And that's helpful how?

Dialog boxes.  Oh how I hate dialog boxes - particularly when there's so little dialogue actually involved!  Anyone use Outlook?  You've probably seen this gem:
Outlookdialog

So the software is smart enough to see that I entered a comma delimited list, but it's not smart enough to replace them with semicolons for me?  Rather than telling me to click cancel and do it myself and press send again, how about just asking if I would like it done for me?  Better yet, why ask at all?  Software is supposed to make hard things easy.  This kind of message is why people hate software: it really makes Outlook sound like a prick.  "You've separated the items in your list the common and accepted way in the English language.  However, for no apparent reason, I want it another way.  Go away and don't come back until you've done it my way."  And, while I'm at it, if you don't have any suggestions for me, don't tell me to select an address to use!  
"Welcome to my restaurant; what would you like for dinner - but before you tell me, I want you to know we're out of it."
"Well, I'd like the roast chicken, and I'd like a salad for a starter, and a glass of the pinot noir..."
"Sorry, you've ordered your main before the entrée.  Please come back in again and place your order in the correct order."
"Seriously?"
"Yes, and pinot doesn't go with chicken..." 

Comments [0]

Thoughts for this week's "young IT" crowd

I’m delighted to be invited to the Australian Computer Society’s Young IT Professionals meeting this week to give a short preso and participate in a panel discussion on “cloud” computing.  For events on this topic in the past I’ve generally spent most of my time talking about the different models – IaaS, PaaS, SaaS – and their benefits and the players in each category.  But I’m going to take a leap of faith in the “youth” of the audience and presume that they will already know this.  Instead, I was asking myself, “what does ‘cloud’ mean to someone who’s just starting off in a career in technology?”  Cloud isn’t a new technology (it’s just a new way of delivering computing resources), but it’s probably the most important shift in the technology field in the past thirty years, particularly for those just getting started in the field.  Why?  Cloud’s democratisation of IT.


The only real requirements to bringing a software vision to life today are an idea and a willingness to work hard.  There’s no need to buy servers or any other hardware, there’s no need to have a distribution channel for getting media to users, there’s not even a need for IT as everything from email to collaboration to finance and sales systems are all delivered in a SaaS model.  Geography, too, is unimportant as it’s more likely than not to have a team spread out around the globe using github and Skype as much as any physical meeting room was ever used.  And as we vendors of “cloud” computing rush to win the hearts and minds of developers, the offers to customers are very generous!  (just look at  programs like BizSpark and DreamSpark from Microsoft for example).  In the past – as recently as 5-10 years past – most good ideas lingered in the minds of developers who worked at a bank or other enterprise somewhere; maybe, occasionally, a domain name was registered in a burst of enthusiasm, but the actual leap was rarely made.  Today, I find it difficult to come up with a good answer to the question, “what’s stopping you?”


Of course, this comes at a price: startup clutter.  It’s so easy now for anyone to just start throwing some code up in the cloud and calling it a business that we’re beginning to miss the filtering function formerly played by the logistical difficulties of starting a software business.    When you did need capital investment and when you did have some tangible risk, you tended to think more deeply about the idea (at least that’s the theory!) and those who actually made the leap had to be discovered from within a much less cluttered crowd.  Now I can go from wild idea over lunch to working prototype that night – and that doesn’t leave much time in between to validate the concept with real customers and a real market.


So, I think this will be the gist of my talk: today’s entrepreneurs have a tremendous opportunity to move against relatively few tangible obstacles, but that makes it all the more important that they validate the idea first (and thoroughly).  Luckily, the lack of obstacles also means that you can fail fast, too.  And we all know that a few failures are key to fertilising success – but that’s a topic for another Young IT lecture…

Comments [0]

Cloud Insurance?

We’ve seen a number of high profile outages recently in some IaaS and PaaS offerings.   All vendors (I hope) have various service level agreements (SLA’s) that try to explain the guarantee that they make about the availability of the platform.  But what happens when they fail?  Most offer relief to the customer in the form of a service credit (and, maddeningly, require the customer to proactively make a claim for credit!).  The cloud vendor in essence says, “you built your business on our platform, and we failed to deliver the availability that we offered.  But don’t worry, we’ll make it up by giving you back a credit equal to 10% of your bill.”  Well, that’s nice.  Isn’t that just like a restaurant saying, “You got a nasty case of salmonella from your dinner, but we’ll make it up to you with this coupon good for 10% of your next dinner here.”  Hmmmm, thanks, but I’ll pass! 

So what can be done instead?  Certainly a PaaS vendor can’t be held responsible for the business impact to all of their customers caused by an outage.  For some customers it may have little or no impact, while to others it may be devastating!  Bad things happen, and some get hurt more than others: that’s how it goes with natural disasters and other accidents, right?  Sure, but there’s something that you can do to ameliorate the impact of those bad things – you can buy insurance!         

Insurance is simple:  I trade my risk of loss to you in exchange for a payment.  If the risk never materialises (e.g. the tornado never strikes), you keep the full payment.  If it does, you compensate me (financially, anyway).  So why can’t we do this for the cloud?  I can purchase various levels of coverage based on the potential damage that would be done to my business in the event of a PaaS or IaaS vendor’s failure to meet their SLA.  Like in automobile insurance, the premiums would vary depending on the amount of coverage (the value of the vehicle), the likelihood of a problem (a sports car vs. a Volvo), and the reputation of the vendor (I’ll omit an analogy here to avoid a flame war ;) 

Some of the benefits:

Clearly there’s a benefit for the business who wants to move to the cloud, but is nervous about the change.  Similarly to how the US FDIC (deposit insurance corporation) helped get Americans to keep their money in a bank instead of under their mattress, there’s comfort in knowing that someone’s going to cover you if anything goes wrong.  Cloud insurance could greatly increase the rate of cloud adoption.

There’s also a benefit for the IaaS and PaaS vendors.  A third-party insurance market could help push vendors towards better and better levels of service as they try to gain customers who will shop for the lowest insurance rates.  It’s like the “five-star safety ratings” that automobile manufactures covet!  A safer, more reliable product will be cheaper to insure, thus more desirable to customers.  It would be much more effective in changing behaviour by the vendors than simply making them give out a 10% credit (not even a 10% refund!)

Finally, it could be a great business for an ambitious insurance organisation!  Insurers are always looking for the next policy to write.  Where there’s fear, there’s an insurance policy to write: and we all know that the Cloud certainly causes fear and trembling in many corporate IT managers today!

So, who's up for it?  Or maybe it's already out there?  Thoughts?

Comments [1]

My YOW Cloud Night slides

The YOW Cloud Night was primarily a panel discussion, I did share a couple of slides about Azure.  By request, here they are...

Click here to download:
YOWAzure.pdf (794 KB)
(download)

The Sydney stop of the event is next week (June 14 - visit http://yownightsydcloudjune.eventbrite.com/) so please come and join us if you're nearby!  The discussions with Clayton Brown from Salesforce.com, James Tyrrell from Cloudfoundry, and Lachlan Hardy from Ninefold are great, and only get better with meaty audience questions and participation!  (and hopefully Amazon's Simone Brunozzi can stick around for the panel next time!)

Comments [0]

We need some Venture Appitalists!

There’s a disconnect between the business models for custom software development of the past and the new age of SaaS and PaaS.  It goes a little something like this:

Frame1

The "exectuive" thinks that this whole PaaS thing sounds like the bees knees.  He can finally get that project he's been wanting done without waiting for next year's capital budget and the inevitable delays of buying new servers and all that.  Get the SI to build it and pay per user per month.

Frame2

The SI conducts his business as usual - charging for work done.  The customer can get utility pricing from the vendor for ongoing use of the app on the platform, but development is a separate expense.

Frame3

Hillarity ensues as the executive realises that software as a service is software use as a service.  Bespoke development doesn't have a new model to go with the new platforms and billing models of SaaS and PaaS. 

Yes, we can have a debate over whether or not there should be any bespoke application development in the future, but for now corporate customers expect Whopper software (have it your way).  And thanks to outstanding PaaS marketing (ahem), they also want it “in the cloud.”  So how do they/we/you deal with the colliding business models of traditional bespoke development and the utility model of SaaS?

 

Imagine a project that’s going to bid in at $50K for custom development work but it will run corporate-wide with 1000 users.  That could be spread out for $4200/month covering the development cost over a year.  That’s only $4.20 per user per month.  But who’s putting the $50K up front?  The venture appitalist, of course!

 

A venture appitalist funds the upfront development costs of bespoke applications that will be delivered in a SaaS model in exchange for a commitment by the customer to use the service for a minimum period.  Why would anyone want to be a venture appitalist?  Well profit, naturally!  Profit comes from negotiating both the monthly utility cost to cover the development costs and duration of commitment.  In the above case, for example, the VA (see, I’m already making a meme) could pin $5 per user per month premium onto the regular per-user cost, and require the end customer to commit to a 24 month contract.  The first year would yield about $10K (.80 per month above the straight repayment costs) and the second year yields $60K.  140% return.  OK, not bad.  But even more attractive is taking that originally bespoke solution and selling it to other companies.  Chances are if one metal fabrication company in Duluth needs a custom inventory system, some metal fabrication company in Dunkirk does too!  As long as there’s company+1, it’s no longer bespoke, it’s just SaaS.   

 

So who pays?

1.       Independent investors who front the initial expense of development in exchange for a share of the ongoing revenue from the utility of the application (pure “venture appitalists”).  This could be a whole new type of venture – not investing in a company, but investing in an app.

2.       PaaS providers who create a device that lets organisations attach some initial development expense to their platform subscriptions (e.g. the PaaS vendors themselves are the venture appitalists)

3.       System Integrators who create their own system to amortise the expense of the development into the utility pricing model.  This is probably the easiest (but least interesting): and it simply means that the SI’s give up their “net 30” mindset. 

4.       The end customers who just need to adjust their expectations to allow a certain up-front investment if they want a custom solution, and then the utility model benefits kick-in once the application is live and running (e.g. no ongoing administrative costs, or IT infrastructure and maintenance costs).

Only options 1 and 2 are really “venture appitalism.”  For a PaaS vendor, it could be an interesting side business and would obviously help drive traffic on their platform.  For a traditional VC, it could be an interesting new business.

 

 

Comments [0]

Your chance to vote...for the Azure contest finalists

The Windows Azure Australia “My Concept Rules” contest has been an incredible success and the winner will be decided tomorrow June 2 at REMIX11!  We have decided to add a “people’s choice” category to the contest, too.  Use the following FaceBook page to vote http://www.facebook.com/msautech?sk=app_112552622166410 and read about the four finalists.  They’re all really outstanding and I wish we could give them all the big prize!  Stay tuned, though…we’ll be doing more with the finalists and other good entries in the coming few months.

Comments [0]

YOW! Nights

Great title for this panel discussion "Hey! You! Get Onto Our Cloud!"  I'll be representing Azure along with speakers from Amazon Web Services, Ninefold, CloudFoundry, Salesforce and others.  A great knock-down, drag-out night of PaaS carnage: guaranteed fun for all!  


Melbourne - June 7th

http://www.yownightmelcloudjune.eventbrite.com

 

Brisbane - June 8th

http://www.yownightbriscloudjune.eventbrite.com

 

Sydney - June 14th

http://www.yownightsydcloudjune.eventbrite.com

Comments [0]

P, as in Platform!

I had an excellent session yesterday with a distinguished (ahem) panel of judges to try to cull some of the entries in our Australian Windows Azure "My Concept Rules" contest down to a short-list of finalists.  Now, I'll say right at the outset that the entries were almost all excellent.  There was not an insignificant amount of doubt expressed by some folks around here that there wouldn't be any good ideas, so I'm glad to prove them all wrong!  The entreprenurial spirit is coursing throughout the land, and I can't wait to reveal the finalists and award a winner (happening, incidentally, at next week's hive of tech hotness, REMIX11 (http://www.noisetosignal.com.au/remix/).  But I'm going to leave the specifics about the entries aside for now (I'll post some details about the winner and other finalists after next week) to muse on something that I noticed in reviewing the entries: very few people have any idea what PaaS give you.

Depressing thought, certainly, for someone charged with growing our Azure business!  But I'm afraid it's true.  We had a field on the contest application asking what about the application idea made specific use of any of the attributes of the platform and, with all but a few exceptions, the answers focused solely on simple elasticity: my app will start out small, but will grow to be used by millions once word gets out there and Azure will allow us to scale to meet that demand.  A fine attribute of Azure and other PaaS offerings, surely.  But hardly the only benefit!  Almost all of the entries looked just like any web application that could have been built in 2003, only hosted in Microsoft's datacentres instead of with some traditional hosting company.  Eliminating the up-front capital expense of servers and traditional hosting for a startup is great, but come on!  We need to move beyond PaaS as just an elastic web hosting environment!

What I was hoping to see in the entries was much more around the unique attributes of "the cloud."  I hoped to see more ideas that could leverage high on-demand processing power (like the very cool LuxRender from NZ-based GreenButton http://geneious.greenbutton.net/apps/luxrender and http://www.luxrender.net/forum/viewtopic.php?f=13&t=4607 on Azure).  Or what about leveraging the Azure DataMarket (https://datamarket.azure.com/) and some of its great collections of statistics from the UN or the European Environment Agency, or everyone's favourite US Crime Statistics (https://datamarket.azure.com/dataset/c663117f-db6d-49e1-bc83-b05390bb3c70)?  Azure has a killer CDN to let you cache static application objects at strategically placed locations around the world: surely that could change how you build apps!  Or, how about hybrid apps that bridge the on-prem world with the cloud?  The new SQL Azure Data Sync (http://www.microsoft.com/windowsazure/sqlazure/datasync/) preview is hotness of the highest order!

Now, to avoid turning this into an advertisement for Azure (but feel free to sign up for a freee trial here http://www.microsoft.com/windowsazure/free-trial/ hehe), this condition, this treatment of the cloud as some sort of next-gen hosting environment, isn't just an Azure problem.  Many other PaaS offerings have similar features and also seem to struggle to get users to think beyond the same kinds of web apps that we were all writing with CFML in 1998.  It's our responsibility as PaaS vendors to help developers understand that PaaS is not app hosting.  If you want that, there's plenty of great hosting partners out there.  PaaS, Azure anyway, needs to emphasise the 'P' far more!  It's a new platform for building applications, a new environment with new capabilities for a new decade.  As a marketer, I'm failing as long as our contest entries fail to recognize that. 

Comments [0]

Apple PaaS?

I've been having a conversation among some of my colleagues here on the Azure team about Apple getting into the PaaS space.  You may have seen this post http://gigaom.com/apple/apple-may-have-snapped-up-icloud-com/ about a new domain they may have grabbed.  While most rumours center on a cloud based music service, I suggested that they may be pursuing an app platform.  Why?  Consider this: the App Store has delivered over 10 billion apps producing revenue somewhere around $2B per year.  Now, Apple earns from various points of the app lifecycle: the hardware, the one-time app store purchase, e.g.  But once those apps have been installed on the user’s device, most connect back to someone’s server somewhere – be it a weather app, or stock checker, travel reservation engine or what have you.  That’s a major part of the app’s lifecycle and Apple currently derives no revenue from it.  If there’s one thing Apple hates, it’s being absent from a revenue stream created by its ecosystem.  If I’m Apple, I have my global content platform (powering iTunes and the App Store, software updater, and the like), so why wouldn’t I want to see all those app developers deploy their server-side code on my platform where I could charge on an ongoing basis?  Otherwise, they’ll be going to Amazon or Azure or Rackspace or elsewhere.  Have a global content delivery platform?  Why not rent out some of it in a utility model to developers delivering apps through the same platform?


Apple has developers, apps, and a global content service: monetizing that all as a Platform as a Service seems like a no-brainer to me.  Certainly a better story IMO than VMWare!

Thoughts?

 

Comments [0]

My WebDU 2011 Presentation slides

Enjoyed kicking off WebDU 2011 with a talk on the difference between IaaS, PaaS, SaaS, DaaS, and all those other pains in the *aaS!

</object><div style="padding:5px 0 12px">View more presentations from Tim Buntel.</div></div>

Comments [3]

About