It’s become pretty clear in the last couple of years that we are heading irrevocably toward a cloud-dominated future in the software business. The evidence is irrefutable. To attempt to get a traditionally licensed PC or enterprise software business funded by an institutional investor would be a suicide mission these days.
Whether it’s SaaS, PaaS, IaaS, Cloud-based, web-based, Internet-based—WHATEVER, it’s all still basically the same thing. Some folks get very snippy about all the different definitions, but they are all just different segments or interpretations of the same model: Software hosted outside of the customer’s premises and available via an Internet browser. Although the technology has improved dramatically over time, it’s really the same basic idea as ASP (application service provider) model from back before the Internet stock bubble burst.
In many cases this trend is happening for good reasons, with the primary one being the simplicity the model offers end-users. But like anything, it’s not the perfect fit in all instances. For example, I’m still not convinced this model will ever be definitely cheaper than solutions that rely more on local computing power. For that to happen, I think we’re going to need to go back to the era of much cheaper dumb terminals to replace our powerful PCs. Having all that desktop power and storage (and the associated costs) sitting on your desks unused is pretty inefficient.
In addition, I also don’t believe SaaS and other cloud-based variants are necessarily the most profitable business models for every software vendor, even though institutional investors love it. I recently had a conversation with a venture capitalist and I asked him why the VC community was so in love with software in the cloud, specifically SaaS-based models. After some discussion about the various elements of SaaS and customer premise-based software models, it really came down to something simple: traditionally licensed software companies are valued at 1-3X revenue and SaaS-based companies are valued at 5-6X revenues. Of course, it’s all about the money and this makes perfect sense. But will this valuation gap be sustainable, or is it a market inefficiency that will go away over time? But I digress, that’s a topic for a different debate….
There are some very good (and maybe not so good) reasons that certain segments won’t come completely under the spell of cloud-based computing. Let’s take a look at a few areas where I forecast the cloud won’t become dominant:
Banking
This is one of the toughest software market segments there is. Banks are notoriously difficult to penetrate, and security is paramount. I believe this will be one of the toughest segments for cloud-based solutions to penetrate, and will be even harder to dominate. Certainly they’ll be a lot of cloud-based applications in non-critical functions. But anything that gets at the core banking functions, including customer data or money will be kept private. That might be a traditional on-premises solutions or private cloud-based apps, but anything sensitive from a security viewpoint will be held tight.
Government
I believe this will be a similar situation here to the Banking market. Certainly the Cloud has already penetrated many areas of the government, and will continue to do so. But there are larges segments of government services where the data is just too sensitive. We’ve seen a lot of embarrassing breaches lately with respect to intelligence data that absolutely needs to remain secret. I think we’ll see a pullback from this data being available via the Internet, rather than moving deeper in that direction.
Open Source and Mobile
Outside of the cloud, these are the two software segments that institutional investors will still put money into. It’s true that many mobile applications have a cloud-based back-end, and a lot of Open Source platforms are used to generate cloud-based apps. But both of these areas represent code that will sit on customer-controlled assets and will slow the adoption of a centralized model where all computing is done in the publicly-accessible cloud.
Buyers vs. Renters
Some folks just like to own stuff. While the rental model works for many due to the reduction in software and hardware investment, which saves capital for other purposes, others feel that renting is wasteful. Indeed, SaaS and other lease/rental-oriented models aren’t necessarily the cheapest in the long run. This is really a psychographic attribute that isn’t likely to change among those so-inclined.
100% Service Levels required
The Internet is a long way from the old AT&T Ma Bell monopoly when it comes to service levels. Have you ever had the power go out at a company you work at? In this day and age, when that happens, everything immediately stops. With the every-increasing reliance on Internet-based technologies (and being accelerated by cloud-based apps), the Internet connection going down can have roughly the same effect. The Amazon EC2 Cloud Services outage in April 2011 gives a sneak preview of what can happen to productivity levels if service levels are compromised on a wide scale or for a long period of time.
Security Conscious (and the Paranoid)
There are many out there among us that have their own safes rather than using a Bank’s safe-deposit box, or are building safes rooms or bomb shelters to protect against perceived threats they view as inevitable. Many others are simply very cautious and prudent, and that means holding things close to the vest and not embracing the newest technologies until they are viewed as bulletproof. The profiles vary from the prudent to the paranoid, but the common thread will be slow or no adoption of technologies that are viewed as giving up control of something important.
As we embrace cloud-based applications at an extremely fast rate, my own feeling is that we are headed toward a major, high-profile event that will slow adoption considerably. I’m not sure what form that will take, but it could easily be a major data security breach that causes real damage to a lot of people, or an Internet-based outage that brings a bunch of businesses to their knees. There are many examples already which support that these types of events are quite possible. Several times a year now I get a notice that my private data has been compromised by one vendor or another. The Amazon EC2 outage discussed above already gave a number of people pause about being held captive by this model.
So that’s my take on some areas we’ll see little (or at least slow) adoption of public cloud-based software models. Do you see other areas I left out? I’m sure this will be a bit controversial as well—some out there disagree and believe the Cloud will take over the world. I’d like to hear from all of you, regardless of your view. Post a comment to add to the debate.
Follow Phil Morettini and Morettini on Management via Twitter, Facebook, LinkedIn, RSS, or the PJM Consulting Quarterly Newsletter. Contact Phil directly at info@pjmconsult.com
You’re correct there will be some high profile data breaches affecting organizations and the clients they serve. Like everything, SaaS, PaaS, et al, has limitations but the benefits usually outweigh onsite applications for VCs and end users . A few thought leaders at a SaaS summit of 50 executives predicted 3 major outages in the next 18 months I attended recently but I don’t’ think that will slow the cloud juggernaut. Banking and governmental sectors have been and will continue to embrace hosted applications; some governmental agencies have been told to give deference to cloud solutions. I believe one of the biggest threats will be ownership of data, I’m still waiting for a e discovery case that requires a multi tenet server to be confiscated. I enjoyed the post and the question – there will certainly be applications that will never migrate to the cloud but those will be a handful of exceptions.
Given your definition in paragraph 2, you could easily replace “The Cloud” in the title with “SaaS” or “Hosted Services”. So, over a decade later we find that the answer, pertaining to hosted applications and data storage, is basically the same. Businesses do not trust their sensitive data and business processes to be hosted outside the company. Government and banking have been rightfully mentioned. IT leaders in those industries have the motto, “We don’t even trust our own people”. I’ll also add healthcare as another industry that houses mountains of sensitive data.
My ‘snippy’ comments:
I hope it’s clear to readers that cloud-based solutions do not require hosting outside the business premise and do not necessarily require a web browser at the client end. Cloud implementations can be hosted on-premise by the customer and protected from external access. Private cloud was mentioned in the article, but is apparently differentiated by The Cloud in the title.
An interesting perspective – thanks. I don’t believe that going forward the models of desktop and cloud will be as distinct as you make our. Specifically you make three points that I think lead to the wrong conclusion:
– You define cloud plaforms as: “Software hosted outside of the customer’s premises and available via an Internet browser”. I think this is wrong. You final words should be “available via the Internet”.
– You then go on to say “I’m still not convinced this model will ever be definitely cheaper than solutions that rely more on local computing power … Having all that desktop power and storage (and the associated costs) sitting on your desks unused is pretty inefficient.” I don’t think costs will be the driver of success – they are coming down for both the cloud and local processing power (PC and servers). The winner will be the best experience, and that doesn’t have to be an either or, but more likely will be a combination of both. Offerings that take the best of the cloud AND the desktop.
– In regard to your comments on mobile you say “It’s true that many mobile applications have a cloud-based back-end … but both [mobile and open source] of these areas represent code that will sit on customer-controlled assets and will slow the adoption of a centralized model where all computing is done in the publicly-accessible cloud.” As I described above, what is happening here is that cloud platforms are leveraging the power of the local device through a rich app – offering the best user experience of all, the capabilities of the cloud with the power of the device (phone, PC etc.)
Overall I believe all business apps will go to the cloud, and will stay on the desktop. The benefits of this hybrid approach are enormous. Apps like DropBox, iTunes, Skype are cloud platforms that leverage the power of the PC, offering users the ultimate experience.
When Steve Jobs launched iCloud he made a statement to the effect that, the cloud should make great software better. That we should lose what took 20 years to build and go back to square one. Instead we should plug the cloud into the applications that people use and love. There are many great software apps available; the cloud is simply an opportunity to make them even better.
In my opinion, it is this hybrid approach – cloud platforms with rich, application interfaces – that will be the dominant model.
Though you are usually spot on – I think you missed the target on where the Cloud will not be used. For example, Government and Banking.
In the 20 biggest economies in the world, maybe just maybe the Cloud will come late to Federal government – but for the 120+ other countries around the world, its the best answer today. I saw a Cloud application under development just the other day for a military organization no less (in a pretty large country/economy).
In local government around the world, where security and infrastructure is anything but industrial strength, the Cloud is also a no brainer.
I also believe Banking apps for secondary Banks (all but those”too large to fail” Banks) are ripe for migrating to the Cloud – and possibly giving these smaller Banks a leg up on the big boys. Though I am no Banking expert so take this with a grain of salt.
On cost efficiency… the cost of the local computer versus paying for CPU hours on the Cloud is not the issue… its the cost to the business of people not being able to get work done while away from their desk. Sure you can solve this problem with a deicated private app – but its just so much easier to implement on the Cloud. And then when you get surprised and need to scale the solution quickly, t he Cloud is a thing of beauty.
Phil,
Enjoyed the article. Especially the requirements and concerns of “decision makers” evaluating cloud versus on-premise applications. The thing is, they all have legitimate concerns, arguments and perspectives. As a web-based application marketer and also a consumer of mobile and business apps, I “get” the pros and cons that exist. I too, am a recipient of “your account information may have compromised, please re-set your password” messages. It is scary and I wonder if it truly is all worth it. To comment on your request for other industries not embracing or slow to embrace on-demand solutions: while some areas of healthcare certainly are embracing the cloud, there are many segments that just will not implement cloud solutions due to insufficient resources (and internal cloud technology strategists), but also due to privacy regulations and just plain old disorganization of internal business processes.