Home > Cloud Computing, Cloud Security, Software as a Service (SaaS), Virtualization > Incomplete Thought: Forget VM Sprawl, Worry More About SaaSprawl…

Incomplete Thought: Forget VM Sprawl, Worry More About SaaSprawl…

September 19th, 2009 Leave a comment Go to comments

A lot of fuss has been made about run-away VM sprawl in enterprises who are heavily virtualized due to the ease with which a VM can constructed and operationalized.

I’m not convinced about the reality versus the potential of VM Sprawl, meaning that I have no evidence from anyone facing this issue to date.  I wrote about this a while ago here.

As virtualization and the attendant vendors push more from enterprise virtualization to enterprise Clouds, what I’m actually more concerned with is SaaSprawl.

This scenario describes how enterprises will deal with managing what could amount to dozens of “CloudSourced” SaaS vendors as companies edge toward Cloud adoption by cherry picking applications for externalization using SaaS as the platforms, technologies and standards catch up to allow those pesky workloads that used to run internally, to do the same externally…

Outsource email, security, CRM, ERP, Legal/HR, Purchasing, Desktop apps — all from different vendors, each with different contracts, SLA’s, data integration issues, security concerns, audit constraints, regulatory compliance hiccups.

What we likely could end up with is another illustration of a “squeezing the balloon” problem; trading off CapEx for what I call OopsEx — realizing what might amount to substituting one problem for another as you trade reduced upfront (and on-going) capital investment for what amounts to on-going management, security, compliance and service-level management issues in the long term.


  1. September 19th, 2009 at 07:44 | #1

    This will be handled in the same manner as it is today – distributors, VARs, integrators, etc. If it becomes an issue thousands of companies already exist to help address it. There is already a healthy and growing market of third party tools to assist with user account provisioning, data migration and other cloud challenges.

    As a VAR/integrator I believe there are as many opportunities in the cloud as there are with on-premise solutions. I'll just trade HP, Cisco and Dell for Terremark, Rackspace, and Amazon.

  2. September 19th, 2009 at 07:49 | #2

    @Joe GLeinser Allow me to suggest that our definitions of "help" are simply a matter of perspective. What you've indicated is what my point suggested, that "help" isn't free, it will come at a cost. Managing complexity by abstraction or indirection (read: using third parties to manage other third parties) isn't solving the problem, it's contributing to it.

    I do thank you for your point because it underscores mine; we're squeezing the balloon not popping it.


  3. September 19th, 2009 at 08:38 | #3

    A very real problem, Hoff. I have to wonder however if the decision regarding enterprise use of SaaS is going to be so 'stovepiped' and uncoordinated that the problems you've raised — contract variation, SLA's, data integration issues, security concerns, audit and compliance — aren't a natural deterrent to SaaSprawl?

    Data integration and security/access control seem to be factors that, early on, limit the enterprise's use of multiple SaaS providers… barriers that, in many cases, will be reinforced by the commercial aspirations of the SaaS providers to keep their enterprise customers from cherry-picking. I'm not saying it's a 'good thing'… just that self-preservation by the enterprise IT organization and the commercial aspirations of the SaaS provider will throw up barriers to SaaSprawl.

    If, however, I'm being too generous in ascribing rationality to enterprise IT decisions, and that the decision makers in various functional units make the decision to go to SaaS in as silo'ed a manner as they have often chosen their selection of in-house software… then you're absolutely correct. It will be SaaSprawl disaster of biblical proportion.

    What's your guess? How many of the early adopters of SaaS will have a coordinated decision making function and, subsequently, a coordinated design and operating function? How many will include the CIO / IT organization to supply the rationality and apply the brakes?

    = Rich

  4. September 19th, 2009 at 08:43 | #4

    @beaker We're in agreement concerning the 'balloon squeeze.' My cost analysis to date indicates cloud solutions are about 30-40% more expensive than comparable on-premise systems. More competition will eliminate the cost delta in the near future. Cloud vendors inevitably operate at a higher economy of scale than their clients. I expect that a 'balloon squeeze' is a typical effect of major transitions.

  5. September 19th, 2009 at 08:49 | #5

    @Rich Miller

    Unfortunately, almost without exception, every customer I am meeting with is already experiencing SaaSprawl. There's no "natural deterrent" because in many cases, the business units are engaging directly for non-core services. In other cases, IT is under pressure to "move to Cloud," so they're picking things they don't have an immediate headache with or working to consolidate multiple services using SaaS to do it.

    At this point, there's not a lot of data integration, but that will change as the nature of the apps being moved to SaaS change.

    Infrastructure-centric apps like eMail usually involve IT. CRM apps, not so much (how much real integration takes place when an enterprise moves to Salesforce.com, for example…not a lot, if any.)

    Integration of things like credentials, controls, etc. (driven by compliance) will generally be the leverage points IT and security use to put the brakes to SLOW (but not stop) SaaS adoption…

    I don't have a guess, really…more like observations of what is already happening. Ultimately we'll see consolidation of these services as SaaS vendors roll-up naturally.

    It's nuts out there.


  6. September 19th, 2009 at 08:51 | #6

    @Joe Gleinser

    Without asking you to give away any secrets, when you said:

    <blockquote cite="#commentbody-18935">

    Joe Gleinser :

    @beaker …My cost analysis to date indicates cloud solutions are about 30-40% more expensive than comparable on-premise systems.

    …can you offer up some examples of (at least) the classes/types of services/systems you're referring to? Mail? AV? Anti-spam? CRM?



  7. September 19th, 2009 at 09:23 | #7


    Note that my world is IT up to about 500 employees. It's all I sell to, and all I focus on.

    Here is a small Platform-as-a-service comparison kept short and simple for the sake of any readers: http://www.gcsaustin.com/blog/post/Brass-Tacks-Co

    I've also evaluated email and offsite backup/DR with similar cost variations.

  8. September 19th, 2009 at 09:42 | #8


    I admit to my initial argument being a 'slow-pitch', and you've made the case by saying: "There’s no 'natural deterrent' because in many cases, the business units are engaging directly for non-core services. "

    SaaS is being hailed as the great liberator, providing not only reduction in CapEx, but the ability to make departmental decisions as to where and how to spend operating budgets. Not much different from the way departmental servers and LANs were heralded as 'breaking the dictatorship of the glass house guys' when business units could make independent purchase and operational decisions.

    Point to take away: There ARE a few places where IT can actively and legitimately act as a source of rational decision making, and avoiding SaaSilos. As you point out, there are leverage points.

    Another point to be disproven or not: The market will soon be populated with additional XaaS providers selling Enterprise SaaS Integration Solutions (ESIS ??). There will be, as well, technology providers offering 'appliances' that are purported to ease the management and administrative burden of multi-SaaS enterprises, and which will be installed on premise at the enterprise.

    OMG… this environment sounds ripe for … (dare I say it?)… SOA 2.0 !! And you think it's nuts out there now? Hah!!

    – Rich

  9. September 19th, 2009 at 13:01 | #9

    The difference with the old world is a matter of degree, but there is a difference. SaaS has a chance to do better.

    Taking business apps as an example, the old way was: Vendors were either one-app plays like Siebel CRM, or "suite" plays like SAP.

    Many customers had a mix, with hundreds of different projects involving multiple vendors. There was a time when "best of breed" was very popular among customers, and then later as the costs became apparent – SAP with its "suite" approach did have a business edge. As we all know.

    A similar kind of thing will play out in SaaS world. Those companies that offer "Best of breed" today will have an early win; those that can offer a suite will have a long term edge.

    And yet, while similarities exist, SaaS can do better. One thing were SaaS can do something new is in the area of standards, because it lives in the Web.

    If standards such as OpenID become adopted widely (and there is a good chance), we may see Best of Breed apps sharing a single sign on approach with each other. Would give this better than 50% chance, since customers will demand it and it will confer a competitive edge.

    Also very likely is that most APIs will be technically similar in the sense of using SOAP or REST. However such tech similarity is not a real business compatibility.

    A longer shot (much less than 50%) is the chance that the real business APIs conform to some standard. That will not happen and is where suites will continue (long term) have an edge.

  10. September 20th, 2009 at 15:16 | #10

    Nod. IT has never been an amazing arbiter of the diversity of business processes and needs. Now the cat heard of business demands was just let loose into the SaaS cloud.

  11. Gordon Haff
    September 20th, 2009 at 16:40 | #11

    FWIW, when a spoke with a senior IT guy at Razorfish a while back, he raised what he called "federation" of SaaS apps (which is at least related to what you discuss here) as something he saw as a big issue–much more so than IaaS or PaaS API standardization.

  12. Vanessa Alvarez
    September 21st, 2009 at 02:17 | #12

    A bigger issue lies in enterprises moving away from traditional business processes to adapt to the changes that cloud brings to the table. Organizational behavior within enterprises has always been silo-ed; now, cloud (further) tightly aligns, IT, LOB, finance, etc in order to make it a more fluid, and dare I say "on demand" business process machine internally. With technology, that's easily adaptable; with people, the challenge is bigger. I believe the role of procurement will drastically change to be in line with this new model.

    So in the short term, I agree with Joe Gleinser, there is a huge mkt opp for them; however, I hope that the movement to cloud will push enterprises to internally realign resources to meet this new model, otherwise, it will seem much more of a hassle to make the transition to cloud, and could be detrimental to mass, accelerated adoption.

  13. September 21st, 2009 at 04:38 | #13

    @Rich Miller

    I love the ESIS concept at a conceptual level, but it is a bit of layering that does underscore the fact that SaaSsprawl does drive up costs. That said, I don't see any way around this issue. As long as there are people out there building better mousetraps-as-a-service, the business will want them. You can have no sprawl through some single vendor solution, but it's likely to underperform the best-in-bread solutions that are out there.

    There's a natural tension between efficiency and effectiveness in very many areas, and this is no exception. Sprawl may be more effective for the people who care most, but the cost of sprawl cuts into the efficiency equation across the enterprise.

  14. September 22nd, 2009 at 12:22 | #14

    I've lived through both.

    VM sprawl happens quickly because it is the simplest, most obvious way to segment projects and teams. It's also one of the fastest ways to get a solution up and running without having to worry about complex policies for how users of the service interact with one another.

    SaaS sprawl has been happening for ages. A company I did work for a few years back had no less than 15 SaaS providers for various corporate services. Authentication federation was out of the question in all of them for various reasons. Net-net: users managed a list of 15 passwords which of course had hideous security repercussions.

  15. Anita Moorthy
    January 14th, 2010 at 17:02 | #15

    Hi Hoff,

    My first post in your blog. I have heard a lot about your from my colleagues who follow you and after reading a few posts, I can see why.

    Coming to the question of SaaSprawl – that is absolutely where we see the market going too. And all the reasons for going to SaaS (avoiding IT, keeping decisions decentralized) is going to make full circle and result in centralized decision making and security. But we don't think this will slow the adoption of SaaS – as some of your other readers said, this will just result in creating other tools and services to manage and secure them.

    anita (twitter newbie: trusted_cloud

  16. July 12th, 2011 at 16:53 | #16

    Opting for "integrated" solutions is one way to beat the SaaS sprawl to an extent. The latest trend on the cloud id that of "convergence". Cloud providers bring more and more functionality in a single solution. Our HyperOffice represents an "integrated" solution in the cloud communication and collaboration space.

  17. July 28th, 2011 at 16:27 | #17

    No wonder why Christian Dillstrom let me know about this post – you must be doing a beautiful job as mobile and social media marketing veteran is pointing towards you!

  1. September 21st, 2009 at 07:26 | #1
  2. March 12th, 2010 at 06:49 | #2
  3. September 27th, 2011 at 13:33 | #3