Amy Wohl's Opinions on SaaS

Chrome: Consider how you use computing now

Before we have all the expected rants pro and anti Google's Chrome OS, perhaps we should first ask the question, "How do we compute, now and soon?"

For those of us who spend much or most (let's bet honest here) of our time on the web, a new interface and a new way to write applications and interact with the web might be a powerful enabler to users and developers alike.  But for those who mainly interact with desk-top applications or data on company systems, Chrome may be nothing but an interesting but largely irrelevant idea.

The question, of course, is how many of each of us are there and what does the trend line look like?  Greatly influencing this, of course, is the surge of cloud computing and the rush to put as many applications as possible into the cloud.  Like all reactions to new technologies, this rush will cool down after a bit and we will sort out which kinds of applications work best on-line and which continue to work best in a desk top or company server mode. 

Of course, if we include consumers in these calculations (including ourselves in our consumer mode), the off-line usage soars.  It's hard to sit in your home office looking something up on your home computer when all the action is in the family room around the big TV screen.  Already, I walk around my home (an eight-room condo the size of a fairly large house) with a netbook and have more netbooks and laptops strategically stashed, all connected with a wireless network.  I'm getting ready to give them a little server to share for shared data like address books, to do lists, and lists of music sites.

Many users are developing that "good enough" mentality (with regard to selecting cloud-based products over richer-featured and better supported desktop products) and if Chrome offers them features that are hard to replicate in land-based computers, that may be enough to push things in a cloud-based direction, for many users.

But remember that everyone will not go rushing off to the cloud for every application.  Certainly not now and perhaps not ever.

The Seriousness of SaaS in the Enterprise

I just came across a great column by Phil Wainewright about how SAP's executives seem to be stepping on each others toes, commenting on their SaaS products and their opinions on the value of SaaS (and Salesforce, of course).  To be fair, you have to read yourself through the thread and find the Talkback from SAP exec Vishal Sikka and Phil's remarks about that comment.

I'd like to make some remarks of my own.

SAP is remarkably late to the SaaS market and has chosen a unique path which I don't think will work:  they want to keep their customers continuing to invest in very expensive SAP ERP applications while buying bits of SaaS applications for small outliers, on SAP's schedule.  For some SAP customers who are not at all ready to move (and that could be for excellent reasons and for a very long time), that could seem fine, assuming that they don't notice that competitors aren't busy building excellent and cheaper SAP-compatible applications for them.  With no need to protect SAP's revenues, these competitors will, of course, include software categories and features which SAP will choose to label as "not suitable for SaaS."

Many of SAP's defenders (or Salesforce.com's attackers; there is some overlap), insist that SaaS is not suitable for enterprise users.  Apparently they are unaware of the fact that hundreds of enterprises are using SaaS right now and intend to use more of it in the future. 

It is unlikely that any particular way of doing computing will every be the ONLY way of doing things.  We still use mainframe computers because they turn out to be the best way to do some large-scale tasks or to run software that is too expensive or too complex to rewrite.  Some of today's enterprise software is likely to be around for a very long time.

But to the extent that it makes sense, companies, and not just SMBs, are looking at the SaaS market as a way of less expensively, more quickly, and with greater flexibility offering computing and applications to their users.  It is a Pandora's Box that once opened will not be closed.

Pricing SaaS Products

As the array of SaaS products broadens and more traditional software ISVs (who will continue to offer existing products) enter the market, pricintg SaaS products continues to be a puzzle.

There is a tension between the right price from a marketing perspective, one that will attract the optimum number of users, and the right price from a profitability perspective.

• The right marketing price for SaaS products is often thought, at least initially, to be "free."  This is done artificially through extended free betas, free trials, and free "basic" versions of the SaaS offering, all intended to remove price as a barrier to initial usage.  In every case, the ISV is hoping that as many users as possible (in some cases all users) will move on to a paid version.  Conversion rates vary, depending on the product, the type of customer, and the competition, from less than 1% to about 25%.  For a few business products they are higher.
• The right price from a profitability perspective is a calculated number, taking into account the cost of operations, including ongoing user support, further development, and marketing.  It is likely to be less than the price of a traditional software product (which makes for hard decisions for existing ISVs in rationalizing their new SaaS prices with their existing products). 

Of course, SaaS products aren't priced in a void.  They must be priced in an existing marketplace and compete with existing and potential products.  First entrants have the privilege of establishing the price point if they have a compelling product and it is sufficiently differentiated that products in other SaaS markets have not already set a pricing ceiling (or floor).

Often (too often) we see great products that have few SaaS competitors pricing their products way too low because they don't understand the value of their product to their customers.  If their is no SaaS competitor and your traditional SW competitor is charging $100,000, $10,000 is not a great price, it's a dumb price.  The customer would be happy to pay $25,000 or $35,000 for a well-featured and supported SaaS product.  Discussions with analysts and potential customers, testing for the appropriate price point in advance of product launch, can substantially affect what's available for the product price (and how soon the company might reach break-even).

On June 25th, I'll be speaking at a Webinar on Profitability:  SaaS Versus On-Premise Solutions, presented by Aspire Systems, an Indian firm that works with ISVs.  Please consider joining me.  Here's some more information.

SaaS is being increasingly adopted and is even considered a savior model in this economy due to the zero upfront CAPEX for buyers. Still the question remains if SaaS is a profitable model for solution providers? Regardless of the longer sales cycles, is the on-premise solution providing faster break-even?

Some of the topics being discussed are:

	Cost factors – developmental, operational and marketing costs
	Revenue factors – sales cycle, pricing models, etc.
	Benefits/drawbacks of both the business models
	Will a hybrid pricing/delivery model provide better benefits?

  Panelists:

Amy Wohl of Wohl Associates

Alex Ginger, Director of Active Operations Management International, a SaaS provider

at SaaS: Computing in the Cloud”. It is both an overview of the SaaS market and a series of recommendations for ISVs, platform vendors, and customers who want to participate in it. The centerpiece of the book is 22 interviews Amy conducted with these market participants, illustrating the state of the SaaS market and its direction.
Alex Ginger Product Development Director, Active Operations Management

Registration is FREE!

Space is limited
Reserve your Webinar seat

June 25 11 am ET:

Head in the Clouds

 

In case you're wondering what I'm up to, I'm reading everything I can find about clouds and interviewing lots of companies (vendors and a few users) about their cloud offerings and their plans.

The hype level is beyond amazing.  I know that there is a point in any successful technology where it is used to label everything, because it's so good at attracting positive attention, but I think we're close to reaching the saturation level here.  I hope so.  If everything is a cloud, then it gets very hard to see anything!

The confusion level is high.  Everyone who is looking at clouds as a buyer, a seller, or a commentator (press, analyst, consultant...you name it) has their own definitions.  I have a rule here.  If you want to communicate with somone about clouds on a serious level, exchange definitions first.  Otherwise, you may just be talking past each other.

In my research (I feel like a graduate student on a quest) I have found a lot of great stuff and some real trash.  I'd like to point you to some interesting stuff while I continue my research project.  I'm nearing the end of this phase and I'm going to tell you what I think soon, but I can't resist passing on a few goodies:

There is no agreement on what a private cloud is.  It might be a new architecture for a data center, it might be a new way of architecting part of a large enterprise's data center when certain workloads are present, and it might be  privately owned public cloud.  I'll talk about all of those in my second cloud computing article (the first is on IBM's cloud computing portfolio), but in the meantime, you might enjoy an interesting glimpse into one problem in determining whether an enterprise should consider a private cloud:  what kind of workloads that private cloud would process.  Try this article from Kenneth Oestreich at The Fountainhead on questions nobody's asking re cloud applications.

Or you could read this article from Treb Ryan at OpSource (he's definitely an interested party since he sells platforms to ISVs that could be considered public clouds) about why a private cloud isn't really a cloud.   I can see where he's coming from, but I just spent an hour on the phone with a CIO who's building his own cloud (from a thousand servers and a lot of open source software) and he can make a very polished argument for why he's got the benefits of a cloud.  I think the jury's out on this one and a lot depends on (as Ken Oestreich suspects) what you're going to run on that cloud.

So back to the books -- actually to another hundred Internet posts.  I'll have more than hints for you by next week.  

McKinsey got it wrong: Cloud Computing is for Enterprises

This week I spent several days at the Uptime Institute's Symposium 2009, a conference where facilities management meets IT.  The focus for this year's conference was on Green.  I spoke several times, moderatig a panel on outsourcing versus cloud computing, and giving a mini-keynote on SaaS and Cloud Computing.

At that conference, McKinsey announced a report on Cloud computing, claiming that the economics only made sense for small and medium sized companies but not for enterprises with their own data centers.  The analysis then compared the cost of using Amazon's service with the cost of a typical data center.  McKinsey also assunmed that the firm would move all of its computing to the cloud -- we don't know of any organizations that plan to do that.

As you might guess lots of conversation and controvery ensued.

I guess I'd say they don't understand.  Cloud computing may not be cheaper if all you're looking at is a comparison of hardware and an assumption of people costs.  The value of cloud computing lies elsewhere:

• In the flexibility of being able to gain immediate access to additional computing (or to shrink your system when you don't need it).
• In the difference in Time to Market for new business opportunities.
• In the additional value (not included in the McKinsey study) provided in clouds that offer to manage the hardware (via systems software and other offerings) or to provide applications (SaaS). There is a brilliant explication of this by Balakrishna Narasimhan which I recommend.

Many large companies are already taking advantage of SaaS -- Workday, for example, has clients for its HR and Financials with more than 20,000 employees.  Public clouds can be ideal places for large companies to use specific applications or obtain additional capacity.  Private clouds (which may not be clouds at all, in the sense that the organization may own the hardware and software rather than temporarily rent its use) are intended primarily for large enterprises.  I am certain that schemes which will make these private clouds into capital-free propositions (if not available on a month-to-month basis) is available to any customer who requires it. 

I expect clouds to be popular under any circumstances.  Given our current financial probems, I expect them to appeal to a broader audience, especially to large enterprises. 

Clouds Multiplying, 2

NOTE:  This post was intended for my SaaS blog but it somehow ended up on my other blog, so we're posting it here, too, with the thought that not everyone reads both blogs.

I am in the midst of writing two big articles on Cloud Computing -- too big to post on a blog -- and I can't quite get them finished because every day brings new revelations.  Everyone wants to be part of the cloud as a platform provider or a cloud application or a user.

Do we know where this will end?  We can probably project its reasonable ending.  Too many vendors without the right experience or resources will enter the market and not all of them will succeed.  Too many ISVs will try to field applications and only some of them will be real winners.  And some customers will learn (once again) that nothing is the solution to everything.  Of course, we don't know when this will happen or who the winners and losers will be, so in the mean time it's going to be a lot of fun trying to sort it out.

In the meantime, I'm going to lots of webinars and conferences trying to meet new players and sniff the sea breeze to see where we are on the curve. 

A conference I'm looking forward to is Under the Radar's Cloud Computing eventon April 24 in Mountain View -- a single day filled with products and demos and sharp questions.  I encourage you to join me; I've been there before and it's definitely worth the trip.

I'll also be speaking about SaaS and Cloud Computing at the Uptime Institute's Symposium 2009 in New York City on April 16 and at IBM's IMPACT conference in Las Vegas on May 6.

In the meantime, enjoy the clouds -- in this case they are not an omen of bad weather, but rather part of a sunny new computing future. 

Surviving in Tough Times

All of us expect SaaS to be a bright spot in a dark economic picture, but Treb Ryan of OpSource points out that even SaaS vendors will have to work hard to survive in a tough market where customers are disappearing and cutting back.

I especially like his suggestion of looking for customers in new verticals where you haven't sold before - he suggested healthcare, the federal government, and bankruptcy lawyers! 

He doesn't like hybridizing your business model -- that is, moving away from the pure SaaS model of delivering on-line in order to get another customer.  I wouldn't change anything for one more customer -- it's just a distraction and you won't actually make a profit that way. 

On the other hand, I don't see anything wrong with testing your market to see if customers would buy from you if you had an appliance offering, for example.  (That means loading your software on a server which attaches inside the customer'r firewall; it's optional whether the SaaS ISV manages the software or whether it becomes the customer's responsibility to manage updates and patches.)

I also think the OEM market can be an excellent "tough times" opportunity.  Look for a partner who also needs a new revenue source and who has customer for your SaaS product (or some part of it).  Let the OEM do the marketing; the two of you can negotiate who does the hosting, operational support, and so forth, at what price.

Tough times call for smart tactics.

Who Reads Page 6 of Your White Paper?

Speaking to Keith Thompson of Vitrium Systems, I discovered that whether a white paper is a success is no longer a matter of how many downloads it gets.  That, of course, is largely a matter of good PR, good location on the vendor's web site (and good navigation), and a hot topic.  Whether anyone reads the thing -- much less whether they read the whole thing is quite another matter.

Vitrium first produced Protected pdf, a product that kept pdfs from being changed or copied.  It appealed to  higher ed.

But customers asked for a product that could give them metrics on pdf documents – who read them, what was read, the amount of time spent on a page.  Docmetrics is that product.

When people get pdfs today they either have to register (which they either refuse to do, abandoning their attempt to get the white paper or they give false info) or they simply click on "free" white papers and give no info – in either  case the pdf owner knows nothing about the reader.

Docmetrics inserts a survey into the pdf (which can be mandatory or not) usually at page 2 or 3.  I suspect my colleague Jared Spool, a UI expert, would say this was one of his seducible moments.  The reader has started to engage with the white paper and giving its owner a little information for the rest of its content seems like a good exchange.

This allows the pdf owners to ask a variety of questions in a customized form which may be required or optional.  Docmetrics places a cookie into instrumented documents so if a document is passed along the survey will be presented to additional readers.  Even if they skip the survey, information about their reading habits will be included in the Docmetrics reports.

Reports include both leads as well as information on pages read or skipped, how much time was spent on each page, and so forth.  (As a white paper writer, this sounds both very useful and a little scary!)

Vitrium has been marketing Docmetrics for six months and has over 100 clients.  About 55% of readers using pdfs instrumented with Docmetrics submit information.

The product has a wide range of pricing starting at $25 per month for a small firm with one pdf and up to 5 leads per month.  Volume pricing is available for large enterprises with thousands of documents (or thousands of readers).  Pricing can be less expensive if the vendor wants to capture reading information but not reader IDs (such as name, email address, and so forth) for leads.

Docmetrics is integrated with Salesforce.com so that a company that uses SalesForce for lead management will have survey data entered directly into its CRM files and salesmen notified of new leads.  Vitrium plans to offer integration with other CRM products.

Docmetrics is hosted at OpSource.  While they intend to offer the software as a service, it is also designed and supported so that they can offer it within the firewall, should a large customer prefer to run it themselves.

At this time, they appear to be the only vendor with this type of product; they feel the barrier to entry is high since it requires intimate knowledge of Adobe pdf, Flash, and each Adobe Reader version as it comes into the market.

 

 

Taxonomies for Clouds

Just as we are trying to think through the architecture of clouds, others are trying to develop a Taxonomy. We found this discussion on alltop.com.  Of course this plays directly into the Semantics section of the survey we are running now (see below if you haven’t participated yet).

You might be interested not only in the brief article on a meeting in California to discuss Cloud Taxonomy (not too successful it sounds like), but more interested in the comments to the post which go to the point of trying to create standards too early in a new market.

Do note:

(1)    Big players can push smaller players to use their specifications as de facto standards (that’s the story of the Windows market or of Walmart, as one commentator remarks), but that isn’t necessarily good for users.

(2)    In a web-based community, where much of the ultimate use of the platform and its evolution will be determined by users and their own activities, great user interfaces may be the most important standard.

(3)    Our survey asked users whether they would demand that clouds be interoperable.  We think that it is unlikely that any one user or organization will be able to get all the services it wants from a single cloud.  That cries out for being able to interconnect clouds easily and use services across clouds.  Will we get that?  Not easily I think.

Lots to think about.

- - - - - - - - - -

We are running a survey on Cloud Computing.
Join the fun by clicking here
http://www.surveymonkey.com/s.aspx?sm=rv2UloiI01rprHf6javzug_3d_3d
and answering the questions.
We'll comment on the collected wisdom and send our comments back to you.

- - - - - - - - -

Diversity in Applications

I am always amazed to see how diverse SaaS applications can be.

Yesterday I got to talk with some folks from ICA, a spin-off and partner with Vanderbilt University Medical Center, who have built a clinical application of informatics.

The original research was done at Vanderbilt,  ICA did the commercialization, and Vanderbilt does continuing R&D.

The technology sits in a layer on top of a hospital or community medical system's existing medical record systems, allowing them to preserve their investment.  Each entity (hospital, physician, clinic, health center) puts its information in a "vault" and a web portal presents the aggregated information about any patient to a practicing clinician in one place in a single common format.  It's magic.

Currently ICA has 5 clients in various stages of development inclulding:

• An enterprise installation at Basset Healthcare in Coooperstown, NY including 4 hospitals and over 30 clinics spread over an area as large as Delaware.   They use the ICA application as their primary application for seeing patients (instead of a chart) while continuing to post through their source systems, often via dictation.   95% of the physicians were up and running within 6 weeks.  ICA says the system requires minimal training.
• Lourdes Hospital in Kentucky is a pilot where physician training took only 45 minutes!
• Health Information Exchange in Montana (Calispell and four feeder hospitals and 40 feeder clinics) is in an area where collaboration and collegial discussion would otherwise be very difficult.
• Mid South EHealth Alliance in Memphis, TN – Includes all their emergency rooms and their primary care sites to make sure emergency rooms are used properly.  The idea is to facilitate patients to primary care where appropriate and the focus is on Medicaid, uninsured, and underinsured populations.

The ICA system can take in and aggregate literally any kind of data -- Word, pdf, data file, images.

ICA is hoping that the $20 billion in the Economic Recovery Act that is intended to improve medical records can be focused not just on administrative record keeping and insurance, but also on the kind of patient care that they have been working on, especially since they can use their informatics systems to sort the data in any way, providing physicians with information by disease, by drug, or by any other useful pattern.  In this way, they could see how other patients with COPD, for example, had been treated, and what the outcomes had been.

Only about 20% of physicians have electronic medical records today and most of them exist in silos, unable to be accessed or aggregated with other records on the same patient -- as you no doubt have guessed when you have to provide the same information to a physician in the same medical center.

ICA is hoping that the idea of secure access from any PC, secure messaging to the staff, and workflow tools with disease dashboards will be a powerful incentive to move forward.

The ICA application can be hosted by any single user in a medical partnership for the other members or it can be hosted by ICA.

I remain fascinated at the number of SaaS applications we are seeing that are not for broad horizontal markets, but rather for specific vertical applications.  I suspect we are only at the very beginning.

  - - - - - - - - -

We are running a survey on Cloud Computing.
Join the fun by clicking here
http://www.surveymonkey.com/s.aspx?sm=rv2UloiI01rprHf6javzug_3d_3d
and answering the questions.
We'll comment on the collected wisdom and send our comments back to you.

- - - - - - - - - -