Hello Jet-Stream.com!

StreamZilla is dead, long live Jet-Stream

Yes, we made the decision to kill the StreamZilla brand. No not the successful service. Keeping two brands alive is quite demanding, so we decided to offer all our services under one brand: Jet-Stream. Check our new services on Jet-Stream.com. I hope you like our new website!

This blog describes how we pivoted (startup buzzword bingo 😉 a few times:

Streaming projects
When I started Jet-Stream in 2003, our focus was on offering bespoke streaming platforms for telecom operators in the BeNeLux. We designed and implemented private CDNs and encoding farms for mobile operators, cable operators and internet providers. This was awesome, since they had zero streaming knowledge and we could help them sell faster broadband. Thanks to our streaming platforms, ISPs could bundle live TV services to the fastest cable and DSL services, so you could enjoy television on your PC and mobile phones. We also built VOD streaming platforms for these operators. Unfortunately, back then, CDNs and streaming were not a priority (operators were too focussed on triple play), so the telcos lost ground to new portals such as YouTube and to global CDNs such as Akamai who did understand where the future was heading.

From projects to services
I knew Jet-Stream could be in danger if the telcos and ISPs did not really invest, so I needed a pivot. Less projects, more services: more scalable. Luckily I met people from publishing companies: they wanted to offer videos and live streams to their audiences, but they did not have any infrastructure of their own, and their existing hosting providers had zero streaming knowledge. They asked me to offer hosted streaming services. That is how StreamZilla was born: a hosted and managed streaming platform. Under the Jet-Stream brand we kept offering bespoke streaming platforms to the telecom (telco, cable, ISP, hosters) industry, while StreamZilla was focussed on offering services (today called cloud or SaaS) to the content (publishers, broadcasters, enterprises, events) industry.

Growth
StreamZilla quickly became very popular. We were the first streaming platform that offered unified support for Real, Windows Media and QuickTime streaming, live and VOD. We also offered unified analytics and traffic reporting. In the following years we added progressive download, 3G streaming support, MP4 streaming, and we introduced Flash VOD trick play which we dubbed ‘Stroboscope’ (you know, Flash, Flash, Flash!). We expanded into Europe. Thanks to StreamZilla, the company grew over 1800% in just 5 years, and we entered the Deloitte Fast 50 and Fast 500 twice for fastest growing tech companies.

Software development
With the revenues, we started to develop our own software. We automated all our customers needs, and our own as well. Our ‘VideoExchange’ CDN software was a small revolution, primarily for three reasons:

  • Software defined: our software could run on common servers, rendering hardware streaming appliances, load balancers, storage servers, etc obsolete.
  • Automation: we slashed operational costs by automating all the processes in a CDN: provisioning, customer workflow and billing for instance.
  • Intelligent CDN management: by splitting control and delivery platforms, and by introducing smart popularity algorithms and delivery tiers, we could scale and architect very low cost, high performance CDNs, for any situation.

Licensing
When we released our new StreamZilla service, powered by our own software, we again doubled the number of customers in one year, while reducing the operational load by 50%. Suddenly, telecom operators knocked on our door. They wanted to license our software and build their own content delivery networks. I wrote a white paper about CDN strategies and shared it on a number of events. I also wrote a technical white paper (took me almost 8 months to write…) and that helped entering CDN licensing deals with some the largest telecom vendors in the world. I recall presenting in front of an audience (a room half filled with people from telcos, the other half were hardware vendors) and all the vendors went really silent when I told the audience that the appliance was dead. That was actually the title of my presentation, haha, that was really fun to do.

Crazy growth
Things got really crazy when I pitched for one of the largest telecom operators in the world. This operator was very close to making a deal with Cisco to build an enormous CDN. Our telco vendor partner pushed me forward. I convinced the customer that they would never monetise hardware based CDNs: imagine spending over one million euros per location on bare metal, without a proper CDN management portal in place. We offered a fully integrated software based CDN portal, and each location would be just a number of Dell, IBM or whatever brand common servers, a fraction of the costs, both CAPEX and OPEX, and all the flexibility. We won the deal, which was the largest in the industry, even for our telco vendor partner it was the largest deal in 4 years, powered by Jet-Stream. Not long after, Alcatel bought one of our competitors and Cisco bought another competitor, both spent serious money (south of 100M $, and never really earned it back, whoops).

Sick of it
You may wonder why our telco vendor partner never bought us: well, it is because they got greedy. Other vendors, such as edge appliance vendors, tried to mimic our technology and tried to push us out of our partnership with our telco vendor partner. To make it worse, we learned that our telco vendor partner secretly tried to build their own technology behind our back, and tried to push us out of the project – the project I won for them- even though our solution was already accepted by the customer. So far for ethics, so far for NDA’s. Their attempt to copy our assets horribly failed of course, but the result was that the customer kicked them out and our telco partner never compensated for our losses. What a mess. What an intellectual poverty. Shame on them, because we could together really have made a punch in the market. Their behaviour literally made me sick. I decided to stop licensing our technologies, it was not worth the hassle. We are still going after these vendors: never let anyone steal your properties. I may one day publish all the details of what these vendors did, to help other innovative companies prevent what happened to us. Insiders know who they are.

Strategies
Fortunately, StreamZilla was still there, alive and kicking. And it became the basis for a new idea. Our original strategy was to license our technologies to as many telcos as possible, and then connect all these infrastructures together: to build a huge federated Content Delivery Network, powered by all major telcos in the world. Such an integrated CDN would help the telcos win back market share from global CDNs. Global CDNs had taken full control over distribution, because the telcos had been sleeping. Unfortunately, the telcos were still sleeping. They could have owned the delivery space if they had not left a wide open opportunity for global CDNs. Federated CDN was never going to happen.

We wanted to try another approach: first we implemented HTTP adaptive bit rate streaming and went beyond what regular CDNs offer. HTTP streaming opened up new opportunities. We thought: what if we could bundle buying capacity from broadcasters and publishers, and add intelligent algorithms to dynamically load balance our customers traffic over all major Global CDNs? We would be in control which CDN gets which traffic volumes. That is a great strategic position to be in. We can redefine who gets traffic, and business. I invested heavily in integrating all major CDNs in our platform. And we succeeded. I also invested heavily in performance and pricing research, which resulted in new insights and new load balancing algorithms: we proved we could reduce TCO streaming costs for content publishers by 50%. 50% lower costs and much higher uptime and guaranteed performance!

MultiCDN
We call this new strategy MultiCDN. Some of our new capabilities are:

  • Enforced load balancing: we instantly load balance traffic, without any dependency on third party DNS systems (if one CDN fails, we instantly kick them out of the pool)
  • CDN stacking: we load balance over a pool of CDNs, effectively building the largest CDN in the world (use the capacity of all CDNs together)
  • Fine granular load balancing: optimised load balancing per ISP, down to IP addresses
  • Deep CDN integration: we control global CDNs as if they are our own edge servers: flushing, monitoring, access control, reporting, log file analytics, all automated
  • Feature rich: all our own features work regardless the mix of underlying CDNs
  • Mix Global CDNs and edge servers: optimise traffic over a hybrid mix of resources: your own CDN and your own edges, plus global CDNs

In the past two years, we have tested and trialed MultiCDN succesfully. We tried to build a partnership to form a purchasing alliance… we actually won the tender, only to be outed by this (now former) partner who too got too greedy. They learned from us, and tried to introduce our inventions and models as their own… One year later, they are failing as we speak… And we are fully prepared to protect our interests.

Anyway, negativism is bad, the best way is to go forward, so we kept building, developing and overhauling: and the result is the all new Jet-Stream.com. Hope you like it. I am proud that we can deliver such a unique, feature-rich MultiCDN service at such low rates. We can provision it in minutes. Nothing compares 🙂

I am proud that over a dozen customers already signed up for our new MultiCDN offering. If you want to read more about our work, see:

MultiCDN page
MultiCDN description
MultiCDN use cases
MultiCDN comparison against simple overlay
MultiCDN pricing

Stream on! 🙂

 

 

Geef een reactie

Vul je gegevens in of klik op een icoon om in te loggen.

WordPress.com logo

Je reageert onder je WordPress.com account. Log uit /  Bijwerken )

Google photo

Je reageert onder je Google account. Log uit /  Bijwerken )

Twitter-afbeelding

Je reageert onder je Twitter account. Log uit /  Bijwerken )

Facebook foto

Je reageert onder je Facebook account. Log uit /  Bijwerken )

Verbinden met %s