The Problem With SaaS: Lack of Flexibility

I’m deep into the Wave research (read my latest status update) for nine community platforms (big list here), in fact, this week and next, I’m in 5 and a half to 6 hour meetings with vendors getting to know them very well. Of course, that’s just one side of the story, I’m interviewing up to 27 of their customers (brands) to get the ‘other’ side of the story. You’ll see me pop up online for “Twitter breaks” here and there, surfacing for air, then submerging back again.

I wasn’t surprised to hear from a few brands that not everyone is a fan of Software as a Service (Saas), in fact, for some, on-premise software makes a whole lot more sense. What exactly was the pain point with Saas?

Of course, the most obvious points continue to come to mind: conservative companies, or those with truly critical data wanted to have the data on site, or wanted to self-support their own architecture, or wanted to modify the software at their choosing, yet I learned about something that I hadn’t even considered:

Some brands felt that being tied to SaaS meant they were at the mercy of the vendor when it came to the software architecture. While this is certainly the case not just for SaaS but all non-opensource software, when a vendor would upgrade a version, some brands were forced to accept the changes. In some cases, some brands were not aware/prepared of some upcoming upgrades to the SaaS software and were blindsided to the changes.

Furthermore, these brands felt that the larger customers who wanted specific feature upgrades were able to influence the vendor through direct pressure or even fueling the R&D work through feature requests. What’s wrong with that? Alot, if the feature requests of the larger customer don’t reflect the needs of a smaller SMB customer. Since the software is delivered over the web, there was little recourse for the smaller brand to deny the changes, they pretty much had to suck it up.

Now you may ask, isn’t this a problem with all software, SaaS or on-premise? Yes but vendors are more likely to allow customers to use multiple versions of on-premise software, of course there’s a date when they will no longer support it. Yet vendors who provide SaaS are far more likely to provide only one version, forcing clients to swallow the changes on demand.

Have you had a bad experience with Software as a Service? Leave a comment, let’s explore this further.

Related reading: Community Platforms: Here comes the CIO.

Update: As usual, the conversation has also splintered into FriendFeed.

  • Mark MacLeod

    I am both a vendor and purchaser of SaaS solutions.

    As a customer, I like that I have a low initial entry price to get World-class solutions such as Salesforce or Netsuite. I also like that these solutions scale as I do.

    As a vendor, I like the predictable revenue stream – a lot.

    I just think SaaS is the right model. VCs seem to think so to. And public markets put a premium on SaaS plays vs. their SW licensing counterparts.


  • Jon Erickson

    I like Mark about am both a vendor and a user for Enterprise solutions. I say that the SaaS solution is so much better for both the vendor and the user. We use to be limited to one or maybe two releases (not new functionality, just fixes and minor upgrades) when we were a full licensed on-premise solution. But with SaaS there is not the same issue, we can roll out new functionality and enhancements on a weekly basis if we want. Sure a major release is still something that needs to be staged and coordinated, but the clients and the vendor will with incremental releases.

    I also am a proponent of major companies have a large say into product direction. They are usually the ones who use the product most extensively and imho the SMBs benefit greatly from these best practices by getting this functionality in the core product. Back in the 90s we would do all of this through custom enhancements and they were protected as “trade secrets” but now they are shared and everyone benefits.

    I also think you are going to start seeing some scripting languages coming out that will allow flexibility for SaaS solutions that does not change the underlying code. This will be a great step forward, think Greasemonkey on Salesforce.

  • Steven L.

    Hi James,

    Many thanks for your great blog. I only wish you spent a little less time on it so that you could finally finish the much expected Forrester research :-)

    I am currently looking at different white label social community platforms and am face with the SAAS vs non hosted solution.

    What I seem to be witnessing is a shift from some of the established actors from a full SAAS model to a mixted model where the customer has both options (one site for example). I see this as a great solution. In spite of its potential shortcoming, SAAS allows companies with no or little experience in social communities to deploy a solution quickly (unless heavy customisations are required), with hardly any impact on the company’s web dev/IT service.

    Past the initial learning curve, the company can then choose to migrate to the non hosted solution, should the benefits of managing a social community be proven.

    This can benefit most from marketers who want demonstrate the interest of social community quickly and effectively within a company without having to fight for ressources internally.

    Well at least that is my opinion at this point…

  • jeremiah_owyang


    Heh, I’m right on target for completing the Wave with our research agenda. It’ll get done, but I’m going to make sure it’s done right. This includes vetting some ideas in public.

    Did you read this post? “Here comes the CIO”?
    It’s related

    -Jeremiah (not James) ;)

  • Steven L.

    Hi Jeremiah (sorry about James :-) ,

    I did indeed read your article with a smile on my face because this is just my life :-)

    I have spent the last ten years or so interfacing between the marketing and IT departments dealing with single sign-on problematics (as neither initially saw the direct benefits), trying to avoid “maverick developpement that would cause nightmarish data transfers and so on…

    So I tought your article was spot on.

  • jeremiah_owyang

    Thanks Steven. no worries.

  • John

    Oh my… the stories I could tell.

  • Scott McClung

    This raises some interesting points about SaaS…and Cloud Computing for that matter. My initial thought is that these experiences say more about the level of customer service the vendor is providing and how it communicates with it’s clients verses the practicality of this technology for the enterprise. This lack of good customer service is what the press has continued to use as reasons to “be cautious” about SaaS but those issues exist regardless of whether the software is on-premise or web based.

    Love the blog, Jeremiah, thanks for the thoughts.

  • Jesse Kliza

    Hi Jeremiah,

    Great points, but I disagree that these are problems with SaaS. They are problems with how many SaaS ISV’s are offering their solutions and running their businesses.

    SaaS as a delivery method doesn’t prevent an ISV from allowing their users to pick and choose which features they want, or when they want to start using them. It also doesn’t force an ISV to be influenced by only their larger customers when making enhancements to their application.

    I’m not saying these aren’t valid points/issues, because they definitely are. Your point about customers not being prepared for changes when they roll out is especially good. However, the title of your post and the generalization that these are issues with SaaS in general, I don’t agree with.

    I think that as SaaS evolves, we will see more innovative models and solutions to these types of problems. Companies like Apprenda (, and other SaaS Platform providers will be enablers to these changes.

  • Paula Thornton

    So I’m REALLY confused. How does the inflexibility of SaaS differ from the inflexibility of $500K software products that have hardwired data models etc.?

  • Rohan Hall

    As a vendor and purchaser of SaaS as well as Enterprise apps, I think as more SaaS apps get into the enterprise structure they will need to provide multiple options. One model doesn’t fit all. SaaS will continue to be a great opportunity for small businesses to have access to great tech at an affordable cost.

    However, enterprises may still need to host their own ‘in house’ version of the app behind the corporate firewall and will require deeper customizations and integration with other existing enterprise backoffice apps (Peoplesoft, Siebel, SAP, Oracle etc.).

    I predict that you will see more of this in the near future (smile).


  • jeremiah_owyang

    John, tell the stories.

    Paula, the crux comes down to choice. A brand who purchases proprietary software for on-premise install has options on when they want to upgrade.

    A brand who purchases SaaS model is forced to upgrade at the whim of the vendor.

    If both models are inflexible, then there’s a lose/lose situation.

  • jeremiah_owyang


    That sounds like one solution –multiple versions of SaaS products, not just one size fits all.

  • jeremiah_owyang


    Right, there’s some issues with this particular case as the expectations between client and vendor weren’t clear –regardless of product delivery. The interesting thing is that the client was frustrated with the SaaS issue –not the lack of communications.

    Jesse, you’re right, thanks for seeing the bigger picture, I’ve a lot of smart readers.

  • Larry H.

    Interesting post, Jeremiah! I don’t think anyone has nailed the SaaS business model yet and most vendors are struggling with how to increase adoption. We will likely see much more conversation on this topic in the near future.

    Re: customers being surprised by upgrades, that seems to be a communications (or lack thereof) issue on the part of the vendors. The latter must inform customers of upcoming changes well in advance, monitor feedback, and adjust their plans if appropriate.

    Re: large customers influencing vendors more than small ones, one of the great things about open source is that this problem is minimized. Of course, that isn’t true for SaaS, which leads me to question whether on-demand offerings need to be better tailored to specific market segments. So Vendor X offers a SaaS suite of functionality targeted to large enterprises and another, complementary suite for SMBs. Any thoughts on this?

    I don’t see much benefit for SaaS vendors to maintain multiple releases of their offerings; indeed, I think it’s not likely to happen ever. #2′s comment above about migrating customers to self-hosted versions of the software might make sense, but it would require most SaaS vendors to radically change their business model, product set, sales force, etc. Major platform vendors are much more likely than startups to provide this type of migration capability, but the complexity for even them is problematic in terms of effect on their business models.

  • Sarah Worsham

    With experience in both the IT and business sides, I think SAAS has proven itself to be the right model for many situations. Internal IT departments are often overworked, so having a SAAS which requires minimal upkeep is good. And the business side gets enterprise level software, usually at a very good price, with minimal investment in terms of internal hardware, software, and support.

    Both SAAS and in-house installations may have updates that are not really wanted. The vendor may stop support for older versions or have security fixes. And software updates include together many changes in either case.

    I think it is important that both IT and business sides be involved in decision making when SAAS vendors are being evaluated. I have seen several situations where the business side chooses a SAAS vendor, without IT, based on business needs but forgets technical requirements needed to fulfill those business needs.

    For either SAAS or in-house, making the right choice should be based on both business and technical needs. Whether the choice is SAAS or not should just be part of your decision making process to fulfill those needs.

  • frank

    I love the debate going on here. A great topic and a lot of good things to think about.

    Thanks for starting the discussion right here Jeremiah.

    At the end of the day it’s an educated decision that needs to be made by the purchaser. There are pros & cons of each solution and the right fit depends on the organization in need of a solution.

    The technology for SaaS will continue to evolve – more elegant ways to handle all the issues will surface over time, but on the flip side you will always have organizations that want to have their data & applications live with them on site.

  • Rohan Hall

    Correct Jeremiah. One size doesn’t fit all.

    The needs of an entrepreneur with a vision to create a unique social network and generate revenue at a low cost is not the same as the needs of a large enterprise with business units, employees, and clients dispersed in multiple countries. Enterprise clients already have a huge investment in technology and have a perceived need to have total and absolute control over this investment. Without this control they won’t easily adopt SaaS technology.

    Therefore the one size fit all SaaS model will need to adjust to this and vendors will need to offer products to address these markets. The core product may be the same, Ex. White Label Social Networking software, but the features, functionality, privacy, security, and integration options are just some items that will need to be different.

    Enterprise clients will also have a greater need for customer specific customizations and features beyond the standard css and html features found in SaaS models.


  • Pingback: Is Your Website SaaS(y)? | Sazbean

  • Philip Soffer

    Hi Jeremiah,
    I’m both a vendor and a consumer of SaaS-based products, after seven years being a vendor of on-premise software.

    I certainly hear your argument about on-premise software being more flexible from the point of view of what version customers deploy and when they get upgrades. But we should be careful about comparing SaaS against an idealized version of on-premise where the business owner has a dedicated staff of IT people who will do all of the upgrades and versioning on the business owner’s timeframe.

    In my experience, this is rarely the case. More commonly, business owners wait months or even years for significant upgrades, fighting with new deployments and other software maintenance in a priority queue that always gets longer. Maybe this is fine for some kinds of enterprise software, but in the social media world it would get old pretty fast.

    SaaS vendors have a strong incentive to 1) make their services configurable so customers can use the features they want and only those features; and 2) get better at communicating their roadmaps and upgrade plans to their customers.

    Add to that the considerable advantages that SaaS can offer around speed of service delivery, quality of service, and generating data to enabling benchmarking of customers, and SaaS seems quite compelling for most applications, especially where the landscape changes rapidly.

    Phil Soffer
    VP of Products
    Lithium Technologies

  • Daniel J. Pritchett

    Thanks for starting this lively discussion, Jeremiah. I decided to break my thoughts out of FriendFeed and give them an in-depth treatment at my own blog. I tried to offer a few ideas on how SaaS providers can plan their way around the major pitfalls you identified.

  • Elliot Ross


    There is one question regarding SaaS that never seems to poke its head out of the sand.

    That is the legal and trade law aspects of the data.

    Who jurisdiction is that data in? the providors or the customers?

    One trade law expert I spoke to said possibly both.

    So, for example, in Canada we have some different relationships with countries of the world that the US does not. (Think Cuba)

    So theoretically, a Candian company has a Cuban client’s data on a SaaS server hosted in California.

    It gives me a head ache even thinking about it :-)

  • Jesse Kliza

    @Jeremiah – No problem, I really enjoy your work and the ideas/thoughts, etc that you share.

    @Everyone – Great discussion here! What I’ve noticed is that there are very few SaaS ISV’s that I’ve come across that are really leveraging the SaaS delivery method to the extent that they could be. Not many are doing really creative or out of the box things. I posted something a while back that may be of interest to some here:

    I’d love to hear what others think about some of these concepts, and be pointed in the direction of SaaS ISV’s that are doing some of those things.

  • Michael Wilson

    As a SaaS vendor, we’re actually surprised at how infrequently the “cloud” vs. “install” issue is a poison pill in a discussion. I think that when positioned as a “speed to market” vs. “greater control” prospects are not as quick to see it as a make or break issue. By minimizing IT resources needed to launch and run a community, SaaS represents a value for companies and brands who are just starting to explore social media strategies and allows them to focus of the community management resources that can help ensure success.

  • Steve Boese

    I would have to agree with most of what Philip Soffer observed about many organizations waiting years to take significant upgrades to on-premise enterprise systems. In the ERP space for example, a major upgrade can be a 12-18 month long project, that puts almost all innovation on hold, as stressed staffs wrestle with simply ‘make everything work in the new version the way it works today’ syndrome.

    I am strongly encouraging my colleagues to get out of the installed and owned software mindset for some upcoming initiatives, and I point to this latest ERP upgrade debacle as my strongest selling point.

  • C.H. Low

    I think the title of this blog may be unfair. I would retitle it as “A Characteristic of SaaS – Lesser Flexibility”.

    One key characteristic of on-premise is indeed the ability to do whatever you want, PROVIDED you have the wherewithal vis-a-vis a budget that is big enough to cover development, IT Staff to run it, backups, redundant systems, business process knowhow (really good developers are in short supply), resources to do Version 2, 3, etc…

    The key characteristic of SaaS is lower cost, unique knowhow of the vendor, and SaaS provides accessibility by clients who cannot get the appropriate technical resources or knowhow w.g marketers being told by their IT/CIO they are too busy!.

    I also challenge the premise that SaaS lacks flexibility. Good experienced SaaS vendors create platforms that are highly configurable from fonts, color themes, components/widgets of the applications, security setup, etc, etc. These provide a lot of flexibility albeit not total flexibility. We certainly pride ourselves that our platform provides such high level of configurability.

    SaaS when done well, is kind of like a menu at a restaurant where you can pick your Vegetables, your meats and your sauces. Enough permutations to satisfy diverse clients.

    On premise software, would be like having your own personal chef. Takes a while to hire a good one. And when s/he is on holiday or quits, you might be in deep trouble.

    Many businesses will have a good enough solution with a good and experienced SaaS vendor. The extra money you saved is better invested in content creation, community management skills, promotion, etc that are just as important, if not more, to the success of social marketing projects.

    Add cost, timing to launch, unique features of a Saas Vendor, degree of flexibility is just a characteristic you need to factor into your decision and also SaaS vendor selection.

    My humble opinion! I have been in 3 previous SaaS businesses since 1986.

    C.H. Low, CEO,

  • Chas Warner

    I work as an engineer providing professional services at a SaaS company and this converstaion really seems to highlight the maturity of SaaS as a business model. It’s here, it’s an optimization that has a place in the market, but the model isn’t refined yet. The flexibility argument is quite valid as I think many have seconded, but I too believe much of this will be overcome soon by the maturing of the technology – API’s, javascript etc. The two most obvious problems to me for SaaS are 1) What services do you provide and not provide – especially boutique services – and where do you draw that line (you can only specialize in so much) and 2) Communication. The complaints about flexibility often mention communication problems regarding versioning and features etc. This is a tiny part of the crux that kills a ton of our man-hours. The whole point of a SaaS company – especially white label vendors – is to integrate with other parties. Not just the customer but everyone the customer is involved with (designers, other services etc.) – it’s about being easy to work with. Not just your software, but the actual company.
    SaaS with find it’s spot in the ecosystem but I think the human side of the communication/integration problem is expensive and painful for SaaS right now.

    Thanks for stirring up this discussion, that’s my two cents.

  • Chas Warner

    Hey Jeremaiah,

    One other thing I wanted to mention about SaaS maturing is that it’s sort of a two-way street. A lot of our customers don’t know how to deal with a SaaS company yet. As a result they don’t know where our services begin and end, they don’t know what to expect and they aren’t sure how to make their resources work with ours. This is strongly tied to the communication issue and it should get better (I like to think) as SaaS is more widely adopted.

  • Clerkendweller

    Something I’ve seen more of recently in my web application security work, is larger organisations in highly regulated industries (e.g. banking, insurance, accountancy) beginning to place additional security requirements on smaller niche SaaS providers. They are asking for security audits, implementation of policies and procedures, penetration testing, etc. It’s actually improving the standards (reducing risk) for all the customers.

  • Pingback: Sample Weblog

  • Steven L.

    Hi James,

    Many thanks for your great blog. I only wish you spent a little less time on it so that you could finally finish the much expected Forrester research :-)

    I am currently looking at different white label social community platforms and am face with the SAAS vs non hosted solution.

    What I seem to be witnessing is a shift from some of the established actors from a full SAAS model to a mixted model where the customer has both options (one site for example). I see this as a great solution. In spite of its potential shortcoming, SAAS allows companies with no or little experience in social communities to deploy a solution quickly (unless heavy customisations are required), with hardly any impact on the company's web dev/IT service.

    Past the initial learning curve, the company can then choose to migrate to the non hosted solution, should the benefits of managing a social community be proven.

    This can benefit most from marketers who want demonstrate the interest of social community quickly and effectively within a company without having to fight for ressources internally.

    Well at least that is my opinion at this point…

  • Maverick Money Makers Review

    well, The SaaS model has now been applied to a supply chain, warehousing, … These businesses

  • Pingback: A website that represents the Web, technological innovation,social network, announcement, recommendations and lessons, how-to website,software, training video tuts and useful sites