GoGrid is one of the most innovative “full service” Cloud providers for over 2000 business customers in more than 100 countries worldwide. Clients include small to medium technology companies, large enterprises, International, Federal, State and local government agencies, and web hosting resellers. Together with GoGrid’s parent company ServePath, GoGrid is able to offer dedicated hosting, public clouds, and a highly flexible hybrid solution, plus a wide variety of managed services that are usually fairly complex and costly to provide, such as disaster recovery, content delivery networks, or load balancing, paired with a “10,000%” Service Level Agreement (SLA) – six minute service outage means ten hours of service credits. Recently GoGrid made its proprietary cloud computing operating system dubbed ‘SkySys’ publicly available to transform existing datacenters into Clouds, allowing multi tenancy including billing, real-time reporting, and support.
DMR: How do you position GoGrid in the market?
John Keagy: From the telecom provider perspective we are next to Amazon the largest Infrastructure-as-a-Service provider with a feature-complete offering. What I mean by that is that we are offering both Linux and Windows servers in the Cloud with integrated cloud storage, integrated load balancing, hourly billing, and we’re doing it for nearly 10,000 customers.
DMR: Are those customers mainly large companies or smaller companies?
John Keagy: Those are Web 2.0 companies and Web 2.0 business units of large enterprises. For example Macy’s with their iPhone app is a good example of a larger business that has a Web 2.0 business unit that’s doing business with us.
DMR: What’s so attractive about your offerings that specifically attracts Web 2.0 players or business units?
John Keagy: These are people with production workloads and they like us because we are true to standard datacenter computing concepts such as persistent storage, persistent IP addresses, contiguous IP addresses – real reliance. Those are standards of protectoral constructs that they’re using already in on-premises computing or in their other installations and that allow them to easily work with GoGrid.
DMR: Are any of your clients, especially business units of larger companies, concerned that unanticipated capacity bursts could increase the costs massively?
John Keagy: No. GoGrid offers very competitive pricing with great volume pricing plans that with scale make our pricing even more affordable than Amazon. And that’s one of the great things about Cloud computing: you pay-as-you go. So cost control is much easier.
DMR: Who is signing the check right now? Is that someone within the Web 2.0 business units ordering services and putting them on her credit card, or is that more like a C-level decision enabling business entities to order services?
John Keagy: It’s typical done at the business unit level, and it may or may not be done with a credit card, other times it’s done by invoice.
DMR: You are an infrastructure provider and not a network provider. Early virtualization of desktops and cloud computing didn’t take off as planned because of challenges in the network. One would have thought that network providers would have a bigger play in Cloud because you need to transport data from A to B, do some load balancing, have many distributed data centers, etc.
John Keagy: I think connectivity is something we do quite well at GoGrid. There have been a number of independent parties that have compared data transfer performance on our network against Amazon and other networks and we’ve done a better job. Not many enterprises have that scale to be able to have the connectivity that you need to deliver good global performance. But there’s not much rocket science to do it.
DMR: So what’s the telecoms’ play then in this space? How do you see telecoms positioning themselves?
John Keagy: Telecoms have to decide if they want to be in the business of compute and storage. Network and access has been commoditized. Well, network costs are still making up a large part of the IT economy, but that’s shrinking quite rapidly. Cloud compute and storage is so much analogous to networks because it’s a scalable business that can be delivered on an automated basis where there can be some efficiencies had from scale.
DMR: Is there any particular innovation you would like to see from telecom or other network access providers that you would love to see to either create a new business, grow or mange your business better, or enable a different kind of business model?
John Keagy: No, because IP transit is really all we need. I think all the innovation that’s going on is happening in compute and storage. Services that typically reside on transit circuits like voice have already been commoditized in IT. So: no, nothing I can think of. Everything’s IP from my perspective.
DMR: That’s very interesting. Telecoms traditionally had large investments in network infrastructure rollouts. We’re seeing next generation mobile and fixed networks being planned, 4G and fiber rollouts. Telecoms fear that they cannot survive on micro-margins in commoditized markets. Is there no opportunity or need for value-added connectivity services for Cloud? Can telecoms only play in storage and compute segments?
John Keagy: It’s the only thing I can think of. I’m sorry to say it so blandly but I think IT communications has been resolved to one commodity: IP transit. And the prices are dropping still.
DMR: What’s the next big step you think the cloud industry is going to take? How are Software-as-a-Service, Platform-as-a-Service, or Infrastructure-as-a-Service markets going to develop?
John Keagy: There is going to be a rush of new vendors into the marketplace. One of the most interesting things will be to see if any of the great brand names of telecom and computing really do anything relevant to what GoGrid and Amazon are doing. Right now you’ve got GoGrid, which is not a main brand, and you’ve got a company that’s well known for selling books leading the charge in cloud computing. When are we going to have a great brand name in computing like HP do something that’s relevant – relevant services at a relevant price point? When are we going to have a great brand name in telecommunications really establish itself as doing something relevant?
DMR: If telecoms partner with you or help their clients using your services, what would be the single most interesting innovation they would look back on in ten years and say “We did the innovation X and that really made us shine again.”? Where do you think are the great opportunities to engage with telecoms from your perspective?
John Keagy: The great thing that’s going on in IT is automation. Anything that’s related to automation is what we will be looking back on as transformational. Since telecoms are well suited for delivering automated services at scale it makes a lot of sense for telecoms to participate in this.
DMR: But isn’t that your business or Amazon’s business right now already, to delivery something at scale? Why would you need or want to partner with a telecom?
John Keagy: It would be nice to have some validation in the marketplace. To have a big name brand really do it well would be good for the marketplace.
DMR: Why is that?
John Keagy: The marketplace in general is lacking a great brand name that’s really doing infrastructure and service well. Until the marketplace has that there will be a lot of doubters.
DMR: You mentioned that automation will be transformational. Aren’t you afraid that your business is going to be completely automated and that you could be reduced to a commodity as well?
John Keagy: There is that risk, but we own the automation technology. We’re not a reseller of somebody else’s automation technology; we’re the ones that are developing the capability of automating IT. If you’re looking at the 1.5 trillion dollar IT economy there’s two huge pieces that I think are going to go away: One is telecom cost and the other is labor. Those are two of the largest pieces of what people spend money on in the IT economy and I think that those are going to be driven towards zero.
DMR: CIOs would argue that the labor costs stay the same but focus changes of their internal IT organization; people don’t do so much plugging, installations, hardware, setup, and maintenance of actual boxes anymore but focus more on business planning, business processes, negotiations, SLAs, and contracting.
John Keagy: Most certainly those are necessary things for businesses to do. But I also think that those are executive level skills, business decision making requirements that CIOs have regardless of what they’re doing. Whether they provide IT in-house or whether they’ve outsourced it: they have to decide how they want manage their business and what capabilities and services they need. I’m not that’s entirely relevant…
DMR: What is the most interesting and unexpected lesson-learned during your operations?
John Keagy: I’m every day amazed how different … There are a lot of pundits in Cloud computing that think they know what the marketplace needs. I compare what the pundits say versus what our actual customers and actual users say. And there’s a vast difference between theory and reality.
DMR: For example?
John Keagy: Features. Features of people supposedly want, what’s supposedly important to customers and what’s not. It’s very different than what people like to blog about. There’s a lengthy list of small items and some big architectural ones.
DMR: Like Cloud-in-a-box – which more often than not turns out to be “in-a-box” but not “Cloud”?
John Keagy: Sure! There are also a lot of folks out there trying to sell cloud operating systems to telecoms. But so far the only one that’s proven to operate to a scale telecoms would need is our SkySys operating system. There’s just nobody else that’s achieving that scale, not even in order of magnitude. In terms of a cohesive system that offers all the elements of compute, storage, billing, and customer management and operations, in something that pundits don’t write about but that is a very practical need, and that’s the operational tool for delivering the service. That’s something that all these other companies that think they’ve got a system that can distribute the ends on the hardware. They think they’ve got the system. That’s one of 36 elements of our system.
DMR: Is there a difference between mature and emerging telecoms and what SkySys allows them to do or how they would use it?
John Keagy: So far we are working only with well established telecoms: SkySys is targeting your existing data center, your existing hardware. Emerging operators often either don’t have that data center yet, or – exactly because they are emerging – they don’t want to have them. Purchase equipment, “rack and stack” that equipment, and manage capacity is not yet a best practice for emerging telecoms. It could be.
DMR: Thank you for sharing your time and insights.
