Amy Wohl's Opinions on SaaS

Can Traditional ISVs Make the Move to SaaS?

Reading Phil Wainewright's column this morning about the move of Steve Lucas (acquired with Business Objects) from SAP to Salesforce.com reminded me that I've had a post writing itself in the back of my mind.  That's because I've been having an ongoing discussion about the subject of SaaS with some of the big, traditional ISVs.

It's been broad:

• Is SaaS a significant and lasting change in how software will be purchased and supported or a fad?
• Who is SaaS for?  Enterprise ISVs still seem to think it's for the SMB market (which they don't sell to much yet) and that the Enterprise is too complex and customized for SaaS.   (That is, that their existing market is protected from SaaS.)  Apparently they don't know much about SaaS in 2008 or who's buying it.
• What do you have to offer to be a successful SaaS vendor and when do you have to be in the market?

So, you know what side of these arguments I'm on: 

1. Yes, it's significant and enduring.  I don't fall in with the SaaS enthusiasts who believe that all software will be SaaS any minute now.  It doesn't work that way.  But I do believe that within five to seven years more than 30% of new software purchases will go to SaaS -- and maybe more.
2. We've already fought the "who buys SaaS" battle.  Of course, it's wonderful for smaller companies, permitting them access to applications they could otherwise neither afford to buy or manage.  But it's also very appealing to large enterprises, especially for mature applications that have little or no competitive advantage (think email and HR), supporting remote workers, contractors, and customers, and permitting the enterprise to offer applications that are used only occasionally or by a few workers.  That lets the enterprise use scarce, skilled IT resources to focus on developing applications that are core to the organization's mission.
3. What you have to offer and how much time you have to get into the market depends on what you do.  Very large vendors who address big markets are sure to attract additional new SaaS ISVs to their market -- the prize is just too large to be ignored.  If they wait, they risk being too late, competing with a newcomer for customers they could have brought to the SaaS model themselves (think of what Salesforce.com did to Siebel).  ISVs in narrow vertical segments have longer -- perhaps forever.

Note that offering a SaaS version of your application that is less than the real thing will not work, particularly if you are the standard bearer in a big market segment.  The customers will find it less than satisfactory and a new SaaS startup will find your mistake their lynch pin to guarantee their market success.  SaaS customers, thanks to the web, know the alternatives and want the real thing.  We can't count on marketing in little curtained cubicles to less-than-knowledgeable customers any more.

So Steve Lucas left SAP whose Business By Design is late and perhaps less than what a SaaS customer might have wanted.  In a segment as important as ERP there will be (there already is) lots of SaaS competition in the market who won't make the mistake of offering less than a full-blown product and who will understand that customers may prefer the speed, ease, and economy of SaaS to the last jot of customization.

Blog Missing in Action

Note that my other blog, Amy Wohl's Opinions, seems to be Missing in Action.  When I went to post to it earlier this week it (and its SP) had disappeared.  We have started a new site for the blog, here at Typepad, and started the laborious work of moving some of the posts from our own archives.

In the meantime, if you liked to read the Opinions blog, which is about to have a great story on IBM's new move into Service Management, please go here and sign up.

Thanks,

Amy Wohl, Editor

Many Clouds Ahead: Strangely, Future Seems Sunny

This has been quite a week (or two) for cloud computing.  If I were a "good" blogger, I'd put this into three or four different posts, but then you might lose the point I'm going to try to make.  So you'll have to read this like an article, not a blog post.

Eyes Up

The first thing to keep in mind is that we have some semantic confusion, as is usual in this stage of a new market, around just what is a Cloud.  We are now pretty sure that what we used to call Grids and what we now call Clouds are the same thing.  But we also have things called "Platforms" which seem to be very much like a kind of cloud (and are sometimes called clouds) and then there is SaaS itself which looks very much like a Cloud with some application software (and some SaaS vendors describe their offering just that way).  I'd say we can agree that a Cloud is managed computing power, often with applications, accessed across the Internet.  And I'll agree that a company can have its own Cloud, if it wants one.

IBM has finally enunciated a vision for its cloud computing which makes sense for IBM and its customers.   It offers a very high-end version of cloud computing (virtually unlimited amounts of power, up to and including z (mainframe) systems, backed by its Tivoli system management.  These clouds can be used as part of a large shared infrastructure, where the customer has access to a large pool of computing resources to handle peak activity but doesn't need to pay for all of this infrastructure all of the time.  Some customers (governments in China and Vietnam, for example), are buying and implementing their own Blue Clouds which they will use to support particular projects such as university research or computing for high-tech start-ups.  And, of course, IBM can create a Blue Cloud for an individual Very Large Enterprise, managed by IBM outside the firewall or by IBM or the customer inside the firewall.  The choices are broad.

IBM actually offers even more choices which may be less obvious.  IBM provides a platform to ISVs who want to offer SaaS applications. In that, they are very much like other platform vendors like SalesForce.com, eBay, Google, or Amazon.  A major difference here is whether the platform owner is also an application provider (like SalesForce.com), whose ISVs are related to his application offering or whether the platform provider is simply offering managed computing, perhaps with some technical assistance and/or some marketing oomph.  (IBM offers both.) 

It was reported last week that Microsoft is moving to the Cloud.  In this case, Microsoft means it is going to offer at least one of its applications, Exchange, as a hosted application (SaaS) from its own cloud.  Microsoft intends to continue on the road of “SW plus Service,” meaning that you to use a PC on your desk with MS Office to make use of the Internet-based services that Microsoft also provides.  What seems to be moving to Microsoft Clouds is not personal productivity apps (a la Google or Yahoo’s Zimbra), but shared services (Mail, Collaboration, Customer Relationship Management).  Microsoft has been quick to note that this is a big initiative and that in five years they would expect half their customers to use Exchange as a service.  This is not just something for smaller customers, either.  Microsoft already has some large enterprise like Coke, with 75,000 seats.

Microsoft also has other on-line (perhaps Cloud) services such as their Live services and their new Live Mesh service for synchronizing devices of every kind.  They are also testing a consumer version of office which will combine a basic version of desktop MS Office with an array of on-line services.

Last Friday, we spoke with Workday, a SaaS ERP company we’ve been tracking since its start.  It’s stirred up quite a bit of attention, boldly claiming last April that its feature set would be at parity with SAP by next fall.  So far, they are on schedule with their product plans, with hefty HR offerings, including Payroll and Expenses, and a substantial portion of their Financial offerings.  Look at their latest Workday 4.0 Offering at their web site.  More interesting is the fact that although they didn’t expect to move beyond the mid-market into the enterprise until more of their product was completed, they already have a number of large enterprises and more in their pipeline.

Recently I attended a Digital Transformation Forum at Penn State as a guest of its host, John Jordan.  Clouds and SaaS were definitely subjects of discussion.  There was a lot of lively commentary on whether SaaS was for big companies and whether it could provide the customization and security they needed (SAP was an attendee).   I believe companies like Workday and SalesForce.com are going to prove that SaaS is about what the application does (the solution) and not the size of the customer.  But I think we shall see – and soon.

Think of it this way. 

·         You can access a cloud as pure computing power, a place to load an application that needs lots of computing or requires cyclical peaks

·         You can find a cloud that provides a platform with APIs to support an ecosystem of ISVs who provide software that you find appealing

·         You can choose a cloud whose platform includes an application (SalesForce.com) that attracts other ISVs to surround it

·         You can simply find a SaaS vendor whose application is appealing – you don’t care about the cloud behind it, just the application

·         You can build your own cloud, managed by a vendor or your own IT staff for some large-scale purpose

And no doubt, someone, somewhere, is dreaming up some other things to do with clouds

I sense several likely outcomes:

·         A messy period with many clouds, many of them not interoperable because they’re not build to the same standards

·         An interim period of standardization

·         A period of consolidation because we really want to buy our applications in a way that they can be aware of each other and use them in a common environment.  The easiest way to do this is for each of several vendors to provide a cloud with all of the applications you need and some level of integration.  Remember that nothing keeps a cloud/platform vendor from being one of several vendors who offers a popular application if the ISV agrees

·         A higher level of standardization where (just as we do on the Internet today) any application can be used with any other application on the platform and through the portal of your choice.

The PC is No Longer the Center of the Universe

At IBM's Business Partner Leadership Conference, held this week in Los Angeles, IBM CEO Sam Palmisano and Google CEO Eric Schmidt spoke on the state of computing and their vision for its future, both separately and togehter.  The big news is that they agree:

• We are moving from a PC-Centric world to a Network-Centric world filled with millions of devices of every shape and description.
• In this Network Centric world, the PC is just another device, of decreasing importance as the emerging economies come online with their preference for mobile platforms which more closely ressemble smart phones.
• In the Network Centric World, Open Standards and published APIs are very imortant so that new innovations can easily be attached to the network and become new platforms for deelopment.
• The cloud (very large scale shared computing, available for hire) will become uibiquitious, probably sold by platform providers who will recruit ecoosystems of ISVs and provide a well-managed, secure platform, freeing both customers and ISVs from the need to implement hardware where they lack the skills or capital.
• This networkcentric Computing model will require tons of storage and new interaces and methods to discover the information and manage it.
• This enables everyone to have more access to information.  Some of it will be stored in the cloud, but some of it will continue to be stored on site.  It's a hybrid world.

Large traditional vendors like Microsoft will have to choose between adopting the open standards, network centric, cloud approach, finding a new market niche, and slowly becoming less relevant.  That's why Microsoft's chsing so hard after the consumer market with the Yahoo.

Microsoft Continues to Go Live

At the Microsoft SharePoint Conference in

Seattle

(which we’re attending together with some 4,000 customers, partners, analysts and press), Microsoft is announcing a broadening of the availability of its products in the on line environment. 

Last fall, Microsoft entered the enterprise SaaS environment with Exchange and SharePoint offerings to companies with more than 5,000 employees.  It is now offering these products and others to businesses of any size, available in the 2^nd half of 2008, with a beta starting now. 

This takes Microsoft’s on line offerings much more directly into the heart of its target audience, an indication that Microsoft no longer believes that it can compete with Web 2.0 vendors with on line offerings without an Internet-based offering of its own.  Of course, compared to a group offering like Google’s Site, announced last week, Microsoft’s more mature offerings look much more polished and feature rich.  Microsoft will attempt to also give them the flexibility of Web 2.0 style applications with its SilverLight application development environment, available for download now.

Pricing (and reseller arrangements) have not yet been announced and appear to be still under discussion.  That’s not surprising given that Microsoft has the tricky job of making the Live offering look competitive against Web 2.0 offerings without disturbing its existing customer revenue very much.  Given that many Web 2.0 offerings are “free,” with revenues based on advertising, Microsoft will have to convince Live customers that superior, more richly featured and refined products, with support, are worth more money.

In that context, Bill Gates comment, in answer to a question from the floor during his keynote yesterday, makes Microsoft’s position clearer.  “For most of these Google products, the day they announce them is their best day.”

Now we will see, I suspect, whether what customers want is good enough (at free or a very low price) or better (some would say much better) at a higher price.

Why can't Enterprise SW Be Reliable and Sexy?

There's a raging debate going on over at ZDNet.  You can pick up a corner of it at Phil Wainewright's site, where you can read Phil'a comments (good ones) and check into the original argument from Scoble and the Enterprise Irregulars. 

It started out with BIll Gates wondering why enterprise software doesn't get mentioned much iin the blogosphere.  (I'd note that he doesn't read the right blogs, but I know what he means -- the sexy, big readership tech blogs are all about Facebook, Twitter, and the latest Web 2.0 software.)  Scoble replied that's because the user interfaces on Enterprise sw are awful.  The enterprise guys said enterprise software is about reliability and scalability and not about nice user interfaces.

Everyone is, of course, missing the point.  First of all everyone uses both kinds of software.  Business users have a life outside the office where they shop and socialize and consumers have to bank and rely on the back-ends of big systems, written with big, complex enterprise software.

Now that we've been forwever spoiled by the introduction of good user interfaces on consumer software we want that same ease of use for business applications (enterprise or elsewhere).  A few enterprise application vendors are trying, but it's pretty early in the game.  Software vendors keep telling me they want to show me an application that the users can customize themselves, but they must have different users in mind than the ones I know.  And even the user interfaces (without fooling around with customization) can be pretty grim.

The recipe is clear.  A user interface is user friendly when the user can step right up to it and do what he wants or needs to do without getting someone to help him.  Of course enterprise applications will always have tons of geekware in the basement which users will never understand -- and don't need to -- but we need to be able to get the interface intuitively.  After all, I don't need to understand how to fix an internal combustion engine to drive a car.

And don't think the CIOs who buy enterprise software don't care.  Of course they put reliablility and other enterprise priorities first, but they pay for the help desk that supports users who have to cope with less than friendly interfaces, so whoever builds great enterprise applications with great user interfaces should get lots of customer attention.

WORKDAY IS WORKING OUT WELL

When I first heard (in 2006) that David Duffield, the founder of PeopleSoft, was going to start a SaaS company to build a full-blown ERP portfolio I was really excited.  It made me feel that SaaS was finally coming of age, moving to mainstream, enterprise-oriented, and mission critical software. 

(Of course, there are lots of contradictions here.  SalesForce.com certainly thinks of themselves as mainstream, enterprise-oriented, mission critical software, but perhaps less so in 2006 than today.  And Workday, Duffield’s new company was not quite so ambitious; the initial goal and the current focus is to build ERP software for the mid-market (focusing on 100-600 employee companies) [update, I meant 1000-5000 employees], but there are enterprise customers already, a year after the first software shipped.

Let’s not talk about contradictions, but rather examine Workday’s plans and see how well things are going.  I got an update this week from Stan Swete, Workday’s CTO, which placed Workday right on schedule for their roadmap.

Workday’s plan was to provide software services in four areas:

(1)  Human Capital Management

(2)  Resource Management

(3)  Financial Management

(4)  Revenue Management

Human Capital Management was scheduled to be available first and went into production last November.  It includes modules for organization, lifecycle, compensation, and payroll management, and workforce development.  Workday is currently using ADP for payroll management, but expects to have its own offering by 2009.

Resource Management includes Supplier Accounts, Expenses, Resource Management, and Procurement.  It is scheduled to ship in 2008.

Financial Management includes Financial Accounting and Reporting, Cash Management, Planning and Budgeting, and Management Accounting.  The Financial Accounting and Reporting and Cash Management modules are shipping now; other modules are expected to ship next year.

Revenues Management includes Customers Accounts, Orders, Billing, and Revenue Recognition.  It is scheduled to ship in 22008.

Workday is currently shipping to about 25 customers, a mixture of technology companies (about two-thirds) and non-technology companies (one-third).  It’s Workday’s strategy to target technology and services organizations, especially those in food services and professional services.  Swete seemed a little surprised but very pleased to note that they are seeing a smaller number of bigger deals than they had planned.  They already have two much larger customers, Lifetime Fitness (17,000 employees) and Chiquita (25,000). 

Generally speaking they expect to be selling to customers who are new to sophisticated ERP solutions, rather than replacing existing implementations, but they have already replaced Peoplesoft/Oracle at McKee.  Of course, they do expect to move into replacement mode later on, just as they expect to garner more enterprise customers.  (SaaS is generally becoming more appealing to larger firms as we’ll talk about in an article on SalesForce.com soon.)

Why, one might ask, are they seeing such early success?  Workday and I are in pretty violent agreement on these.

·        The ERP solutions built by traditional software vendors are expensive to buy, very expensive to implement, and expensive to upgrade (since customizations must often be entirely rewritten).  That makes them not only expensive but inflexible, a barrier to the more agile and dynamic world of the Internet.

·        Rewriting a traditional ERP solution to a SaaS platform is tough.  ISVs have to address both business model and architectural issues.  Efforts often flounder on the unwillingness of ISVs to understand that they’re writing an application that might be used by an enterprise (whom they’d like to keep on their current, more expensive software) as well as an SMB.  Applications are often written to be deliberately less appealing to enterprise customers, shooting them in the foot.  Of course, it’s a lot easier to write a great SaaS application if you have no existing business model or installed software to consider.

·        David Duffield has a great reputation.  Many customers are delighted to consider any product he’s associated with.

We asked if Workday had to sell SaaS before they could sell Workday.  It’s clear Workday’s experience mirrors what we’re seeing elsewhere:

•         Prospects are much less concerned about going to SaaS; concerns about security have stopped being the barrier they once were.

•         On the other hand, customers are much more concerned about whether they can get enough integration with their data and applications.  Workday is confident they can supply these needs.

•         Customers are also looking for the ability to use configuration of the application as a way to provide some customization.  Modern SaaS applications are designed to provide this.

We still think that Workday’s presence – and success – in the marketplace are evidence that SaaS is growing up in both maturity and in the size of customer who will find it compelling.  We expect to see more companies join it in the business software marketplace over the next few years.

IBM's Blue Cloud Platform

A while ago I noted that we would see the emergence on the web of platform vendors, each providing a set of services and APIs and each collecting an ecosystem of partners around them.  I noted that the first platform was SalesForce.com and that we were likely to see more, including IBM, Microsoft, Oracle, and Google -- and perhaps a few that don't exist yet but rise, Google-like, from the web.

Of course, for this to get started in a big, enterprise-specific, way, someone like IBM had to make a move.  They have, announcing their Blue Cloud utility service.  Now we shall have to see how, over time, this platform evolves and how richly it is surrounded with IBM and partner services (and applications). 

Nothing matters unless customers decide they like this idea, but I would be amazed if some large IBM customers (and quite a few smaller ones, up and down the mid-market) didn't find this a tantalizing proposition.  Getting IT for less from a first class vendor who will provide what you need and want and offer it with a QoS guarantee will be very appealing. 

We'd guess that a lot of IBM's efforts to enrich it's already bulging software portfolio (such as its recent announcement to acquire Cognos) are designed to provide the broader integrated offering that a web platform vendor needs.

There's a great article on the subject by Dan Gardner at ZDNet, if you'd like to read more.

SaaS as a Platform

Sitting in the audience at the Office 2.0 conference last week (you can read my musings on that at my Amy Wohl's Opinions blog), I was struck by the thought that SaaS is going to move in a new direction.

In fact, it's moving in that direction already.

There is no possibiity that customers, particularly enterprise customers, are going to agree to buy dozens of individual SaaS applications from individual very small vendors and then try to deal with their different user interfaces and APIs to say nothing of how to provide any level of interoperability.  We doubt that standards will ever provide a satisfying level of interoperabiity and integration across all applications -- there are just too many ways to make improvements and not be standard.

So we think you should look at SalesForce.com and realize that it isn't really an application vendor any more (although it does, indeed, sell a popular SaaS CRM application), but rather a platform vendor, providing a SaaS platform on top of the web.  In SalesForce.com's case theiy have chosen Model A -- they provide a platform and APIs and make the rules and build an ecosystem on top of that platform where partners gather.  They then make substantial revenue from the success of these partners.

But there's a Model B coming.  That will be vendors who provide not just the platform but a large portfolio of applications.  It might be a horizontal portfolio (perhaps with partners providing vertical market specialization) or it might be a vertical portfolio, aimed at a specific market.  In this case, the platform vendor will also own the appications, probably buying most of them from the hundreds of applications now available in the Web 2.0 marketplace.  This platform vendor could choose to provide any level of integration and interoperability he choose to support, such as a single user interface, single sign-on, and application integation.  He could still offer APIs for further refinement of his applications by business partners -- or by the customers themselves.

We don't expect too many platform vendors.  It's a role that calls for deep pockets and the ability to do significant technical work and systems operations and management.  We wouldn't be surprised to see such vendors as Google (some might argue that they're already at work), IBM (they have the SaaS infrastructure complete with rich middleware, which they offer to SaaS ISVs), and perhaps Microsoft and Oracle.  Of course, there's always the chance that one or two next generation Googles might appear.

Eventually, we might begin to see interoperability across the major SaaS platforms, but never the same level of integration as you'd see within a single platfrm vendor's portfolio. 

Reports of SaaS's Death are a Bit Premature

From time to time we have had a SaaS provider fail to provide his promised services.  Usually it's because of a server or software failure.  Occasinally it's because of an inability to handle increased volume (scale) or an act of nature (power outage).

Each time, naysayers use that as an occasion to say that SaaS can't work.  Perhaps it might be better to use it as a way to look into the failure and figure out a way to evoid it n the future.

SaaS Naysers will enjoy John Dvorak's article on the occasion of the failure of Microsoft's WGA Server, which provides authentication.   It took them most of a day to get back in business.  Personally, I wouldn't judge the SaaS business on Microsoft's performance -- they are scarcely a seasoned performer here.

But it is good advice to consider where your applications are running in the distant cloud and just how secure the system behind your delightful software might be.  Happily, the days when ASPs insisted on running their own (tiny) data centers is over and most SaaS ISVs take advantage of specialists who make sure that the services they provide can keep up and running even when a server goes down or a power outage requires transferring users to a server in another physical location.  If the application is criticl to you, asking who and how its infrastucture works is cheap insurance.