Cloud computing for SMBs

Cloud this, Cloud that – the word cloud is clearly an overhyped word and reminds me of the beginning of the hype around hosted models and ASPs (application service providers) in the late 90s and the term SAAS today.  Anyway, as I look at announcement after announcement released about cloud computing platforms, one thing is pretty clear to me from an investment perspective.  First, I am not going to invest in the next hot cloud computing infrastructure service that will compete against Amazon, Rackspace, Microsoft, and every other large tech vendor in the world.  This is suicide and far from capital efficient.  Secondly, while everyone looks in the consumer space, I want to look at how software companies can deploy new enterprise-based applications in the cloud, particularly for small/medium sized businesses.  In other words, show me the arms merchants with a recurring revenue model and frictionless sale and I will definitely be interested.

Some of the companies that fit this parameter include Rightscale (founded by Thorsten von Eicken, a cofounder of former portfolio company GoToMyPC) and one that I am looking at in the email archiving and compliance space which has a number of OEM partners reselling its service. Rightscale is an on-ramp to Amazon EC2 and other clouds and provides automate systems management.  It kind of reminds me of a next generation Tivoli or Openview.  The beauty is that the whole sales cycle is quite frictionless and all web-based which means an oppotunity to scale quickly.  There are a number of other recent players I have seen including one for BI in the cloud (not exactly sure what the killer app here is yet) and many others.  Of course the trick here is not to get enamored with the word "cloud" but to really understand the business problem that is being solved and why leveraging a cloud computing platform offers better economics, scale, and competitive advantages.  As I dig deeper into some of these companies, it is clear to me that software purpose-built from the ground up to live in a cloud has a huge advantage since it is hard to retrofit off-the-shelf software to leverage all of the benefits offered by Amazon, Rackspace, and the like.  Secondly, many of the better companies have built some slick tools and services to solve difficult problems like how to make customers feel like they have their own privated, dedicated systems while still keeping costs low.  Finally, from a go-to-market perspective, a number of the companies I have spoken with have not gotten the question of whether or not they could scale as they quickly point to their backend provider and move to the next objection.  So, if you have an application targeted at the SMB market that is taking advantage of cloud economics, please feel free to contact me.

Selling to large enterprises costs big dollars no matter how frictionless your sale is

I have written a number of times about frictionless sales and how on-demand companies have a huge opportunity to reduce their sales and marketing costs and subsequently scale their business more efficiently.  Here is an excerpt from a prior post:

Frictionless sales means reducing the pain for customers to adopt and use a service/product and consequently reducing the cost of sales and marketing to get a customer and generate revenue.  As I mention in an earlier post, "The less friction you have in your sales and delivery model, the easier it is to scale. The easier it is to scale the faster and more efficiently you can grow." The lowest friction sale can be a user clicking on a web page and the content owner getting paid for it.  The highest friction sale is spending lots of money on marketing and trade shows and having a large, direct sales force of expensive reps pounding the pavement for months trying to close a large deal with an enterprise customer.  Follow that with a 3 month implementation process to get the customer happy.  There are various grades of friction between these two extreme points like open source business models, software as a service, and reseller/OEM-type models as other forms of packaging and delivering a product/service.  And of course, each of these models requires a different methodology and way of marketing and selling to a customer. Ultimately what you want is sales leverage where every $1 you spend on sales and marketing equals multiples of that in terms of revenue.

The perception that it is much easier to scale definitely holds true if you are selling to consumers, small businesses, and workgroups within large organizations.  However, it seems that many public on-demand vendors are feeling the pressure to deliver growth and ultimately need to feed the revenue machine by going after larger customers.  And what many companies are learning is that no matter how on-demand your software is, if you are selling to huge enterprises you are going to have to spend huge dollars in sales and marketing.  Sales cycles are long no matter how you slice it and even if there is no massive hardware and software installation, many large companies want to have their service customized and integrated, even lightly, with other systems.  in other words, many of these high flying on-demand vendors are starting to look more like the old software companies they are trying to replace.  As per a Wall Street Journal article today, it seems that many of these public on-demand companies are finding out the hard way that no matter how frictionless your sales process is, the bigger the company you sell to, the more it is going to cost you. 

There is nothing to install, so workers can start using online software without the aid of the tech department. That makes it easier for companies that sell online software to get into a business than their on-premises competitors.

Seizing on this, investors bought into online-software companies in a big way. During the first 10 months of 2007, shares of 15 online-software companies tracked by Thomas Weisel Partners increased in value 61%. Since then, however, these companies have lost about a third of their value.

Wall Street has realized that it isn’t enough to simply offer online software—you have to have a sales strategy that can make your offering a corporate standard. It is possible to get individuals, project teams or small businesses to buy online software through word-of-mouth marketing, but it is hard to make money from these groups—at least the kind of money necessary to become a billion-dollar company.

In order to get there, they can’t operate like an Internet start-up, letting their technology spread virally as end users hear about it. They need to sell to the same executives and information-technology professionals who made purchasing decisions before online software was an option. Businesses have a lot riding on the decision to use one product or another. And while having pockets of workers advocate for a particular piece of software is a plus, the execs who sign the big checks still want to see demos, vet the seller and do all the things they have always done when they buy software.

So if you are an on-demand vendor, either stick to your focus of scaling with SMBs and consumers which requires a completely different sales and marketing approach more rooted in traditional online budgets and telesales or be prepared to spend some real dollars if you truly want to go after the big guys. 

Need homework help – try

I got an email from George Cigale, CEO of, yesterday to check out the New York Times article on his company’s service (full disclosure-my fund is an investor in and my partner Dan is on the board).  The article, "On Demand, On Time and for a Fee, an Army of Tutors Appears," highlights Michelle Slatalla’s experience with’s service where her two daughters were able to get instant homework help by going to, logging in, and clicking on their grade level and subject matter to find a qualified instructor.  Thankfully Michelle had a pretty positive experience as we have had time to hone the service and continue to find great tutors as we currently complete thousands of sessions each day.  Anyway, next time your child asks you for help on Algebra, you may want to visit and try another method.  From a thematic point of view, this is another example of how companies can leverage the power of the Internet to offer an on-demand service leveraging a distributed and free agent workforce.  I just love those types of models!

Do you believe in the Red Shift theory?

The first time I heard the term Red Shift was from my portfolio company, Greenplum.  Greenplum has used red shift to characterize the nature of the existing database market where exponential data growth driven by network computing and internet applications have outstripped the capacity of existing mainstream vendors.  Hence, a new approach was needed (our database software running on commodity clusters) which would allow companies to load and query terabytes of data at 10-100x performance and scale over traditional vendors.  Ok-enough of the sales pitch.  Moving on, it is clear that red shift data requirements are only a fraction of what’s necessary to meet this exponential growth as it will put tremendous strain on the existing IT infrastructure consuming ever-increasing amounts of CPU cycles, energy, storage, and more.  If you want to read more about this red shift theory, I suggest checking out a great article by Richard Martin in Information Week.  Martin neatly summarizes Red Shift as defined by Sun’s Greg Papadopoulos to be:

  • Red Shift refers to companies experiencing exponential growth in demand for raw computing power
  • Red-shift companies tend to be Web 2.0 focused like YouTube and MySpace, or big financial, energy, or pharmaceutical companies
  • Those companies, Sun CTO Greg Papadopoulos says, will experience similarly high levels of growth in users, revenue, etc., while blue-shift companies will grow relative to GDP
  • Along with the cost of powering and cooling in-house data centers, the red shift is driving a surge in utility computing and software as a service

Based on my experience with both consumer Internet and companies selling infrastructure, I can say that this all feels right to me.  It is also no wonder that virtualization which helps IT consolidate servers and increase capacity utilization and utility computing are top of mind again.  Think about Amazon’s S3 and EC2 which I have written about before as utility storage and processing for the masses. I am definitely meeting more and more startups which are starting to offload some of their computing requirements to these services.  And of course, while Greg Papadopoulos is pushing this vision of the red shift, he has put Sun in a great spot to execute on this with new platforms and ways of keeping up with this exponential demand.  The only question as Mark Anderson points out in the article is not if there will be an exponential increase in servers sold but how many of them will be Sun servers running Solaris versus open systems.  Either way, it looks like the stock market has been voting with its feet as Sun has been performing quite well as of late.  And as a VC whether you believe in the red shift or not, we would all like to find companies experiencing hypergrowth where one of the main uses of capital will be for scaling the infrastructure to meet demand.  That is what I call a good problem to have.

Do it yourself (DIY) in the enterprise (continued)

Last year I wrote about the newfound productivity of the prosumer, the consumer who is bringing technologies into the workplace in a DIY (do it yourself) fashion.  If IT can’t or won’t get something done, users can simply check the Internet for the latest web-based service or software download to help them solve their problem.  In this month’s CIO Magazine which landed on my desk somehow, the cover article is titled "Users Who Know Too Much and the CIOs Who Fear Them."  The subtitle is "They’re smart, productive and using IT you didn’t provide.  How to manage the modern user."  I think we are at the very beginning stages now of IT’s recognition that the world is changing and like Jeff Nolan says the balance on the continuum of systems and people should move more towards a people-centric vision of technology.  What do the people want and how do we provide them the ability to get things done while at the same time balancing our need to keep a safe and secure environment?  Sometimes these issues are directly competing with one another.  It is still quite early in the CIO’s recognition of a user-centric IT world but the fact that CIO magazine is focusing on this means that it is becoming more critical to its readers.

Over the next couple of years, it will be interesting to watch how the battle between top-down, conservative IT and bottom-up DIY employees gets resolved.  IT wants control, security, and compliance while users just want to get things done.  As the article advocates, the smart CIOs will figure out how to balance the needs of their users and the role of IT.

This will require CIOs to reexamine the way they relate to users and to come to terms with the fact that their IT department will no longer be the exclusive provider of technology within an organization.  This, says Smith (Gartner analyst) is the only way to stay relevant and responsive.  CIOs who ignore the benefits of consumer IT, who wage war against the shadow IT department, will be viewed as obstructionist, not to mention out of touch.  And once that happens, they will be ignored and any semblance of control will fly out the window.

Whether or not CIOs get it, does not really concern me as the nature of sales for many of these DIY apps and services should be focused around the end user vs. centralized IT.  Given this, the sale should be much different, less costly, and with much less friction.  If a user wants to track his sales force productivity, they can go online and sign up for or create their own through a SugarCRM download.  There is no on-site installation as the web helps deliver the product efficiently.  From a sales perspective, as these companies grow over time, much of their sales can be done over the telephone or through a WebX or GoToMeeting session with only the large accounts reserved for an expensive direct sales rep.  Given this bottom-up, web-based model of selling and delivering software, it will be interesting to see how the incumbent vendors respond.  For example will users adopt a collaboarion platform from IBM that IT has pushed down on them or would it be better for CIOs to figure out what their workers are using and standardize on that?  Does this mean that the smarter incumbent software vendors look to buy startups that already have bottom-up traction versus building their technology from scratch?  As I was writing this post, I just noticed that IBM just signed a deal to pipe Google gadgets through its Websphere portal. 

"These sites are not just valuable to consumers. Businesses want the same content. Why would we keep these two universes separate?" said Larry Bowden, vice president of the IBM Lotus division for portals and Web services.

While Internet access, and thereby Google Gadgets, may be easily available to consumers, many businesses restrict access to the latest Web applications for security reasons, to make network management easier and to limit employee distractions.

By allowing Google Gadgets to work within its WebSphere Portal, IBM is making it easier for companies to give employees access to popular Web applications while keeping control over how they are used. Companies can decide which Google Gadgets they can see.

"The end user decides: We no longer need to go off and call a technician," Bowden said. "The power has been turned over to the people who know best. You know best."

It looks like IBM gets it and is trying to help its IT customers strike the delicate balance between control and giving users what they want.  All I can say is that the intersection of the enterprise and the web-based platform will be an interesting space to watch over the next few years and it is clearly heating up.

GOffice – what’s the big deal?

It is not a surprise that Google officially launched Google Apps Premier which is a bundled package of their hosted offerings for word processing, spreadsheets, email, calendaring, and instant messaging.  I wrote about this in the fall of 2004 when Adam Bosworth joined Google from Microsoft and wrote a lengthy blog post on the web-based platform.  Google has clearly been executing on this vision over the last two years, but I do not see this as a Microsoft killer.  While I am huge fan of web-based software and data in the cloud, there is one big problem – you always need to be connected.  For the last two weeks I have been living in a web-based world as I had to send my laptop back for service.  While I could do everything I needed to do, I must admit I was about 60% as productive as usual.  This lack of productivity partly came from clicking and waiting in my web-based Exchange offering and partly due to lots of travel which meant I could do absolutely nothing on the airplane.  What I see Google Apps doing is breaking the market into two segments – those who want to easily share and collaborate information with others in a lightweight manner and the power users who live, eat, and breathe in their productivity applications.  I certainly see myself using Google Spreadsheets to post some information on my blog but it will be a long time, before I even think about replacing my desktop productivity applications. In the meantime all of this is great for consumers as competition is forcing Microsoft to rethink their whole application strategy by incorporating a SAAS component into most of their offerings.  I can only assume that Microsoft will get better at this and make it easier for their users to work online and offline in a seamless manner.  In my web-based world, disconnected applications with an online component will rule.  Let’s see what the Adobe Apollo platform brings to the world later this year.

Small business startup kit for 2007 – mostly free!

A friend of mine called me the other day to ask for advice on what services (email, voice, apps) he should use to run his business with the caveat being that he wanted to spend as little upfront capital as possible and also have minimal ongoing maintenance headaches.  As I started thinking about his question, I remember what it was like setting up our office in 1998 and the headaches and cost of buying a Nortel phone system and phones and hiring a Microsoft networking expert to get our office set up for file sharing, back up, and email.  What a nightmare!  What was even worse was that we had to have this guy come in at least once a month for general maintenance.  So when we moved in the beginning of 2004, I vowed to outsource as much as possible.  In the end, here is what we did:

1. Exchange server – – pay monthly based on number of mailboxes and mailbox size and eliminates the headache of ongoing maintenance and backup.  also can add mobile devices like Blackberry, Good-enabled, etc. and easily provision without cap x.
2. Voice-outsourced VOIP, we have a direct pipe to a local provider, we leased some Cisco phones, and once again no upfront cap x and lots of great functionality, we pay a base monthly fee for unlimited calling.
3. Security – we bought some Cisco gear but have a small IT firm as our managed service provider remotely monitoring and updating the software with the latest patches and release.
4. Connectivity = We are networked internally on Windows and have a shared drive where we can access files.  In addition, we have a VPN for remote access to this share drive.
5. Productivity – Microsoft Office

Going back to my friend’s question, if I could set up my office now, here is what I would do:

1. Exchange server – I hate exchange and I would bail on this as soon as I can.  Instead, I would get all of my email and calendaring functionality through Google Apps for your domain – it is free and provides 2 gb of email, integrated calendaring with your email, chat and simple voice chat, and an ability to create simple web pages.  Yes this is basic but it is easy.  In addition, I expect a lot more to be offered once Jotspot is integrated along with some of the other basic Google Office apps such as word processing and spreadsheet functionality.  My one big beef which is holding me back right now is the lack of simple syncing with wireless devices.  There are some apps you can plug in to sync Google calendar but they still need some work.
2. Voice – if I want something more robust I would get a Fonality PBXtra for $995.  If you choose to go the really simple route, the PC-only VOIP providers of today have come a long way since 2004. I am partial to Gizmo Project (wait for our new version which will be accessible through a browser – also, full disclosure, I am on the board) but Skype and other services can once again offer you pretty decent voice communications and functionality like the ability to buy your own phone number, call forwarding, and dual ringing on your computer or cell phone.
3. Security – not as important if your files are hosted offline and backed up remotely (try xdrive which is free for 5 gb or (free for 1gb). 
4. Connectivity – a simple wifi network in the office can get you simple file sharing without an IT professional’s help.  If you want to collaborate with remote workers, you can use a wiki like Jotspot  or Socialtext or some of the shared storage services I mention above.  As far as remote acccess, no VPN is needed as a simple GoToMyPc account ($19.95 per pc per month) or LogMeIn (free for base functionality) can get you the access that you need without the headaches and upfront cost of a VPN.
5. Productivity-Microsoft Office but the online apps are getting better and in fact for collaboration or sharing would consider Google Office apps like spreadsheets and writely

What is amazing to me is how far and how fast we have come during the last 2 years.  The big difference is that the functionality is even better and so is the price – mostly free!  Given this, I wonder what we will be looking at 2 years from now?  Yes, one problem is that all of the solutions I list above are dependent on having an Internet connection.  What if I am not online and need access to my calendar or some office documents?  Since this is a pretty clear problem, my prediction for 2007 is that online apps get better offline client like functionality.  Maybe it will be the new Adobe Apollo platform that makes it happen for us?  What is clear is that one of the benefits of SAAS for developers is that they don’t have to code in multiple platforms.  Once you start diving into the murky world of multiple operating systems and developing clients for Windows, Mac, and Linux, it can quickly become quite messy and resource intensive.  That is why I also see 2007 as the year that offline apps become big as the Apollo platform is released and allows web developers to build an application on one platform that can be deployed cross operating system.  Also keep an eye out for Microsoft’s WPF/e (windows presentation framework everywhere see an earlier post for more info on wpf).  This is a big deal and will help SAAS-based apps continue its upward trajectory and spread from consumers to SMBs and even further into enterprises.  As an example, take a look at Jeff Nolan’s recent post about how frustrated he is with Exchange and how GMail provides a nice alternative.  With the ability to get my whole office set up with a few clicks, it is no wonder that Microsoft is running scared and embracing SAAS rather than fighting it.

Is the bar lower for a tech IPO?

I am not sure if you saw the news, but recently filed for an IPO to raise up to $50 million. On the book is Thomas Weisel Partners, William Blair, Needham, Pacific Crest, and Wachovia. According to the S-1 filing: is a leading provider of on-demand compensation management solutions. Our comprehensive on-demand software applications are integrated with our proprietary data sets to automate the essential elements of our customers’ compensation management processes….

In addition to our on-demand enterprise software offerings, we also provide a series of applications through our website, which allows us to deliver salary management comparison and analysis tools to individuals and small businesses on a cost-effective, real-time basis…

We offer our solutions principally on an annual or multi-year subscription basis. Our direct sales group markets and sells our solutions primarily using the telephone and web-based demonstrations. From the introduction of our solutions in 2000 through September 30, 2006, our enterprise subscriber base has grown to approximately 1,500 companies who spend from $2,000 to more than $100,000 annually, including companies such as Wal-Mart, Home Depot, Procter & Gamble, Merrill Lynch, UPS and Cisco Systems. We also sell to both individual consumers and smaller businesses through our website.

From April 2001 through June 30, 2006, we achieved 21 consecutive quarters of revenue growth. During the years ended March 31, 2004, 2005 and 2006, we achieved positive operating cash flows of $0.3 million, $0.9 million and $1.8 million, respectively, and used $0.7 million of cash in the three months ended June 30, 2006. During these periods, we have consistently incurred operating losses, including $0.8 million for 2004, $1.9 million for 2005, $3.0 million for 2006 and $0.8 million for the three months ended June 30, 2006. As of June 30, 2006, we had an accumulated deficit of $21.8 million.

I would usually put IPO filings in the nonevent category but as I dug deeper into the company and financial performance, it did raise some interesting questions for me.  First and foremost, the traditional rule of thumb that most investment bankers have quoted me in the last couple of years was that in order to go public a company needs to have an annual run-rate of $40-50mm of revenue and a couple quarters of profitability.  While the numbers are strong (read the S-1 here), they are not close to those metrics.  In fact, during the last 3 fiscal years for the company, it did $6.4mm, $10mm, and then $15mm in revenue.  The trailing twelve month number is closer to $20mm in revenue.  While slightly cash flow positive, the company is not GAAP profitable.  So the natural question for me is to ask whether or not the barrier for a private company to go public is much lower today and whether or not this will signal an ongoing trend in the future.  This is obviously relevant for a number of reasons.  Outside of a few outliers, most of the returns generated for VCs have been from M&A transactions.  If the IPO markets open up again, it would give investors and entrepreneurs another option to create value.  Using a back of the napkin analysis, most companies sell about 20% of their stock to the public, so one could assume that is valued at around $200mm pre-money implying a 10x multiple on trailing twelve month revenue.  I must say that sounds quite appealing.  Anyway, we should all watch this company as it goes through its paces because if it does well, it could open the door for plenty of other companies like it.  There must clearly be an appetite from the institutional money managers who are looking for more upside from rapidly growing small cap companies.  By the way, one other interesting point about is that is an on-demand application play with some web-based advertising thrown into the mix.  It is also mostly a subscription-based business which means it has a highly predictable revenue stream which is great for forecasting future performance.  Finally, the company only raised $5mm of VC dollars so it is highly capital efficient.  If you read from the S-1 above, most of the sales are generated through the telephone or through web-based demos, all of the traits for a nice frictionless sale and great business model.

Utility computing for the web and startups

There is a great BusinessWeek article outlining Amazon’s ambition to be a utility for web businesses.  This reminds me of a conversation I had last month with a founder and former CTO of one our of our prior portfolio companies who said his goal was to have a highly successful SAAS play with 1 operations guy instead of 20.  When I asked him how he would do it, he quite simply said Amazon – Amazon EC2 (Elastic Compute Cloud) and Amazon S3 for storage.  Sure, I had heard about this before when Amazon launched it during the summer, but what really got me thinking was that here was a guy who had been there and done it – scaled a SAAS business to incredible numbers and he had been playing around with Amazon’s infrastructure and was willing to offload a majority of his new startup’s business on the Amazon infrastructure.  When we talk about the commoditization of technology and how cheap it is to launch a new business on the web, we think open source and commodity servers. Now think about being able to launch a new web-based business and only paying for what you use.  If it takes you awhile to scale you don’t have to burn alot of capital upfront and only pay for minimal usage.  If you are hugely successful, then you don’t get caught with your pants down because you have the opportunity to quickly load a few more virtualized images on the Amazon EC2 infrastructure and pay more for that usage – bandwidth, storage, and compute time.  Think about it –  the upfront cost of starting a new web-based business if you went the Amazon route (when it is ready for primetime) has been driven down another order of magnitude as you can get started with little to no capital expenditures. The numbers are pretty incredible too – $0.15c per GB per month for storage or $150 per terabyte per month, $0.20c per GB for bandwidth, and the use of a pretty standard server (1.7Ghz x86 processor, 1.75GB of RAM, 160GB of local disk, and 250Mb/s of network bandwidth) for $0.10c per hour or $72 per server per month.  Not too bad when you think that you can scale up or scale down pretty easily.  It will be interesting to see how many startups look to use the Amazon infrastructure after it gets more publicized at the Web 2.0 conference.  As a startup, your job is to allocate your scarce resources as efficiently as possible – time and money.  If you can stretch either of these and give your company more of an opportunity to hit critical milestones or get better product out the door, then it is a huge win for you to spend your dollars on making that happen, rather than on capital equipment.

Revolutionary technology with evolutionary implementation

I was riding on the train this morning and was talking to a friend about one of my fund’s portfolio companies.  She mentioned that the management team had done a great job during a recent sales presentation because instead of going for the "rip and replace" strategy, they went with the "co-exist" philosophy.  Too often, entrepreneurs can get too enamored with their own technology and forget that the customer may not need every feature that you are offering today.  In fact, while revolutionary technology and vision is great, what the customer may want is an evolutionary approach to implementation.  What I am talking about here is reducing the friction in your sales process (See an earlier post on frictionless sales).  Convincing a customer that your technology or product can coexist with an existing investment is a much lower barrier to sales than convincing them to "rip and replace" or "forklift upgrade" a significant prior investment.  The sales prospect will have a hard enough time buying a product/service from an unproven startup, let alone ripping out an existing investment from a safe choice, a much larger public vendor.  Once you land the customer, you will always have the chance to expand your footprint.  That is why I continue to be enamored with SAAS and downloadable software because I believe that it is inherently a more efficient and cost effective way of selling and delivering a product or service.  Granted, most of the initial target market opportunities will be the small/medium business market but I still firmly believe that this market is untapped and offers great upside.