Amy Wohl's Opinions on SaaS

Changing to a Services-Based World

I've been busily working on a number of SaaS-related things in the last month:

• Judgkng a small portion of the many SaaS Solutions submitted to SIIA's software contest this year.
• Working on the next SaaScon conference (April 17-18 in Santa Clara, CA), as a member of the advisory committee. 
• Just working -- consuming and creating information -- which means using all kinds of web-based services, all of them software-based.

As I did that I came to several conclusions:

1. It's getting hard to tell that the service you're using is software.  Most people just use the service without every thinking about that.  This was clearly true for some of the contestants I was judging in the SIIA beauty contest and it's certainly true for most of the web-based services I consume (access to two bank accounts and all my credit cards, on-line shopping -- I bought most of my xmas presents on-line this year, billilng customers (many of whom use on-line purchasing systems now), ordering movies (NetFlix), and so on.  I also buy travel (hotels and air and limo reservations), send cards and gifts, buy postage, order personalized forms and stationery and on and on. 
2. It's getting hard to separate a service from its underlying software and, in many cases, the service user doesn't care.  Most would actually prefer to use the service without thinking about or "seeing" the software -- they just want to see the service and the desired result.
3. Many of the vendors who demoed their software services for me were puzzled when I called this to their attention.  They think of what they're providing as a service.  Of course, it runs on software, but they think of that as part of the plumbing.  And therein lies the tale.

So it was with great pleasure that I read Phil Wainwright's article on how Software as a Service is becoming more of a Service -- and how service providers are buying SaaS vendors to automate their offerings while at the same time smart SaaS vendors are bundling other services with their offerings.  You'll want to read his whole article.

Phil speaks at some length about the notion that the market might eventually coalesce around one (or more) platform providers or service aggregators.  I have another idea to add to the mix.  I think that for many users access to this new world of services may come through an existing relationship with a trusted service provider from the physical world -- their telco, bank, or bank card company (the source of their Visa or Mastercard).  If these compoanies become service aggregators (or, if they make aggregations of services, collected and hosted by others, available to their customers), they may convince a much larger and more mainstream market to try on-line services as a way of life.

I have little doubt that this change in how we deliver function is coming -- I think the unanswered question is not "Will this happen?" but rather from whom will customers buy their services.

Defining SaaS

Phil Wainwright has written a definitive article on what SaaS is and why it's a very broad set of services, not all equivalent in architecture, concept or value.

It was particularly timely for me because I've been spending the last week working on judging SaaS players for the Codie contest.  There's no better way to see how many different kinds of approaches fit inside the SaaS tent then looking at a dozen on-line services and asking questions.

This afternoon, one vendor felt that he owed me an explanation as to why his very definitely SaaS application did not use a browser interface, but rather a rich client.  (They have a browser version.)  Rich clients typically offer a better user experience and may offer better performance because the interface and some of the data is kept on the desktop, so there's less need to go to the server for every keystroke.  They may also offer a solution to the desire to being able to work off-line. 

There's nothing that keeps an SaaS application from using a rich client -- nothing but the fact that some "purists" require that these applications be Web 2.0 compliant which apparently means (I'm not sure I agree) that everything happens within a browser, with or without the uplift AJAX can offer the interface.

Read Phil's article and see just how stretchy this market is, and how silly it is, especially in emerging markets where we're still making up the rules, to insist that everyone must be just alike.  I especially enjoyed his comments on how SaaS vendors who know what they're doing and where this is going are using SOA.  This is a pet theory of mine, so I'm delighted to see I'm not the only one who sees this intersection between SaaS and SOA as being a key element to how software will be constructed and delivered in the future.

One Giant Computer (or Five)

There's a provocative article on ZDNet today, an intervirw witih Greg Papadopoulos, CTO of Sun, in which he asserts that the world will end up with five (or some small numberO of big computing systems, suopplying SaaS to everyone else.

Sun, of course, would like to be providing the computer hardware to the eBays, Amazons, and SalesForce.coms of that near-future world.  I'm not sure why they think they're better positioned to do this than, say, IBM.

There are also a lot of comments,  some of them silly, because they compare Greg's remarks to Tom Watson's infamous (and ancient) prediction that the world would only need five computers.  Need we explain that the five enourmous networks of computers Greg is describing, probably not in any single place, is very different than five single mainframes?  (Even IBM is likely to provide networks of computers, mainframes or otherwise, for any large computing environment.)

But Sun and ZDNet's editors do bring up some tantalizing questions:

1. Who will be the big providers of computing services to future users, not just consumers, but businesses?  I suspect that some of the big providers didn't make the ZDNet list because they were thinking more in terms of the consumer providers like Google and eBay and the Web providers like SalesForce.com and they forgot that hosting the infrastructure may be a different job than providing and supporting the applications, especially for business applications.  I'd nominate companies like IBM and the telcos for that job.
2. Who will provide the hardware and infrastructure software for these giant hosters?  Sun has nominated itself, but there are other players who may be more likely.  Again, don't forget IBM is high on the list, but HP could be a player, too.  Microsoft doesn't build HW, but it would nominate itself for infrastructure software (and some mid-market customers might prefer it).
3. Will SaaS be the only mode of computing?  I am the most enthusiastic of SaaS promoters, I confess, but I know that old habits die hard.  I would imagine that some customers will continue to keep things in-house and some applications and data sets are going to stay in-house for good reasons for a long time, if not forever. 

In computing, we add new concepts and technologies to our bag of tricks, but we rarely completely replace what went before.

Is SalesForce After "Containing" Traditional Enterprise Applications?

Lots more ink on the SalesForce.com Apex Connector announcement, including a very fine article by Phil Wainewright on SalesForce.com's rebranding adventures.

It points to a very interesting idiea:  do aggressive SaaS ecosoystems like SalesForce.com, with their ability to connect to traditional applications, make it easy for IT to stop further development to these expensive (and sometimes unpopular) environments, containing them as they are (since we can't seem to live without them) and doing future development in the SaaS environment, largely by using offerings that require nothing more than light customization?

Of course, such containment strategies would inevitably lead to the tail not just wagging the dog, but becoming the entire creature, complete with bark, as SaaS developers kept adding function from the SaaS end back.  Eventuallly, the back-end would be nothing but the data base and some policies to be moved into the SaaS application.  It does bear thinking.

Defining SaaS

Reading an article in eWeek on SaaS, SalesForce.com and Oracle's foray into SaaS, I was struck by the definitional problem.

Vendors like Oracle like to point to SalesForce.com and other net-native SaaS players and claim that customers can't get the necessary application and data integration that way. The article carefully points to a customer who begs to differ. 

I'd note that the wonder of SalesForce.com is that it wants to make everything work with its platform.  If you can live with the fact that its platform is proprietary, you can connect to your legacy apps (typically with pre-built connectors) and build customizations that upgrade with SalesForce.com upgrades via its Ajax development environment. 

And if you think that Oracle is a much bigger SaaS vendor than SalesForce.com I have a question for you:

     What is a SaaS vendor?

I thought a SaaS vendor provided a multi-tenant environment where multiple users could share servers (the users are, of course, unaware of this process, except in the cost savings it generates) and paid for usage on a subscription basis.  That is, SaaS users don't own the software they're using or the hardware it's running on.

Oracle thinks that by providing management services at the application level (surely something useful) for customers' hosted applications (customer must own the software and possibly  the hardware), it's in the SaaS business.  I don't think so.  Others (like IBM)  do this and they're quite clear that there's a difference between providing hosting, however many services are usefully provided, and providing SaaS (which IBM also offers in partnership with a number of ISVs).

We need to think about the obvious differences and try to make sure we know what we're buying -- and what vendors are offering to sell,  particularly when they're claiming apples and oranges are something other than fruit salad.

Comparing SaaS to Hosting
	SaaS	Hosting
HW	Multi-Tenant (Shared Servers)	Dedicated Servers
SW	Subscription Model, based on usage	Purchased by User Organization
Application Management	SaaS Provider	Hosting Provider
Integration with Legacy Applications	May be Provided by SaaS Platform (a la SalesForce.com) or partner	May be Provided by Hosting Provider or partner
Flexibility	High Add/Subtract Users Leave with little penalty	Low Need to buy additional HW/SW Penalties for leaving

The Move to SaaS

I've been working (with great pleasure) on next April's SaaScon program, hoping that you'll be able to join us in Santa Clara on April 17-18 and continue the conversation we began in September in San Francisco.

That means I'm even more immersed in things SaaS than usual, trying to guess what will be important nearly six months from now.

We're all certain, of course, that the ramp is up (way up) for the SaaS market -- the presence at SaaScon San Francisco of users, not just vendors talking to each other about the market to come -- was proof of that.

Now, along comes a McKinsey CIO Survey that notes that CIO's are much more enthusiastic in their plans for SaaS in the future than they were in the past, with 61% of them planning to use SaaS for some applications next year. 

The study notes that many CIO's cite a "hybrid" model, with some applications being run on SaaS platforms (and the vendor being responsible for performance as well as for the application's features) and others remaining in-house.  (These are the big corporates, of course; CIO's in McKinsey's study are in the $1B-plus revenue column; we'd expect some smaller companies to be interested in an entirely SaaS model.)

But keep in mind that there are rarely absolute changes from one model to another.  Note my comments on that subject in my Amy Wohl's Opinions blog (November 27 2006), where we use SaaS as an example of technologies that are unlikely to swing a market -- in this case software -- 100%  in a different direction.

Comoing Attractions

Coming Attractions

I've been traveling like crazy which is why I've been less than faithful to keeping up the blogs.  But I have a ton of postings I'm working on -- and a special issue of the occasional Opinions newsletter for the things that really don't fit a blog format (we'll provide a link from the blogs for those of you who really don't want to read it as a newsletter).

Look for:

• Lots on SaaS - SaaSCon and SalesForce.com's Dream
• Office 2.0 - the conference and the concept
• A book I'm working on -- How to Succeed at SaaS

The Invisible IT Market

Rick McGee, IBM's VP of SaaS, opened SaaScon with a party -- and some remarks about the state of SaaS.

He figures 2006 will be a tipping point, a year in which, in retrospect, SaaS gained irrevocable prominence in the marketplace, based in part on the support for SaaS by important vendors like SAP, Oracle, and Microsft, all of whom have SaaS initiatives now and are expected to have more in the future.

He also pointed to important acquisitions and mergers, like the purchase by ADP of SaaS HR firm Employeez, as well as to IDC's predictions of 20% per year growth into the foreseeable future.  SaaS is now estimated to be a $13 billion market.

McGee likes the catchphrase "The Invisible IT Market," because it points to how SaaS expands traditional IT functions.  It's all about removing complexity from the acquisition, implementation, management, and operation of IT.

Think of it as a Consumption model of IT with buyers willing to give up architectural decisions in return for solutions and guaranteed service.  Customers don't want to select hardware, middleware and software, they want to buy solutions. 

McGee says there are winners and losers in this model:

WINNERS

Large enterprises with failed ERP deployments

SMBs who couldn't find the right value proposition for complex software -- or the resources to implement or manage it

Providers who have a new business opportunity in SaaS

LOSERS

Traditional IT firms that can't change

(Editorial note:  I'd add customers who don't take advantage of outsourcing applications with no competitive advantage.)

With this much going on at a Sunday night party I can't wait for the fun to begin tomorrow!

Getting Ready for SaaScon

For me, part of getting ready for SaaScon was recording a Podcast interview with Paul Gillin.  You can find it here:

http://www.computerworld.com/blogs/node/3288.

While you're there you can also check out another 10 or so podcasts with other SaaScon speakers and check out the program.  If you're coming, see me at the IBM reception on Sunday night or racing around the convention center on Monday.  (Unfortunately, I'm racing on to DEMO in San Diego on Monday night so I'm going to miss Tuesdau.)  If you attend and see something wonderful, send me a note.

The SaaS Ramp Grows

Each week I look at about a dozen SaaS applications, often from companies I didn't know before.  Sometimes I come across them on the web (often referenced in other weblogs).  Sometimes they call or email, having found me by this blog -- or by the speeches I give at SaaS conferences.

We talk about how the SaaS industry is doing, what they've got to offer, and how they're intersecting the market.  I have lots of standard questions to ask -- and a few trick ones. 

But we always end up talking about the State of the SaaS market.  What's its status everyone wants to know?  How do I make that judgment?

The answer, of course, is partly experience, partly research, and partly magic.  Sometimes the answer just is "it feels right."  We all agree that SaaS is here to stay this time (as opposed to the ASP attempt of the late nineties, a kind of trial balloon for most vendors, with a few lucky exceptions).  I'm pretty sure that we are headed up the ramp (into the period of explosive growth) and the only questions are:

1. Where are we on the ramp? and
2. What's the rate at which we're accelerating?

It's always nice to have some reinforcements.  Just this morning, I read an article by Sun Federal VP Bill Vass about how SaaS is growing.  Of course, he picks some companies to highlight from a Sun point of view, mainly enterprise vendors with a Sun connection, but that's okay.  The SaaS tent has room for many players.

In fact, Enterprise SaaS is much healthier than many vendors (and analysts) think.  Although early SaaS players were just certain that the SaaS market was for the SMB customer (and, therefore, unlikely to cannibalize their enterprise efforts), enterprise customers have always found SaaS offerings and bought the ones that made sense for them. 

Now, we're seeing something more interesting going on:

• Enterprise vendors are realizing that they need an enterprise SaaS offering.
• Web-based SaaS vendors (who never had a traditional, inside the firewall offering) are heading straight for the enterprise market with no apologies.  I offer as an example Projity, a project management software product that is compatible with Microsoft's Project, but claims to be far easier to use.  It also doesn't require the set-up time typical for customer-based software in this category.  Yet it's fully featured (which is usually the trade-off). 
• Of course, there are more offerings than ever for the SMB market, by both traditional vendors, trying to expand into a new market, and web-based vendors, choosing the new platform.  You'll find everything from accounting and human resource software to specialty offerings like business intelligence for retail chains. (May we introduce you to SeaTab who can do what the big boys do for one-tenth the price?)

Next week we're headed for SaaScon (San Francisco, September 25-26) and we'll come back with more companies to talk about and more news of a growing story.