, are positive and for some it has become a discovery engine, an unexpected but welcome feature Gil assured me. So far, INEZE has risen organically in Google search results for some semantic queries and will probably continue to do so as the site is refined. 

Screen Capture of a Product Comparison on INEZE

In the meantime, there are still a few bugs to fix in the semantic search: a query for “touch screen phone” returns some non-touch screen models to the top - and some tweaking of their UI: graphic layout of product data. But the site continues to evolve daily as feedback from users is absorbed and responded to. As for a business model, Gil says the INEZE team is laser focused on building a compelling product and “being the best at one vertical, learning from user feedback, then building out each additional vertical.”  Next on the list: Cameras, TV’s, Tablets? That hasn’t been made public yet, but that’s not for fear of competition. INEZE’s team invites competition. They see it as validation that they are on the right path and that there is plenty of opportunity for everyone. They even cheered when a “competitor” tackling the same problem in a different vertical, was acquired by Google.

Since winning both Seed Camp events, INEZE has amassed a war chest of resources and free services which they have yet to deploy. These will definitely come in handy for the march up the hill to conquer the paradox of choice for ordinary consumers. Good hunting.