November 29, 2006
The cult of SaaS. Nobody expects the SaaSquisition
Posted by theotherthomasotter under IT Related, Irregulars, SAP, SaaS, enterprise 2.0Nicholas Carr, Vinnie, Jason, Phil , David and several others have picked up on the McKinsey Report about SaaS.
For ages I have been trying to figure out what SaaS is. I’m still no clearer, and I have read masses of posts, analyst reports, marketing materials and irregulars emails. It seems there are different forms of SaaS, including the highest forms of SaaSdom, “pure” and “true” SaaS.
Jason asks an important question.
How did McKinsey define SaaS? — Put 20 technologists in a room and ask them to define SaaS, and you’re going to get at least a dozen different answers. Is SaaS any software solution that’s provided in a hosted environment? Is multi-tenancy a necessity? Is subscription pricing requisite? If McKinsey doesn’t tightly define SaaS to the CIOs surveyed, the potency of this data loses much significance.
This is my problem with SaaS. What is or isn’t “allowed” to be called SaaS seems so arbitrary. Have a look at the wikipedia entry. A good bit of it is about what SaaS isn’t.
I’ve mentioned the Monty Python Spanish inquisition before when discussing the high priests of SOA, but it may be appropriate here too.
I never expected the SaaSquisition. Haaa! Nobody expects the SaaSquisition
The unique property of SaaS is that it is hosted and that is multitenant..
Haaa! the 2 unique properties of SaaS are that it is hosted, multitenant and subscription based..
Haaa! the 3 three unique properties of SaaS are that it is hosted, multitenant, subscription based and uses AJAX
Haaa! the many unique properties of pure and true SaaS are that it is hosted, multitenant, subscription based, uses AJAX , REST, RSS , is completely brand new, is bought by the lob, requires no training, isn’t tainted by the evil dark side of on-premise, sells virally without a salesforce, has a seasonal release cycle and knows the secret SaaS handshake.
Make no mistake, here in starship enterprisey we see the SaaS wave. There are some great solutions out there in the “cloud” I see some HR executives buying SaaS offerings instead of deploying in-house applications simply because the in-house IT is too busy doing “mission-critical” supply chain stuff to talk to the HR folks. Successfactors is vendor benefiting from this.
SAP’s approach to SaaS is cautious to date, but it would to foolish to imagine that we are ignoring it.
I find the talk about “hybrid”, “pure” Saas and “true” SaaS very Peoples Front of Judea. I’m tired of all this purity stuff.
David is uncomfortable with my assertion that SaaS is bureau rebranded,
Of course there are still some who question whether SaaS is just an old idea with new marketing spin. In recent dialogue another Irregular, Thomas Otter of SAP, believes that SaaS is simply bureau computing from the 60s and 70s, or ASP from 5 or 6 years ago rebranded. I’ll argue the case for the differences another time, but the key thing is that these sorts of multi-tenanted, hosted (true SaaS) solutions are set to become a much bigger component of the average company’s solutions portfolio in 2007.
Show me what is a) new AND b) unique and I’ll repent, kill a goat and join the cult. The technology has moved on since punchcards, but the business model has not changed. AJAX, SOA etc are not the sole preserve of the SaaS vendors.
SaaS is simply the latest evolution of the bureau. This is a fine thing. Using the term bureau in the 1990’s was uncool, so the term ASP was born. Many ASP’s slipped on the 2000 banana skin. New players needed a new name, something to distance itself from the ASP days, and something hip that would make hosted applications cool again. As much as I hate the 2.0, 3.0 game, SaaS is Bureau3.0. There is nothing wrong with running a bureau well -recently ADP marked the 32nd Consecutive Year of Dividend Increases.
ADP, not Salesforce.com is the largest SaaS provider.
In 1949, when one of Henry Taub’s two partners in their joint venture chose to leave because growth was simply too slow, Henry decided to buy out the remaining partner. Although he was only 21 years old at the time, he never made a bigger business decision in his life. So, for the sum of $6,000, he became the sole owner of the small enterprise that offered, for its day and age, an unprecedented service — but up to that point had only one client. The business was called Automatic Payrolls, Inc. It manually processed company payrolls —providing everything from doing the calculations to preparing the checks and the payroll register.
The idea for the service actually belonged to Henry. As the story goes, one day he was visiting a nearby company. A key employee had taken ill, the payroll wasn’t done, and the workers weren’t paid on time. Company managers, who knew nothing about doing payroll, abruptly put aside their other duties, and together
they finally got the payroll out. Henry was astonished at the disruption that was caused. Employees were angry. Managers were frustrated. Productivity suffered. All because a key person wasn’t available to prepare the payroll.
He wondered how many other companies faced similar crises, because a payroll wasn’t done on time. From that observation
came the unique idea that launched Automatic Payrolls. He would offer a payroll preparation service that any business could use. One that was not only accurate and dependable, but also affordable.
Sounds exactly like the SaaS promise to me.
Read the history of ADP here. You’ll see that they do a lot more than just payroll.
November 29, 2006 at 12:48 pm
Thomas, thanks for this. Over the past 6 months I have been asking people regularly what the difference between SaaS and ASP is, and have come to the conclusion that the only difference is the marketing pitch around them. That doesn’t mean that the hosted/multitenanted model is bad - increasingly it’s the preferred route for our customers at InfoBasis - but we try to avoid using the terms altogether. The problem is that once an ill-defined buzz word gets popular, you can trip yourself up when someone understands a different meaning to yours.
November 29, 2006 at 4:19 pm
Thomas, I agree. This actually looks like something I should be ranting about! Thanks for doing it for me.
It’s a nice new acronym for an old idea.
I’m on board with your cynicism.
November 29, 2006 at 8:37 pm
Donald,
There are vast differences between SaaS and ASPs, but unfortunately (much like any promising technology trend), too many people who aren’t SaaS are painting themselves that way.
A lot of existing software companies, including a good many that started out as MSP/ASPs but hung around during the lean times, are calling themselves SaaS and OnDemand.
All that really means it two things…they are charging on a subscription basis, and they are hosting the software instance themselves.
To me (I’m a bit of a purist here), those are hardly the things that make SaaS such a potentially disruptive trend within software development and deployment.
Multi-tenancy is where the potential TCO values come into play, no way around that. Unfortunately being able to deploy a really useful set of services in a secure and scaleable cloud environment isn’t easily done; thus we’ll continue to see small and large vendors pretending they’re SaaS when in fact they’re really just offering traditional software code in a slightly more fluid way.
Make no mistake, there is value to the customer by providing better pricing AND offering to host an implementation; but if that’s all SaaS was about, it would hardly be worth the attention we all give it, nor would it warrant the focus SAP, ORCL and the VC community are putting behind it for their own reasons.
November 29, 2006 at 8:41 pm
[...] My friend Thomas has been trying to steal my job of pointing out hype of late, so I thought I better write a quick post to try and reclaim my title. (although this has less to do with hype. . .) [...]
November 30, 2006 at 3:26 am
[...] SAPs Thomas Otter expresses his ‘confusion’ about the term SaaS David Terrar focuses on the financial case and its impact in the larger organisation Jason Wood looks at the economics, broadly agreeing with McKinsey but raising the odd alarm and wanting more market detail Phil Wainewright weighs in with some fascinating insights into where the market goes next. Vinnie Mirchandani disagrees with analysts and proposes an alternative based on his observations as a cost negotiator and outsourcing specialist Nick Carr whacks Thomas over the head with his idea of a clue stick: Thomas Otter, a smart technology blogger who works for the corporate software giant SAP in Germany, finds himself baffled and befuddled these days. He can’t figure out what “software-as-a-service” means. [...]
November 30, 2006 at 7:07 am
nowhere in this discussion and elsewhere is the question being addressed - what are the thousands of incumbent SAP customers going to do? Many cannot or will not move to SaaS. SAP and Oracle and salesforce.com are too busy arguing the purity of SaaS and chasing after new customers. Incumbent customers do want lower maintenance and to be able to leverage some SaaS concepts like utility computing and fractional resources and automated or at least painless upgrades as I wrote in SaCS - software as a customized service below.
http://dealarchitect.typepad.com/deal_architect/2006/11/sacs_software_a.html
Why is SAP not encouraging itself and its ecosystem to provide that to current customers?
November 30, 2006 at 8:47 am
That’s not strictly true Vinnie. Depending on how you wish to define SaaS from an apps perspective, SAP is building out a massive ecosystem of developers using all this good stuff.
I was surprised to find an entire gorup dedicated to building widgets. How current is that? Now assuming they get these things to do useful, non-trivial ’stuff’ then there is no reason to assume this won’t spill into the main SAP community. Is there?
SAP devs are no different to any others. They like to share and show off their technical prowess. If that brings benefits to the business, and, as seems likely at low cost and low risk, then does your initial thesis really stand up?
As regards core - probably - but as regards value around the edge? I’m not so sure.
If on the other hand you’re focusing on the cost card then your thesis stands up no problem.
SaaS has many dimensions.
November 30, 2006 at 12:12 pm
[...] Following some e-mail dialogue and recent posts, my fellow Enterprise Irregular Thomas Otter wrote a very entertaining post referencing Monty Python’s Spanish inquisition sketch. Thomas believes that the SaaS term is all hype - just 60s and 70s bureaux services rebranded, or ASP services from the late 90s repackaged by vendors so they can distance themselves from being associated with the dotcom boom and bust. Over on Rough Type Nicholas Carr gives him a good thwacking - Dennis has invented a new term for this - I’ve been “Nicked”. Thomas has looked at the Wikipedia definitions, and he suggests that they say more about what SaaS isn’t, rather than what it is. I can understand an element of his confusion, because there are plenty of vendors with a bureaux approach, or with a hosted application who want to jump on the SaaS bandwagon and use the emerging branding for the sector. That’s why some of us vendors start to talk in terms like “pureplay” and “true” SaaS to try and differentiate ourselves from the pretenders who are stealing our clothes. It does begin to sound a bit like a religion, so the inquisition reference is spot on. In the comments on Nick’s post, Charlie Wood suggests: “I think Thomas is right that “SaaS” as a category has very fuzzy boundaries, but I’m not sure it matters. Mark (Crofton) gives SAP’s standard response of “customization”, but that’s just a euphemism for “complexity”.”In his post, Nick uses Glovia as a good SaaS example, but Dennis wonders: “But to be serious for a moment, the whole point is SERVICES (plural) beyond the hosting. I have no idea who Glovia is - so my bad - but the description you give and their collateral sounds a lot like SoSaaS (Same old software as a service) to me but with a fresh pricing spin.”Vinnie brings the debate back to economics: “customers have little interest in the purity of SaaS models or moving to SaaS whether offered by SAP or Salesforce.com, but are looking for creative ways to lower maintenance and to leverage SaaS type concepts in their incumbent apps portfolio - utility computing, fractional resources etc. It’s what I call SaCS - software as a customized service”I’m not sure inventing a new term is such a good thing to add clarity to a topic that already has “fuzzy boundaries” and people dressing their old hosted applications as SaaS (SoSaas to use Dennis’s pejorative term). However, at the end of the day the customer is much less interested in the details and semantics, and only really wants a stable, secure, scalable, good value solution. But I am interested in the detail, because that is what leads to the value. So when I’m looking at a new SaaS or On-Demand vendor, these are the questions I ask to test if they are part of the one true faith (red robes, soft cussion and comfy chair not required): 1. Is the software behind the service properly web architected? If the software is based on some previous client/server or windows product that is using Citrix or some other technology to web-enable it, it may well be working inside a browser, but that doesn’t make it SaaS. The solution will hit performance, usability and very definitely scalability problems. We started like that, hit those problems and re-architected. One of my UK competitors uses web-enabled approach. They give out a multi-page guide on how you configure IE to make it work (and it doesn’t always work), so this is one of the signs you can look for to spot the difference.2. Is the solution using a multi-tenanted, 1 to many model? Although the vendor may have hybrid, or other implementation options available, is their core solution multi-tenanted? This is the key way they can make use of economies of scale within the ingredients of their solution, and pass those benefits on to the customer with more cost effective subscription pricing at a much lower level than traditional software, when you take Total Cost of Ownership in to account. 3. Do they use true “pay as you go” pricing? Changing the pricing to monthly rental is not enough. Does the customer have the ability to discontinue on a month’s notice? If so, then that is SaaS. Is there some sort of 3 or 5 year contractual agreement along with my monthly subscription? If so, then all I’m doing in practice is leasing the software. 4. Do they have a “self service” component in their approach to support? New style SaaS vendors will tend to have better online documentation, online help, FAQs, forums, wikis and IM based live support. This will vary greatly from vendor to vendor, but you should be able to see a cultural difference in their approach to implementation and support that is learning lessons from the consumer oriented, web 2.0 style solutions. 5. Do they have a SaaS business model? You’ll need to do some investigation to check this out properly, and it overlaps with question 4. True SaaS vendors have to adopt an entirely different business model to traditional software companies, which affects every aspect of what they do. The structure of their sales force and sales approach will be different, and they have to pay their sales people in a different way - no big up front licence and service contracts to pay big commissions. A SaaS customer might start with a small pilot project and a few users, to test things out, and so the risks of the project are suddenly transferred away from the customer and on to the vendor. The vendor needs the small pilot to be a success, because they only really begin to make money when the solution is rolled out to a wider community of users, which might be in year’s 2 or 3 of the project. The vendor’s development and release cycle is completely different. Instead of a 12-18 gap between new releases, the SaaS vendor can do monthly, or even weekly product releases in much more controlled circumstances (only one platform and several browsers to support, rather than multiple operating systems, multiple database platforms, and complex PC client implementations to test and get right). It means the SaaS vendor can get new functionality out to market much quicker than the traditional or SoSaaS vendors. And they have to put more emphasis on all of the elements of support and helpdesk. As well as making sure the various implementations succeed so the population using the service increases to a full roll out, there is always the risk that customers could pick up their data and move to a competing service at a month’s notice. Now, obviously this wouldn’t be an easy change for the customer, like switch electricity suppliers, but it would be significantly easier on a SaaS platform compared to a traditional software implementation. The SaaS vendor has to be mindful of the increased risk of attrition, and so puts more effort in to support. For a particular vendor, the answers to 4 or 5 might not be so clear, but the aggregate of all the questions should help you see the true, pureplay SaaS vendors from the ones who have an old (and perhaps very successful) bureaux application, or an ASP approach that they have simply relabelled. Technorati Tags : Enterprise+Irregulars, SaaS, on-demand, ASP, enterprise, bureaux, Monty+PythonPowered By Qumana [link] [...]
November 30, 2006 at 12:16 pm
Jason
Thanks for your thoughts, and for your excellent blog on SaaS exemplar Salesforce.com.
I take your point that multi-tenancy is the key to TCO gains for clients. It means gains for software providers, too, like ease of deployment and not maintaining a long tail of legacy installations.
So, yes, there has been a shift in the delivery of software. I suppose my concern about SaaS is one of communication: if the market doesn’t agree what it means, is it a potentially mis-leading term to use?
November 30, 2006 at 1:32 pm
Donald, Funkmeister,
I’m afraid I don’t agree. See the trackback above this comment to my post on the 5 questions to ask to differentiate from a true SaaS vendor over an ASP or bureau provider. There are plenty of ASP and hosted solutions who are misrepresenting themselves as SaaS. Part of the confusion is that the difference doesn’t just involve technology, but the business model adopted by the vendor. Although web architected (as opposed to web enabled), and 1 to many multi-tenancy are important components, the real differentiation comes in the different business model. For a (true, real, pureplay, member of the faith) SaaS vendor, everything from sales, through production to support is different when compared to a traditional software vendor. Some ASPs have some, but not all of the characteristics.
November 30, 2006 at 11:28 pm
All this discussion about what is and isn’t SaaS is interesting, but mostly misses the point. The point is not about the features of SaaS, but the benefits. Within the benefits live the differentiators.
Customers deserve to be able to 1. focus on their business and business results - not technology 2. be ably and easily supported in reaching their goals and c. always have the most up to date tools at their disposal so they can nimbly and seamlessly meet changing needs.
These benefits are the ones that, to my mind, truly differentiate SaaS/On-demand. To the extent that these nicknames help us communicate those benefits, they are great. Once they start getting in the way, they are just a confounding element.
More on the blog.
November 30, 2006 at 11:31 pm
[...] A post over at Vendorprisey (with a very clever title) got me thinking about Software as a Service (SaaS) again. So often, conversations about SaaS involve an argument about what is and what is not considered SaaS. [...]
December 1, 2006 at 10:59 am
Hi Max,
I’m completely with you on this… in fact my company’s strap line/motto is “think Business, not Technology”. The benefits of the SaaS solution (achieved through a combination of particular technology approach and different business model) are the important things. I just get frsutrated when so many “not SaaS” solutions dress themselves up in SaaS marketing clothes. When customers investigate, and do a proper benefit analysis and TCO they should identify the differences.
December 1, 2006 at 6:52 pm
[...] I’ve been thinking about BPO and SaaS alot lately. They aren’t exactly the same, but they have a lot in common. Jim from Gartner has provided some definitions in response to the SaaSquisition. [...]
December 1, 2006 at 10:07 pm
[...] Wow, talk about a (relatively) hot topic! Recently, a web of blog posts full of coined terms, rebuttals, and re-definitions was spun by some of the blogs I frequent. I recently read articles by Dan Farber, Phil Wainewright, David Terrar, Thomas Otter, Nick Carr, Dennis Howlett, and Vinnie Mirchandani (whew, monster list!) that discussed the topic of “What is SaaS?” After sitting it out and thinking about what I read, I figured I would finally jump in. [...]
December 5, 2006 at 10:05 am
[...] When Thomas was suggesting the SaaS term was all hype and marketing spin and just bureau solutions revisited, I responded with the 5 questions I ask to differentiate SaaS vendors from SoSaaS, ASP or hosted solutions. Where the vendor’s solution is more like traditional software, but hosted, they often focus on the on-demand term, rather than SaaS, as this is less clearly defined. But as several people pointed out, from the customer’s point of view, if the solution is available through a browser, does the job, and at the right monthly charge, does it really matter? From a buyers point of view, I still think it is valuable to make the differentiation, because the true SaaS provider is more likely to have a scalable solution that performs well as their user community expands, and a sustainable business model that continues to allow the vendor to provide high quality support, and allow the application to be regularly enhanced to keep pace with the market. Jim Holincheck, at BlogERP, has weighed in with Gartner’s definition of SaaS, in an good post that puts SaaS in to context amongst other deployment options. Although most of their material on the topic is only available to subscribers, Jim explains: The Gartner definition has three requirements: 1. The application is owned, delivered, and managed remotely by one or more providers 2. The application is based on single set of common code and data definitions which are consumed in a one-to-many model by all contracted customers at any time. 3. The application is licensed on pay-per-use or subscription basis I would suggest these 3 make good sense, but are the minimum requirement. I would add the following: 4. The application behind the service should be properly web architected, rather than an existing client/server or Windows application simply web enabled. 5. The vendor should have a true SaaS rather than traditional software business model. In dialogue on this topic with the Irregulars, Jeff Nolan agrees with me that this last item is actually the most important. To reiterate, using the words from my earlier post, it affects every aspect of what the vendor does. The structure of their sales force and sales approach will be different, and they have to pay their sales people in a different way - no big up front licence and service contracts to pay big commissions. A SaaS customer might start with a small pilot project and a few users, to test things out, and so the risks of the project are suddenly transferred away from the customer and on to the vendor. The vendor needs the small pilot to be a success, because they only really begin to make money when the solution is rolled out to a wider community of users, which but be in year’s 2 or 3 of the project. The vendor’s development and release cycle is completely different. Instead of a 12-18 gap between new releases, the SaaS vendor can do monthly, or even weekly product releases in much more controlled circumstances (only one platform and several browsers to support, rather than multiple operating systems, multiple database platforms, and complex PC client implementations to worry about). It means they can get new functionality out to market much quicker than the traditional or SoSaaS vendors. And they have to put more emphasis on all of the elements of support and helpdesk. As well as making sure the various implementations succeed so the population using the service increases to a full roll out, there is always the risk that customers could pick up their data and move to a competing service at a month’s notice. Now, obviously this wouldn’t be an easy change for the customer, like switch electricity suppliers, but it would be significantly easier on a SaaS platform compared to a traditional software implementation. The SaaS vendor has to be mindful of the increased risk of attrition, and so puts more effort in to support. If you turn those business model characteristics around in to the benefits they give to the user, they’ll be key to providing better value, better service, a more competitive Total Cost of Ownership over a 5 year period, and an application that is steadily enhanced, compared to the maintenance mode and software tax of traditional solutions. So I would propose these are the 5 characteristics that buyers should be looking for in differentiating true SaaS within the current on-demand market. Technorati Tags : Enterprise+Irregulars, SaaS, Gartner, ASP, on-demandPowered By Qumana [link] [...]
February 5, 2007 at 3:04 am
i am ready to join your cult only on one condition if your conditions are favourable cause i am desperate
May 22, 2007 at 11:48 pm
[...] but some good debate has been sparked. David Terrar wraps up the discussion with contributions from Thomas Otter, Nick Carr, Dennis Howlett and Vinnie Mirchandani, and then presents five questions as an SaaS [...]
July 10, 2007 at 10:45 pm
[...] happened to all of that bit of the creed where SaaS would be driven by viral user adoption rather than herds of sales [...]
December 1, 2007 at 3:12 pm
best info for payroll guide
December 19, 2007 at 11:06 pm
[...] ADP isn’t a big noise marketing company, they just get on with it. You could say they have been doing software as a service since 1949, depending how you define SaaS. [...]