The On Demand Journal

Talk, Talk, Talk

The past couple of weeks I've had the pleasure of talking to a wide group of people. On September 15, I was invited by Josephine Cheng, IBM Fellow and Vice President of IBM Almaden Research Center to give a talk to 100+ PhDs. I, of course, took the opportunity to tell them seven research projects I'd have them work on. For those of you who work at IBM the talk was recorded. Check in with Merridy at merridyh@almaden.ibm.com if you want to know more. A few days later on September 17, I joined Mike Maples, Jr, Managing Partner, Maples Investments  Ravi Mhatre, Managing Director, Lightspeed Venture Partners and Peter Sonsini, Partner, NEA on a panel entitled "SaaS 2.0: A New Wave of Innovation".  Prashant Shah was the event host at the TIE Silicon Valley sponsored event.  The word is it will soon be online at the TIE website.  Finally, on October 1, Feyzi Fatehi, Chairman of the Orange County Chapter of the Technology Council of Southern California invited me to give a breakfast talk. Entitled "The End of Software - The Beginning of the Cloud" I took them thru the seven business models and ten key things any software company needs to think about when moving to the cloud.  They did a great job of promoting the event and getting nearly 100 people to show up even at that early hour.

Stanford Class Starts

We've started the fifth year of the class on software powered businesses. The lineup of speakers is at cs309a.stanford.edu so it promises to be a very good year. If you're interested in seeing the lectures (available the next day). Sign up to watch. I did the opening day lecture. Slides are available here.

CloudWorld Keynote Panel

CloudWorld's 2009 Keynote Panel will be held on August 12 and is entitled "Assessing the Real Market Opportunities and Obstacles for Making Cloud Computing Mainstream". The panel has been told the discussion will be kicked off with some questions. As preparation I thought I'd write a response to each of these questions...

Where do you think the cloud computing market sits today?

As with all markets it's appropriate to partition the market. Consider a 2x2 matrix where cloud computing can be applied first to Applications and Infrastructure; and then to Business and Consumer uses.  In Consumer Applications there is no doubt this is the dominant mode of computing -- Google, Amazon, eBay, Facebook, Expedia, etc. need I say more.  For Business Applications we're beyond the Early Adopter phase but not quite to Main Street.  That being said there are at least eleven application software companies that have gone public since 2000. The so-called Elite Eleven includes: Salesforce.com, RightNow Technologies, Concur, Taleo, Kenexa, Omniture, DealerTrack, Constant Contact, Vocus, Success Factors and Netsuite.

While Applications are well on their way, Consumer Infrastructure and Business Infrastructure cloud computing are still in the very early phases. Amazon with their work in compute and storage as a service has led the way, but their offerings are only a few years old.  We're clearly going to see much much more from many other players. In many ways IBM's appointment of Erich Clementi, reporting to CEO Sam Palmisano earlier this year shows how dramatic the shift might ultimately be.

What aspects of the cloud computing market gets you most excited?

I'm going to focus my comments on Business Infrastructure cloud computing.  Innovations in this area promise to deliver the long-tail of thousands of specialized applications to power businesses. Whether that's managing a poultry business, automating a funeral home or managing fuel in a gas station we've only scratched the surface in business applications of computing. Yahoo list 5000+ applications in over 45 sectors Google list 4000+  applications in over 35 sectors . These lists give us a sense of what is possible. But that's looking in the rear-view mirror.  With more cloud infrastructure services, more open source, more cloud applications services the number of business applications could easily be 100,000.

The other aspect of the market is the advent of private cloud services. If a cloud service means a repetitive, standardized infrastructure then the difference between private and public clouds is who gets to specify the cloud. In a public cloud someone else gets to do it (e.g., Amazon) in a private cloud the purchaser of the services makes the specification. So the idea that an AS/400/iSeries private cloud, or a private cloud for media and entertainment apps makes perfect sense. 

What aspects make you most concerned?

The only debate is the rate. But the rate can be slowed if the industry does not continue to invest in improving the overall level of quality, reliability and security. Here are two specific challenges. In a increasingly complex environment the fight to achieve high reliability and availability will only grow. We all have our own personal stores of how even losing DSL for an hour caused all sorts of havoc in our household. While perfection might be difficult to achieve the challenge should instead focus on how to get applications and infrastructure to have fast, predictable recovery time.  If every service, when it fails, would be able to tell you accurately when it will return then in time engineering can cut the number by 50% and so on, and so on... The second challenge is how to select and take a security patch from R&D into production - again in a known, repeatable time.  Change management has always been the bane of reliability so developing engineering techniques to move any important change from development to production will only increase in importance.

What events or evolutionary advancements need to occur to for the market to continue to grow and become truly mainstream?

The move from Early Adopters to Main Street will require many of the classic requirements to first attract the early adopters (which has begun) and then find the Bowling Alley (s) where there is a dramatic adoption around some particular application.  That was Visicalc for the PC; ATM applications for Tandem computers.. we'll have to see what it will be for business infrastructure cloud computing.

What are the biggest potential impediments to market growth and expansion?

Simply, the cost of sales and marketing.  In business computing we know how to sell and market items at $1,000,000 that are imprecise (see Oracle, SAP sales reps if you want to know how). We also know how to sell items that are 10 dollars and precise (buying a book on Amazon.com) the challenge is how to sell things that are $10,000 and imprecise. Salesforce spends 50% of their revenue on sales and marketing; Webex spent 35% of revenue on sales and marketing.  And you have to reflect on the thought that selling web conferencing still requires this much money. The problem is that while web conferencing might be simple to many of us, trying to explain how it makes a difference to your business, how to set it up and invite people is not simple.  You only need think about the popularity of the Geek Squad to realize that even in simple consumer electronics the distance between what I need and how the products accomplish this is not simple.  The industry will have to develop technology, processes and whole new businesses to climb this hill if the technology will see it's true potential.

Cloudbook

The new book, Cloud: Seven Clear Business Models, is now available for purchase. As important, we also launch Cloudbook.  At the core of Cloudbook is the new book, but in addition all of the contributors mentioned in the book (and more) have their notorious content indexed into the book. As such, today with nearly 200 contributors and 600+ inteviews, blogs, videos, articles, webinars, Cloudbook is the single best source of information on the people and products shaping cloud computing.

Some recent stats:

• On track to hit 100,000 page views by early June. The site has already attracted over 2000 unique visitors.
• Seven cloudTV shows have been produced including ones featuring Sun, AT&T, and NTT. The cloudTV shows have generated over 70 leads.
• One company did 2 cloudTV shows. In the US: 33 registered: 15 attended, in EMEA: 6 registered: 2 attended. The company did NO email marketing.
• cloudTV Guide is sent to 4000 people, who have all expressed interest in SaaS and/or Cloud Computing
• 2000+ people are following the Twitter feed.  Each new Contributor, each new piece of content and each cloudTV show provide an opportunity to tweet to our group of followers.
• Cloudbook is optimized for natural search; we expect it will begin to show up early in any Google search. Two weeks after going live over 15% of the visits are coming from Google natural search results.

Stanford Autumn 2008

We've nearly completed the fourth year of CS309A: Software as a Service. We've had a great list of speakers. I'm going to use this blog entry to be able to have the students gain access to some of the powerpoint and to summarize the quarter.

We began the year with Jonathan Schwartz. Jonathan spent most of his time talking about the value of free. Which in this context is talking about how free products reduce the cost of customer acquisition. Depending on who you talk to for companies like Amazon or Priceline the cost can be $30-40. While companies like Vonage and salesforce could be closer to $400. And we all know that traditional software companies spend 25-40% of their dollars on sales and marketing.

Tony Hsieh, founder and CEO of Zappos delivered the 30 September lecture. Many of us were unaware that Zappos is poised to be a $1B company and is already the leader in online shoe sales. Tony talked a lot about the culture of the company - but one quote stood out: "Customer service is not a department'. Zappos takes a completely different attitude about customer service and it has changed how it's people and infrastucture are organized in a fundamental way.  If you're interested contact them and get a copy of the yearly company culture book and you'll understand what I mean. We all look forward to the day of Zappos Air.

Marc Chardon, CEO of Blackbaud (full disclosure: I am on the Board) was our next lecturer. He talked a lot about similarities and differences between Blackbaud's software and conventional CRM software. Perhaps it was best described as software directed more at knowing more about the donor, the customer and less about the state of the transaction. With Blackbaud's move to deliver software in Model Four and the recent acquisition of Model Six companies (eTapestry and Kintera) this quarter the subscription revenues will be twice the conventional license revenue.

Dave Girourd, President Google Enterprise spoke on the 14th of October. For those that are students and interested in a copy of his presentation let me know and I'll email it to you. I was surprised at how little Dave talked about enterprise search. But he did highlight the scale of the operations and investments Google is making in enterprise productivity apps - in particular the size and cost structure of Google Mail  As they gather more early adopters it will certainly be interesting to see how deeply Google penetrates the enterprise with more business oriented apps - not that we don't all use Google search everyday in the office.

CEO of Netsuite Zach Nelson was up next. Netsuite went public last December and has been growing ever since. While traditional back office software is not simple to implement, Netsuite has been making great progress in the SaaS model. They clearly see the trend to specialization or verticalization and are taking steps to allow CEMLIs (customizations unique to an industry) be transferable from one customer to another. For more details see his presentation.

John Lervik, founder and CEO of FAST (recently acquired by Microsoft) seemed to be happy to have the resources of Microsoft to bear on what has been a life's work.  I asked him - today - the SQL business is probably close to $15B (Oracle, IBM, Microsoft...), but search is probably only $1.5B - Why?  I think his answer was: Google did us a great service in making search important (but that only happened over the past 6 years) - but they also made everyone see search as a search bar. John sees search much more as a new platform to build search-based applications - apps driven by search not SQL. For more detail see his presentation.

Charles Phillips spoke last week. He's the co-President of Oracle and has been one of the driving forces of Oracle's M&A activities. They've completed nearly 50 acquisitions in 50 months - so Charles spent a lot of time talking about the methodology of doing this. It might have been lost on the students how difficult this is. He spent some time on - on demand - and how the @Customer model (Model 4) was doing well - but he made the point that large companies aren't going to move everything tomorrow - particularly when they have a large scale operation.

Yesterday my old friend Javier Barerras came from Mexico City to deliver a lecture on how technology is reshaping the hotel business. It's not widely known but Grupo Posadas the largest hotel chain in Latin America generates 50% of their profits from non-hotel (and more specifically software) -based businesses. This includes AltiusPar, which is the moral equivalent of Sabre for the hotel industry. Javier also pointed us to CitizenM, a new hotel chain powered by technology and a "real CRM" system that knows what you like to eat, drink, watch and listen to - and customizes your hotel experience based on information. For more detail see his presentation.

Stanford - CS309A - Autumn 2008

For the fourth consecutive year we're running the CS309A class - called "Software as a Service" - or as some have said - "people Tim would like to listen to". The class schedule contains the list of lecturers and a little bit of an introduction. Jonathan Schwartz, for the 4th year in a row started out the series with a great lecture on "Free Business Models". For those taking the class for credit there is a requirement to write a paper. Last years class papers are available for download.

Seven

The new book, Seven, is now available for purchase. Seven is a follow on book to The End of Software, which predicted the rise of software delivered as a service. Now that this transition is happening Seven provides a framework for the producers and consumers of software to take the next steps. Seven is written with a chapter focused on every functional area: R&D, Operations, Sales, Marketing, Finance HR. In addition, You'll also find lots of campfire stories from people who did these jobs at leading companies such as eBay, Salesforce.com, Webex, Netsuite, Oracle and many others.

Click here if you'd like to purchase a copy.  Also, if you're interested, an early version of an active book is available at www.softwareseven.com. It's under construction, so you should come back often and see how it's changed. If you have any comments please feel free to email me.

Mabel and the Cloud

Jim Morris, who is Dean of Carnegie Mellon West, invited me to speak at Sofcon 08.  The conference was on the Mobile Future and speakers included David Pogue; columnist for The New York Times, and Bob Iannucci; Senior Vice President, Chief Technology Officer, Nokia.  When Jim asked me to speak - I told him I didn't know much about mobile, other than of course, I have a mobile phone. But nontheless he convinced me to say yes.  I took the opportunity to reflect on mobile computing and also learn from some great people. You'll soon be able to see the video of the presentation but here is the gist of my thesis.

Mobile computing (in particular the massive success of the iPhone) is happening. Google says more search is coming from the iPhone than any other mobile device; Bank of America has said 30% of their mobile banking transactions are coming from an iPhone. The ability to deliver the full web experience coupled with a flat data rate plan have been two particularly important key success factors. Special thanks to Bob Borchers, who is Senior Director, Product Marketing for iPhone for educating me. For those that haven't seen his rap Hit Me On My iPhone it's worth checking out. I spent some time with Michael De La Cruz, SVP CRM, SAP and his personal opinion is that it won't be long until half of the 30Million SAP users will be accessing SAP thru mobile devices.

Of course, none of this is particularly new news. But what is interesting to think about is - what's the implication on "the cloud".  An iPhone is about 1/10th as powerful as a Mac laptop, which is by the way 1/10th as powerful as a Mac server.  So if we're going to get rid of our laptops and move to mobile devices - you could make the argument - the cloud, the computing infrastructure needs to be 10x larger.  Beyond the simple math - there are four reasons why demand for services will be larger.

1. Mabel will work in a service business.  Most enterprise/business software was developed for transaction processing (debits, credits, general ledger, parts inventory); and around manufacturing companies (consider the history of SAP). But today 80% of the US workforce works in a service business: financial services, health care, software, etc -- they all don't manufacture anything but instead work with information.  Enterprise 2.0, a term coined by Andrew McAfee at Harvard Business School, is "is the use of emergent social software platforms within companies, or between companies and their partners or customers.". Many companies are experimenting with FaceBook, executive blogging, wikis as a means of discovering this emergent social software platform. But we all know email as a means of running service businesses is dead. I spent some time talking about Openwater Networks and how a service network blends social networking and enterprise search- so that information and people can work together. If you want to think of a simple example from the Consumer Internet consider Flickr.  Flickr without the people is a photo database.  Flickr without the information would be a chat room --- together -- well you get the point.

2. Mabel Knows where she is. This is a big difference that mobile devices know where they are. Bob Borchers said that this feature has been slow to take off because of fears of privacy loss.  I spent some time talking with Anne Bezancon, Founder & CEO, 1020 PlaceCast about the technology they've just unveiled. 1020 was founded to bring information about place into the ability to target ads that are more relevant to the mobile user. 1020 argues that it's more than latitude-longitude. Place is also context - not only where, but when. One good example is if you know you're at the Oracle Arena and it's during a Warriors game the ad you see might be much different than if you were there the next night during an Amy Winehouse concert.  All of this can be done without knowing who you are - just the context.  And remember most of this computing will need to occur in The Cloud.

3. Mabel can talk.  While you can put your laptop up to your ear; or wear a headset. A mobile device was built for talking.  Many advances have taken place in speech recognition and synthesis.  Thanks to Steve Pollock co-founder and EVP at Tuvox I had them listen to simple demo of the technology at work for AMC movie theatres. If your thinking about The Cloud, you'll clearly see how much more computing is required to deliver this kind of service.

4. Mabel has a camera. As a somewhat amusing (although apparently in the mobile circles, often used) example of the usage of a camera I walked the audience through a system that is being used in Japan that allows a consumer in a grocery story thru scanning a 2D bar code to know the complete history of a head of cabbage.

My argument is that all of these reasons are going to lead to a Cloud that is much much bigger. I left them with an interesting statistic.  On an average day the NYSE processes about 10 million transactions, Visa processes about 100M transaction, Google processes about 200M transactions (and uses 500,000 servers to do this).  There are 3,000 million phones in the world. If everybody scans a head of cabbage once during the day how big will The Cloud need to be?
 

– 

 

 

Sloan Wars

A few weeks ago Jim Roth, who is attending the 10-month Stanford Sloan Program invited me to give a lecture to his MBA class. It's curious that MIT also has a Sloan program -- are they the same Sloan?  Anyway Jim took my CS309A class last quarter. All of the essays he and others in CS309A wrote were put together in a document called "The Beginning of Software 2008". You can download a PDF for free, or buy the printed version at cost.

I took the opportunity to deliver a lecture loosely based on a forthcoming book called "Seven". I heard Werner Vogels talked about it on a conference call last week. Seven is a follow on book to "The End of Software". But rather than try to convince you that software delivered as a service is going to happen the book takes the point of view that any software business needs to figure out how they are going to move to on demand delivery -- the only question is how.  The book is structured around seven software business models (from traditional to the Internet) and has a chapter dedicated to the major functional areas - sales, marketing, finance, R&D, operations, HR.  Each of the chapters contains campfire stories from people who had the experiences including Aaron Ross, who ran lead generation for Salesforce & Lynn Reedy who ran all of operations and R&D for eBay.  More on the book when it comes out, which should be in May.

Pebble Beach 2.0

The folks at CSC invited me to speak at an executive conference they had at Pebble Beach a few weeks ago. So as you can imagine it wasn't difficult to say yes. While Pebble Beach was beautiful and the event as high class as any from the good old days I was even more fascinated that the topic was going to be Enterprise 2.0.  As with many of us I thought this was an overused buzzword that was beginning to fade; but instead discovered there was a lot of interest and much discussion.

Don Tapscott, author of Wikinomics, led of Day 1. While he made many points perhaps the one that stuck with me was the changing demographics of the workforce.  And furthermore the implications on businesses of what it means to attract and retain this next generation.  Perhaps the best story was what email meant to the Facebook generation.  As the story goes one of Don's younger colleagues wanted to meet a certain young lady. Don said - "Why don't you send her an email?" -- the response was -- "Why would I do that I don't want to marry her.."  Clearly every business represented in the room has to think about the implications of the changing workforce.

Day 2 had both Doug Neal, a research fellow at CSC and Chris Turner, CTO at Unilever articulating how they were seeing enterprises becoming next generation enterprises. In Chris's case he talked a lot about his plans for Unilever. Clearly P&Gs usage of mass collaboration to drive innovation is influencing everyone in consumer packaged goods.  I was the closing act of the play -- if you're interested in the .ppt - it's here.  I did take some time to have the group use wiki technology to build an online book as just a glimpse into the implications of Enterprise 2.0