The world of technology is a magnet for buzzwords. Perhaps it is the need of technology marketers to mystify something that in reality may not be that mystical. Or maybe it is that the technology sector is inextricably bound to the VC world, where hackneyed phrases become the secret passwords to the club. It is probably a combination of both, with a little bit of linguistic laziness mixed in. It is always easier to grab a “hot” word or phrase than to concisely and precisely describe what a product or service really is. Seriously, does use of the non-word SoLoMo help one understand the value of a new application better than a sentence or two that actually tells you what it does and how it provides value? Buzzword haters are easy to find in the blogosphere. A trio of PC World writers took a crack at 23 buzzwords that must die on the PC Advisor web site. DOOH veteran and champion of clarity Dave Haynes took a digital signage angle two years ago, and it still rings true. Yet the beat goes on, and new words and phrases emerge, annoy and eventually recede.
One phrase in particular has my hackles raised when it is used and misused in the digital signage space: cloud computing. It is often deployed as the updated buzzword for SaaS (software as a service), when in fact they are not the same concept at all. Cloud computing refers to a highly elastic, on demand computing infrastructure that is massively available and billed based upon usage. SaaS refers to a software application provided as a service on a subscription basis. The best discussion of cloud computing that I have found is from Scott Maxwell of OpenView Venture Partners in Boston, describing its flavors and uses in a white paper entitled, Leveraging The Cloud. It is well worth downloading and reading more than once.
Maxwell’s paper makes it clear that cloud computing is about infrastructure, while SaaS is about applications. Can a cloud infrastructure be used to host a SaaS offering? I suppose that the answer is yes, but unless the SaaS provider plans to charge separately for infrastructure and application services, than how is that different than a private infrastructure and a simplified subscription plan? To the SaaS provider, it comes down to an analysis of cost, reliability and security. To the end user, the server is not behind their firewall, so they should not care unless there is an argument to be made based upon any of those same three factors. So it is easy to be cynical when you see a software provider claiming to be “cloud based”. You don’t get to muddy the waters just because the Internet is often depicted as a cloud in Powerpoint decks. Being in that “cloud” does not make a solution cloud based. If it isn’t the fully elastic, on demand, billed by usage infrastructure as described by Maxwell, then it isn’t cloud based. It is an Internet-based SaaS offering. Stop trying to go all Web 3.0 on us when you aren’t. Sometimes retro works out just fine:
That does not mean that there is neither a present nor a future for cloud computing in digital signage, because there is. Using our own company as an example, while our SaaS offering runs off our own servers, today we utilize cloud-based virtual machines for tests and simulations and after much planning, we will be moving our fail over environment from physical servers to virtual cloud-based servers in the near future. Could we (or anyone else) move production servers to the virtual, pay by the drink world of the cloud? Of course, but because we have predictable and forecastable CPU requirements, the need for elastic capacity is limited on a day-to-day basis. Costs for the physical environment are known and manageable, and we like having control over security as well. As such, moving production to the cloud today would not result in either a cost or service benefit to our customers. So for now, our primary servers remain on the ground.
Moving forward, there may well be emerging uses of cloud computing for digital signage. The first may be for enterprise licensees of an application. Today, that arrangement would require the customer to stand up test and production servers behind their firewall before installing the application. As we look ahead, customers may opt to use virtual, on demand servers in the cloud instead of spending capital and maintenance dollars to deploy physical servers in their IT shop. They would utilize what Maxwell describes as a Private Cloud or Hybrid Cloud. Another high value application of cloud computing may well be ad servers. Since the serving of ads can vary from a one location event to a national event spanning tens of thousands of media players, the need for elastic capacity is obvious. Since ad serving is or could be inherently on demand, and contemplates a revenue generating activity, the idea of billing for cloud infrastructure on a usage basis is particularly appealing. And because many campaigns will span multiple networks, cloud-based serving and billing will more accurately allocate costs. It seems to make good sense. There may be other applications of cloud computing that come to mind. Please feel free to continue the discussion with a comment!
Cloud computing is a hot sector for investors and entrepreneurs. Overzealous marketers have already started misusing the term in an attempt to benefit from some kind of buzzword halo effect. But SaaS is not cloud computing at face value. There are current and future high value applications of the cloud infrastructure for digital signage providers and users. But let’s agree to use the term correctly.
Ken, Thanks for the nice reference. I completely agree that marketing messages are clouding the nature of the evolving infrastructure (pun intended)! My general belief is that the software infrastructure has evolved to a point that companies can get both lower costs and better SLAs by outsourcing all or part of their environments to cloud service providers and that more an more companies are figuring this out, but each cloud service provider is very different right now and I expect that they will continue to evolve as they specialize on particular customer segments and tune their platforms to those segments.
Much more to come!
OpenView Venture Partners
Thanks for your excellent paper and the comment. As the cloud infrastructure becomes more cost effective and secure, it also challenges one of the core tenets of SaaS, that being the magic of single image/multi- tenancy on the hosted server. With virtual servers in the cloud, the idea of multi-tenancy sort of fades away, as each customer could run on its own virtual server (which enables usage-based billing). The challenge for SaaS providers will be to maintain the development discipline and efficiencies of a single image. Multiple image models in the cloud would be very costly to maintain, just as they were in the client-server era. This can be avoided in part by not making the infrastructure a marketing feature, as it doesn’t have to be, nor is it today. Looking forward to what is next!
Agreed. We (Rise Vision) have the same challenge. We are a Saas and we use a cloud approach for providing the infrastructure. What I find interesting is that most of this nomenclature and education resides outside of the user levels. I’m not sure about Neocast, but I rarely run into this discussion when talking with users. They just want it to work.
If that’s the case with us (and I’d love to know if it is with you) then I find myself much more worried about ensuring the user knows what to do than the VCs.
Thanks for the response. In cases where the customer does not have the equipment on their own premises, you are right, they just want it to work and be highly reliable. I only recall taking one prospect in to see our production server rack. They became customers by the way, but I think that visit was more diligence than tipping point. My fear is that companies improperly using “cloud” as a selling point will confuse the buyers and make the word meaningless.