Archive for the ‘paas’ Category

Maturing Google’s Platform-as-a-Service

Monday, December 22nd, 2008

Zoho recently launched a new product named Zoho Creator, which is basically an IDE for Google App Engine (GAE). I have previously tinkered with Google App Engine, creating just a really simple app. App Engine is different from Amazon’s portfolio of cloud computing services because it’s really a platform (using Google’s “LAMP“-like stack), as opposed to a discrete utility service like S3 and EC2. Programming GAE is more like programming for Salesforce.com’s Force.com platform, except GAE is a lot less proprietary.

One of App Engine’s drawback is that it only supports one programming language: Python. With this new offering from Zoho, no coding is required (see video clip below). Naturally, my first instinct is that this is similar to a Web site editor (e.g. DreamWeaver, et al). You can slap together an OK-quality static Web pages, but anything more complicated, like creating a dynamic “shopping cart” function would probably still require that you popping the hood and getting your hands dirty with the lower level details for more granular controls.

At any rate, I still think that this is a great step forward for cloud computing/Platform-as-a-Service (PaaS) because it’s lowering the barrier to entry for non-expert users. That’s one of the signs of a mature technology product: when non-experts can use it with ease. We’re not there yet, but we’re definitely making good strides.

Case study: Bell Atlantic and AT&T’s vendor lock-in battle

Wednesday, August 13th, 2008

One of the reasons I really hesitated in getting the eyePhone is because among other things, I truly dreaded the 2-year mandatory contract. I hated the idea of guaranteeing someone a consistent revenue stream and possibly be locked-in to their demands should they raise their prices.

Case study: Bell Atlantic and AT&T vendor lock-in battle.

In the 80′s, Bell Atlantic spent $3 Bil on AT&T 5ESS switches for Bell’s telephone network. AT&T’s switches were much more superior to Northern Telecom and Siemens at that time.

However, Bell didn’t properly size the vendor lock-in.

The 5ESS switches ran an operating system proprietary to AT&T, so whenever Bell wanted upgrades or new features, it was pretty much at the mercy of AT&T’s pricing weather.

Case in point: Bell Atlantic wanted its systems the ability to identify toll-free “1-800″ calls. AT&T didn’t provide (of course they didn’t!) any documentation or API for Bell to develop this feature themselves, and quoted Bell $8 Mil for a software upgrade just to do that. Bell had no choice and bent over. Voice dialing? $10 Mil! (really)

This extortion was a fat consistent revenue stream for AT&T, and made up 30-40% of AT&T’s switch revenues. AT&T’s position was further solidified by using its proprietary OS to prevent others from developing compatible equipment that may cannibalize sales from AT&T’s product line.

Bell Atlantic could not just throw AT&T out because (1) the switches had a lifespan of over a decade (2) removing and installing was expensive (3) the used switches had low re-sale value, because nobody not already locked-in would want to be locked-in ;)

In other words, the switching costs were astronomous, and Bell was hurting real bad in the wallet. It sued AT&T in 1995 for monopoly.

Shifting gears. To draw a parallel, in many ways, traditional on-premise software vendors use such tactics to .. well, play their hand.

With SaaS, this problem goes away. The customer can switch vendors on a dime; without the safety net of a perpetual licensing scheme, vendors have to constantly prove themselves by continuously delivering innovation and value to their customers — or risk losing them to the competition.

A flat world combined with fierce competition to innovate can only mean more and better options to the consumer :)

Unlike traditional on-premise vendors, SaaS vendors can’t rely on their own product development “baggage” to milk a drying revenue stream.

Reaggregating SaaS/PaaS results for a competitive advantage

Monday, August 11th, 2008

In a previous post, I discussed how cloud computing and the Grameenphone microfinance endeavor fit into a McKinsey article about the benefits of unbundling production from distribution. This weekend I decided to revisit the article again just to see if I would see anything differently this time around.

The world is indeed getting flatter. The article’s section on “Tapping into a world of talent” talks about how technology today fosters interactive online collaboration which in turn enables companies to outsource increasingly specialized aspects of their work and still maintain organizational coherence.

[...] technology permits them to decentralize innovation through networks or customers, it also allows them to parcel out more work to specialists, free agents, and talent networks.

Top talent for a range of activities-from finance to marketing and IT to operations-can be found anywhere. The best person for a task may be a free agent in India or an employee of a small company in Italy rather than someone who works for a global business service provider. Software and Internet technologies are making it easier and less costly for companies to integrate and manage the work of an expanding number of outsiders [...]

This trend should gather steam in sectors such as software, health care delivery, professional services, and real estate, where companies can easily segment work into discrete tasks for independent contractors and then reaggregate it. [...] Competitive advantage will shift to companies that can master the art of breaking down and recomposing tasks.

Globalization is inevitable, and increased competition means keeping businesses on their toes, which in turn translates to increased benefits to the consumer. In short, it’s healthy for both consumers and producers (unless you’re just lazy).

In a way, this also parallels SaaS/PaaS. Look at the SaaSCon sponsors list for a glimpse of some of players out there. There’s no shortage of on-demand providers filling gaps in the cloud-computing/SaaS value chain and gaps left open for disruption by on-premise incumbents.

Each cloud computing/SaaS vendor mostly specialize in one verticle and strive to dominate that niche-delivering a continuous stream of value (innovate or die) for less (save customers money or be undercut by your competitor). Jeff Bezos has explicitly said it before that with Amazon AWS, he wants to innovate there by reducing operating expenses, increasing efficiencies infrastructure through economies of scale, so that (here’s the important part) “.. the cost savings can be then in turn be returned to the consumer.” Ok, so I paraphrased, but he said it in a video clip somewhere online and I can’t seem to find it right now.

The point here is that he’s trying to save the consumer money (and that’s a great brand promise!) The jury is still out on that one, given that AWS is still relatively young, but if anything else – it’s a makes a good sell (who doesn’t like to hear that their vendor is actively trying to save them more money?), but ok .. I’ve digressed too much on Bezos. I just can’t help liking people (and companies) who genuinely want to help others (the customers) be successful, so that they themselves can be successful too. Pay-for-performance? Pay-per-drink? Cloud computing? ;)

Just to name a few vendors:

  • Google Docs -> on-demand “MS Office”
  • Amazon AWS -> on-demand computing power, storage.
  • Salesforce -> on-demand CRM
  • CODA -> on-demand finance application (built on Force.com!)
  • NetSuite -> on-demand ERP
  • WorkDay -> on-demand HR, payroll, procurement, business intel, ERP

Odds are that your company is already using some kind of on-demand solution for one of its functions, even if you do not realize it.

The way I see it, if you think of each of these functions as discrete tasks with each farmed out to a particular SaaS vendor, then the need for the reaggregation for each of the function’s results is obvious. I agree with the article that companies that succeed in recomposing these tasks would hold a competitive advantage.

It would allow executives to conduct business at the speed of thought (asking questions like “how can I reduce operating expenses here today, can I realistically turn the ship around fast enough in anticipation of this tectonic shift/change in competitive landscape”) – as opposed to the speed of “how fast can I line up all the columns in this Excel spreadsheet from that tabular data in the PDF spreadsheet and .. hmm, it would be really nice if I could overlay on this the results from some SQL queries.. oh wait I have to get those from John in IT first ..”

The $200-300Bil business solutions market is open for disruption by Platform-as-a-Service.

Web 2.0 – all grown up and ready to change the way we do business.

Not withholding innovation by decoupling from low(er)-level constraints

Tuesday, August 5th, 2008

Marc Benioff, chairman and CEO of Salesforce kicked off this month in cloud computing and SaaS news with a guest post on TechCrunch.

Some key highlights:

Web 1.0 was about the emergence of the “killer app” from companies like eBay, Amazon.com, and Google. Although we thought of them as Web sites at the time, they were really amazing applications with a level of functionality, ease of use, and scale that had rarely been seen before by the average consumer. Transactions, not just of goods but of knowledge, became ubiquitous and instant. The efficiency and transparency that was once the domain of global financial markets was now at the command of individual consumers and businesses.

It’s about empowering the everyday worker (especially the small business guys) with powerful tools previously only available to mega-corps with deep pockets. Web 2.0, cloud computing, SaaS leaves a taste of benevolence on my tongue, and I like it.

That’s “benevolence”, the way Paul Graham says it:

Surely Microsoft isn’t benevolent? But when I think back to the beginning, they were. Compared to IBM they were like Robin Hood. When IBM introduced the PC, they thought they were going to make money selling hardware at high prices. But by gaining control of the PC standard, Microsoft opened up the market to any manufacturer. Hardware prices plummeted, and lots of people got to have computers who couldn’t otherwise have afforded them. [...] Microsoft isn’t so benevolent now.

I guess I like rooting for the underdog and taking on the incumbents ;) Any big problem is a big opportunity, right?

More from Marc’s post:

Web 3.0 changes all of this by completely disrupting the technology and economics of the traditional software industry. The new rallying cry of Web 3.0 is that anyone can innovate, anywhere. Code is written, collaborated on, debugged, tested, deployed, and run in the cloud. When innovation is untethered from the time and capital constraints of infrastructure, it can truly flourish.

Emphasis is mine on that last sentence, since I think it’s worth noting. It’s why I absolutely <3 SaaS/cloud computing and believe in it. Too many times, as a software engineer .. you cut back on truly innovative ideas because of the voice of fear that speaks softly to you from the back of your mind, “but dude, you’re opening a can of worms; you need to do this, that, and it’s mostly all prickly infrastructure stuff that you’re not an expert in and would take you days/weeks/months to get up to speed! gahh .. can’t we find another less innovative solution that is easier to implement?”

And sad to say, most would take the easy way out — innovate lesser, in smaller incremental chunks (to keep the pain points low). And that’s infrastructure holding back otherwise creative problem solving.

According to Daryl Plummer, managing VP of Gartner (IT), about $8 of every $10 spent on technology in corporations is for maintaining systems, as opposed to innovating. Talk about a serious baggage.
“It’s hard to turn a big ship very quickly [...] You have technologies that are like cement in these businesses—they’re hard to change and get rid of.”

I believe that with cloud computing and SaaS, we’ll see more bottom-up (of the org chart) punching of holes in the “corporate policy” firewall, because business units needs stuff done, and IT departments can’t keep up. This is especially true in huge bureaucratic companies.
(more…)