Subscribe to Morettini on ManagementI have this incredible feeling of déjà vu.
Cloud computing and Software as a Service is all the rage. In my practice at PJM Consulting, I am very involved in software startup activity. Nearly every new software company that I see today is being built on the Software as a Service business model. It’s all the rage–so much so that it appears that any self-respecting software entrepreneur would be embarrassed to start a company using the traditional software licensing model. Even if an entrepreneur was so inclined, good luck finding a VC who would even consider funding such a company. No one wants to look like a dinosaur.
It’s all well and good–there is definitely a real trend toward SaaS and Cloud Computing, with many good reasons for it. But most high technology trends are initially a bit over-hyped, and tend to get ahead of themselves. In addition, this particular story seems ever so familiar to a tech veteran that’s been around for a few of these cycles.
The first bit of history this reminds me of is the old terminal/mainframe model from the early years of computing. There were some real advantages to this model, but also some big disadvantages as well–which opened the door for the golden age of PCs and networking.
The second era that the current SaaS wave reminds me of is “Web 1.0″, when Web-based hosted software (then called ASP rather than the modern SaaS terminology), was first going to take over the world. The current trend seems so very similar because it was around the Web 1.0 years of the late 90s/early 2000 when the traditional software license business model was first proclaimed dead. At that time nearly every new business plan was based upon an ASP model.
So some of this fast-moving Cloud Computing or SaaS trend is new–but much of it could be viewed as recycled from past trends. Let’s look at the Pros and Cons of this computing model:
ADVANTAGES
* Enables “Utility-Style” computing – variable expense instead of. capital investment
* Allows an end run around overwhelmed IT departments (like PC networking did)
* Supposedly “On-demand”–use only what you need, when you need it
* More efficient use of compute resources by time-slicing large farms of cost-efficient computing resources
* Web-based allows anywhere, anytime availability
* Off-site storage of data assists disaster recovery preparedness
DISADVANTAGES
* Immature and inherently more difficult Security
* More difficult integration with other applications
* Internet latency
* Internet reliability
* Data resides outside the company firewall
* Costs over time aren’t necessarily lower for customers
* Lower margins for software vendors–aren’t always accounted for in current pricing
SUMMARY
I believe that the trend toward computing in the cloud will continue, but there will be some stumbles and pullbacks along the way. Cloud Computing and SaaS has some inherent strengths–but also some under-publicized weaknesses. Many software vendors are overlooking the weaknesses at this time, as is typical of any new and hyped technology. Traditional licensed software hosted by the user still has its strengths and a definite place in the market. Like many mature technologies and business models, the death of traditional software licensing has been greatly exaggerated. Once the early hype passes, decisions on whether to computer within the firewall or in the cloud will once again be made on the individual merits, costs and user needs for a particular application within a particular company. That’s how I see it–post a comment with your opinion so we can look at all viewpoints.
Phil Morettini
PJM Consulting
http://www.pjmconsult.com/
Niall Halpenny
/ May 11, 2009Hi Phil,
I liked your BLOG. While I do believe in the promise of Cloud/Saas in terms on bringing greater efficiencies, I don’t see it as being all things to all people from day one.
Therefore I’d be interested in your view any detailed market segment analysis you’ve come across which indicates exactly how big the realistic target market for these technologies will be in the next 3-5 years.
As you’ve pointed out, there are real issues such as the underlying quality of the connection and availability, etc. which are real obstacles to many potenital cusotmer adopting these new models.
Regards,
Niall
Phil Morettini of PJM Consulting
/ May 11, 2009Niall,
I don’t know of any market analysis, but I haven’t been looking. I’m certain that there are projections out there on a trend this important. But I would caution against you using such studies as a basis to making a business decision–these types of projections have a habit of being notoriously and dramatically wrong–usually on the upside. The best advice I can give you is to analyze your market and user, and meet their needs. But if you are contemplating a new product, there is also nothing wrong with coding your product in such a way that it can be delivered EITHER via the cloud as a service, or behind a customer’s firewall as a traditionally licensed product. This gives you flexibility to serve users of different psychographic profiles, and effectively hedges your bet.
cprimault
/ May 25, 2009Hi Phil,
Thanks for your Post. I have also lived the proprietary to Unix period 20 years ago and well, the mainframe guys still make good money
A quick question for you. Rather than whemn, where do you think IaaS and SaaS will take off fist?
Have you made any analysis on the segments that will first get value out of these models?
Thanks for your insights
Christophe
Phil Morettini of PJM Consulting
/ May 25, 2009Christophe,
The segment taking off first in SaaS is unambiguous–it’s CRM. And that is primarily Salesforce.com.
I don’t necessarily think it’s because of the segment itself, but the execution of Salesforce. They are now upping the ante and trying to become a Cloud Computing platform, rather than just a CRM application. Using their technology to enable others to create SaaS apps that can be integrated, etc. The thing to do is to analyze the segment you are interested in directly: How important is the data portion? Will people be uncomfortable that this data isn’t behind their own firewall, from a security standpoint? How big of a problem will Internet latency (or completely down Internet connection) be? How complex is the app to deploy each time within the end-user’s firewall? And so on. This will give you your answer on an segment’s near term SaaS potential -Phil
cprimault
/ May 27, 2009Thanks Phil, good tips.
Mark McClure
/ September 15, 2010This quote from your article speaks volumes to me:
“Allows an end run around overwhelmed IT departments (like PC networking did)”
The ‘battle’ between embedded IT orgs and what the business thinks is available out there ‘in the cloud’ to reduce costs (and shrink IT to some core specialists and governance types) is gathering momentum… but I imagine that concerns around risk and compliance are heavy anchors which may slow down, or even stall, adoption by many large corporations for a while.
Eddie Soong
/ September 19, 2010Thanks for your view Phil.
Things go in and out of style, like fashion. While fashion get its cue from consumer fancies, technological trend is driven by needs, and the technological advancement available to support that trend. By the same token, this advancement was responsible for the changes in the way we work, live, and affects many aspects of our lives, e.g. the invention of smart phone. We all know what Internet has done. Mobile computing is happening. Their impact is not over but constantly involving. The user community has grown into main street. The only way to serve this huge and growing user base is utility computing, currently identified with Cloud-based and SaaS solutions.
The problem however is that some companies would emphasize SaaS and Cloud when describing their products for the sake of getting on the band wagon. You can quite easily identify them when their description of SaaS and Cloud just don’t make sense.
The points I’m trying to make are:
- the feeling of deja vu is because solutions are constantly morphing to meet the needs and changing market demand in the best possible way given the current state of innovation
- old technology, like mainframe, client/server, etc… will still be around, albeit in niche areas, until such time when a better solution becomes available supported by the latest innovation.
Eddie
Mike Harris
/ September 19, 2010Hi Phil,
Great post. I laughed when I saw ‘ASP’ and remember a prominent San Diego startup that spun off of Titan years ago. It blew through a ton of money and eventually went by the wayside.
Of course that was then and this is now. As you mentioned, when a company’s data resides outside the firewall huge security issues ensue. I believe this was the key reason the earlier ASP models failed. I don’t see that this has been satisfactorily resolved yet.
When it gets resolved, however, we’ll see Web 3.0 take off like the proverbial rocket. (My definition of Web 3.0 is the addition of artificial intelligence to the cloud so that when I text my friend asking if he’s aware of a good tire store I may receive the latest tire sales and promos in the area where I live. Of course the permission issue is equally important as the security issue.)
admin
/ September 19, 2010Mike, interesting ideas about adding AI to the cloud. I haven’t seen anyone talking about that to date.