Eric Knipp

My colleague Ray Valdes has just completed his work on the latest Gartner RIA MarketScope. Ray was kind enough to invite a number of analysts, including me, to participate in this important research.

While the cat-herding of analysts can be a thankless job, I think you’ll agree that in this case the results are well worth it.

The RIA Marketplace is in a state of flux. This year saw the introduction of Microsoft Silverlight 3, which in many respects is equal to Adobe Flash in terms of support for enterprise use scenarios. Ajax remains the dominant RIA choice, and HTML5 is poised to expand the power and flexibility of the browser-only approach.

Now that Microsoft has validated “heavy RIA” in the eyes of many enterprises, interest in RIA technologies is increasing across the board. Frequent Gartner inquiries indicate that clients pit Ajax vs. Flash vs. Silverlight against each other in evaluations for new RIA projects. What does this mean for JavaFX and other technologies? Tough to say for sure, but my bet is that the “heavy RIA” arena comes down to a battle between Adobe and Microsoft, and that there is enough room in the market for both to be successful.

Finally, vendors are shifting to meet a new demand from the enterprise – for an end-to-end RIA solution that involves both client- and server-side components. To that end, new “full-stack” vehicles for delivering the user experience will emerge over the next 12-24 months.

For more interesting thoughts, and to see the vendor-specific analysis, check out the 2009 RIA MarketScope.

Oh, and have a Happy New Year!

If You Build It, They Will Come. If You Don’t, They Might Build It Anyway.

Etsy, a Web proprietor of handmade goods, didn’t have a public API. That didn’t stop the “Etsy Underground” from hand-rolling one.

Etsy did have an API for internal use only. While it wasn’t promoted, it also wasn’t well-hidden from the public. Some enterprising developers figured that out, and reverse-engineered the API for their own nefarious purposes. This resulted in third party applications, code libraries, and widgets all leveraging Etsy’s site and content.

By the time Etsy got wind of what was happening in a new developer ecosystem, it was too late to steer the direction of the community directly. They could either shut off the community’s ability to innovate, or add fuel to the fire. The company chose the latter, and launched an open API which replicated all of the features the Underground had come to know and love, and took it further.

The result? According to Etsy, the API now gets 9 million invocations a month, and has spawned 17 applications, including 5 for the iPhone. One site in particular, craftcult.com, is a leading driver of traffic to Etsy’s site.

Handmade goods are interesting, but what about your business? Chances are, your data is already being scraped by somebody; maybe you can make it easier for them, and in so doing foster a community that spreads your brand.

The urge to control access is great, but what if you can become a bigger broker in the conversation by sharing instead? Clearly this won’t work for all businesses – sometimes the content is just too valuable to share, and loses its value to the enterprise once it spreads.

Or, in this case the hammering on Best Buy’s Remix API. According to @bestbuyremix:

5.7 million API calls yesterday. The 19th to yesterday (5 days) accounts for 61% of our total so far this month.

The day before, @bestbuyremix reported 5.3 million API calls.

I spent half a day last week at the Business of APIs Conference, where I had the opportunity to meet with folks from Best Buy, Hoover’s, Etsy, CBS, the New York Times, and more who are involved in open API programs.

It is pretty clear to me that APIs are going to become an invaluable – and common – tool in the enterprise’s public Web presence toolbox. When we finally hit “critical mass” – more retailers with public WOA APIs than without – will API traffic become a new barometer for predicting Christmas shopping economic outcomes?

UPDATE – Proof that an API community must be closely measured and managed. Your key partners might need help at a critical moment, as according to @bestbuyremix:

We had to increase the API rate limit for @miloshopping because of all of their press so far today! NYT article: http://bit.ly/073VfE9

Happy Turkey Day Everyone!

Adobe, Reloaded.

That’s the thought I walked away with after my experiences at MAX this past week. I used to think of Adobe as this cute, fuzzy bunny of a company with great designer and RIA tools (plus that PDF reader thing). Friendly, happy-go-lucky Adobe, loved by all and inspiring fear in no one.

Adobe’s roots, postscript fonts notwithstanding, are in delivering great tools to designers. The Macromedia acquisition of 2005 extended Adobe’s capabilities to delivering rich user experiences to end-users with Flash (and to a lesser degree with ColdFusion). Macromedia was also a company with a narrow focus. Together, the new Adobe could offer end-to-end rich internet solutions from designer to developer. A great story, but RIA was still in a nascent stage. Adobe remained a fuzzy bunny of a company albeit one that did a great job as a steward of its acquired property in Macromedia (primarily Flash and ColdFusion – I can’t imagine any Flash or CF developers being justifiably upset at how Adobe has advanced those platforms).

RIA is no longer in a nascent stage as evidenced by the entry of Microsoft. I’ve characterized the RIA landscape as a “battle royale” between Flash and Silverlight, and I believe this is becoming truer every day. Ignored by Microsoft for years, Adobe is now in a battle  against one of the largest “megavendors” on the planet for mind and market share. Of course Adobe has a huge lead, but Microsoft has plenty of resources to play catch up (or at least to try).

The emergence of a well-funded, competent rival in Adobe’s core business forces the company to become more strategic about its growth. We can see the beginnings of a strategic shift underway.

The Killer Rabbit of Caerbannog

Adobe’s success has been tied to PDF and Flash for years, but lately all of the company’s eggs have transitioned toward the Flash basket. They’ve got a Flash-based BPM product (LiveCycle), Flash-based developer tools (Flash Builder, Flash Catalyst, the Flex Framework), even Flash inside of PDF (PDF Portfolios). With the acquisition of Omniture and the various LiveCycle/Flash services – for now collaboration and distribution – and the unification of the Flash runtime for mobile and desktop, plus the Flash iPhone development announcement, the company is erecting all kinds of barriers to imitation to keep Microsoft at bay.

Can Microsoft copy some of this stuff? Sure. Microsoft can buy an analytics company, they can create various Silverlight-oriented cloud services, and they can conceivably buy enough influence and expertise to get Silverlight on tons of mobile devices. It will cost a lot, but Redmond has scads of cash if they want to commit it to the fight. I’m not sure they’ll do so, but even if they do, it will take time to harvest their investment.

Rather than focus on core capabilities of Flash itself, Adobe is creating complements that improve the value drivers of the Flash runtime. Microsoft, to this point at least, has focused on making Silverlight functionally equivalent to the Flash runtime itself, a task which took the better part of two years. Now, I am not putting down Microsoft; delivering Silverlight 3 in two years is impressive considering how much time has been invested into Flash. Nor do I believe that Adobe is sitting still in terms of Flash player technology (the announcement of Flash 10.1 should put that to rest). I do believe that Adobe has created some very nice complements to the Flash runtime that Microsoft will be hard-pressed to repeat in less than 12-18 months (unless they’re already working on them, of which I have no knowledge).

I commented to several of my peers that I believe Adobe’s shift signals an intent to become what Gartner calls a “megavendor.” What’s a megavendor? To tell you the truth, I think its one of those things you understand when you see one. Today we recognize SAP, Microsoft, IBM, and Oracle as megavendors. These firms offer huge breadth of product lines. Adobe’s not there yet, but I think it would like to be. The fuzzy bunny has teeth.

disclosure: I am a paid shill for neither Adobe nor Microsoft. =)

My wife and I had a wonderful weekend. We first attended a wedding, and then celebrated her birthday over dinner with friends and family last night. As usual, topics of discussion ranged widely, but a particularly intriguing subject came up – Wal-Mart’s new packaging for its private label goods:

This packaging looks, well, cheap. When I first saw it, I was reminded of the classic “private label” beverage of choice in many 80s TV shows:

The person who brought up the “new look” of Wal-Mart private label goods found this cheap look unappealing, and believed that consumers would not respond to it.

Savers First

Without going into too much detail, I’ll start by saying that I believe that consumers are retrenching, and that the mentality of “conspicuous consumption” is gone for at least a generation. Too many kids and young adults in this country and others have seen their parents get wiped out by the economic downturn. They will become savers first.

Many white label products are just as good as their name-brand counterparts. In fact, many are identical, rolling off of the same factory floor. I admit I’m no expert but I’m unable to tell a difference between Great Value pasta and more expensive brands. The same is true for many other commodity items. I buy on price first.

I Beg to Differ(entiate)

At a high level, every company must decide on one of two “generic strategies.” Either be the cheapest (”cost leadership”) or provide more perceived value to the customer (”differentiation”). The customer’s perception of value can be influenced with more features, marketing, convenience, etc. The idea is that these value drivers will provide more of a reason for customers to pay a higher price.

Wal-Mart doesn’t make any bones about it – it is a cost leader. I don’t shop at Wal-Mart because I love the piped-in audio or the ambient lighting – I shop there because it’s CHEAP, plain and simple. I don’t care what the packaging looks like. More to the point, if the packaging looks cheap, it accentuates Wal-Mart’s cost leadership position. In fact, it starts to make cost leadership itself into a potential differentiator.

Imagine this scenario: Wal-Mart could charge $1 for a box of Ziti, which sits next to other brands that cost $1.20 or more. In the past, Wal-Mart’s Great Value packaging looked more in line with that of other brands (it wasn’t cheap-looking). Now Wal-Mart’s packaging looks cheap. It is designed to make the product look even cheaper than it is. Wal-Mart could raise its prices by a nickel a box, and customers might not even notice.

For a more in-depth discussion of generic strategies, competitive advantage, and corporate strategy I strongly recommend Gordon Walker’s book. It was perfect for a strategy novice like me, and it is a regular reference on my shelf.

How does this relate to Technology?

IT product and service companies have the same strategic choice to make – cost leadership or differentiation?

Most companies choose the latter, because differentiation offers more room for innovation – marketing campaigns, cool new features and delivery models, smarter consultants, etc. Cost leadership is about relentlessly pushing down fixed and marginal costs to provide the cheapest alternative. It’s really hard and you’re never safe – there’s always someone else who’s willing to do it a little cheaper if they can figure out how.

In our recent special reports on Cloud Application Platforms, we noted differentiators again and again – but rarely suggested considering a vendor based on price. It is not in the interest of vendors to compete on price, and enterprise CIOs would do well to remember it.

What were the skies like when you were young?

Today we’ve published our most extensive work thus far profiling the providers of Cloud Application Platforms. Our research includes forty companies, all of which deliver some type of platform enabling development of Web applications using a Cloud Computing model. A few weeks ago I blogged that the Cloud Application Platform market might be on the verge of shakeout; whether I was right or not, we should all agree that the market is young compared to the enterprise software market in general.

They Went On Forever

Because we found so many qualified vendors, we decided to split the research into two reports – one covering “Enterprise Generalists” with a heritage outside of Cloud Application Platforms, and the other covering “Cloud Specialists” who focus solely on this new type of development, deployment and runtime environment.

I’m stunned to tell you that even after publication of this extensive work, we’ve found another (albeit much smaller) batch of vendors in this space. Rest assured that we will be covering them too! If you are a provider of SEAP, APaaS or other types of Cloud Application Platforms and believe that you should be on our radar, please get in touch with me so that we can set up a briefing.

They Were Beautiful, The Most Beautiful Skies as a Matter of Fact

The platforms we profile in our report range from the “heavy hitters” of the industry in Salesforce, Google, and Microsoft to smaller, focused startups like Heroku, LongJump, and Bungee. Differentiation is the norm in this market; there are wild swings in platform architecture, programming language, developer toolset, and target audience from one Cloud Application Platform to the next. Some are designed to leverage the skills of professional developers, while others are meant to enable “Citizen Developers” outside of IT in building applications.

Its an exciting time to be involved in technology industry research.

Before I joined Gartner I was a developer. I was pretty deep into a handful of technologies and had broad understanding of many others. In my day-to-day work I was always looking for shortcuts and improvements to make me a better and more productive developer. Most of these related to the technology tools I was already working with.

I think most decent developers do this, and some try to stretch by learning brand new unrelated tools (even easier if you’re a consultant moving from project to project). IMHO if you bother reading developer blogs at all you’re already several cuts above the average developer.

A Peek Behind the Curtain

At Gartner we like to slice and dice everything and everyone into patterns, quadrants, matrices, etc. I’m not exposing any state secrets when I say that our business (at a very simplistic level) is divided roughly in half between producers of technology (vendors) and consumers of technology (end users). We get to find out what kind of horrible pain our end users are suffering and advise them on remedies. We also get to find out what vendors are planning to do, and advise them on strategy. Sometimes it feels like cheating.

As you might expect, by and large technology producers are much stronger at nuts and bolts than end users are. After all, for end users technology is a means to an end. For technology producers it is their core competency and competitive differentiator.

The Enterprise Web Developer

So, in my past life I was one of those enterprise Web developer guys who doesn’t really know anything compared to the engineers at Google and Microsoft. On the other hand I was much more in touch with the reality of doing business where technology is an enabler rather than a differentiator. The reality is that the best enterprise developers need to be slightly ahead of the pack technologically, and also good at the business side. You could say that from the perspective of the indivudal developer as an enterprise, use of technology becomes a differentiator for them.

If you want to be a highly-valuable (and hopefully highly-compensated) enterprise Web AD guy, you have to do a few things, probably not limited to the following (which are not in any specific order):

1. Be really good at the technology that you work with at your company.

2. Have a sense of what is coming down the road with technology that will affect your enterprise and work proactively to use it.

3. Be really good at talking in business terms to internal customers who really don’t care that you just built a cool widget using only 3 lines of jQuery. Be equally good at understanding business speak from the end users.

Do those basic things and you will be in a great position to make good bucks and maintain some job security and job selection. Why am I telling you all this?

Enterprise Web AD is Nearing an Inflection Point

The growing importance of the Web is pushing more transactions online. All businesses are under tremendous economic pressure.  The explosion in elastically-priced Cloud application platform technology is creating a new ecosystem for doing business on the Web, one in which costs are lower, time to market is better, and developer productivity is higher.

Five or ten years from now, most Web applications will not be developed for “traditional” app servers like Java EE, .NET, ColdFusion, etc (although possibly these app servers will be re-engineered to become elastic; hard to be sure at this point). Furthermore, the amount of data developers are used to working with in a relational DB is going to change dramatically – you will be working with ginormous data sets and SQL just won’t cut it anymore. Most likely the tools you are working with today will change dramatically. You might use different tools altogether or the ones you’re using might morph into new ones.

I can’t predict which platforms will win and lose. Its too early for that. However if you want to continue to be a highly-valuable, highly-compensated developer I have a few pieces of advice for you (also not in any specific order):

1. Get familiar with non-relational horizontally-scalable databases. (Ex: CouchDB, Google BigTable, Amazon SimpleDB) I am not a database guru and we have guys for that here at Gartner. But I can tell you that this stuff is different enough to be alien to most developers, even the good ones.

2. Get familiar with platforms and languages that support highly parallelizable applications. (Ex: Erlang, Clojure, Hadoop, anything that uses MapReduce internally; maybe even break out the old LISP) At a minimum bust out some recursive programming techniques in the platforms you already work on.

3. Get familiar with composition methodologies. Give up “Not Invented Here” (I know its hard – developers like building things). The best Cloud application platforms – at least for building business applications – will offer high productivity boosts with generically reusable components. If you don’t get on board with composition you won’t be able to wire together applications quickly from reusable pieces, and the crummy cut and paste code monkey sitting next to you will whip you.

In the late 90s, the Internet boom created a lot of new jobs and obviated a lot of old ones. The simplicity of the Web removed the need for as many desktop applications. We’re about to have another such shift. The best enterprise developers stay ahead of the curve. Its time to skill up for Cloud AD.

Industry Shakeout 101

According to my strategy professor, Gordon Walker, “the trigger point [for shakeout to begin] is when one or more firms achieve a level of productivity that neither weaker rivals nor potential entrants can match. Thus, the shift in entry and exit rates is ultimately caused by successful and sustainable growth strategies. Only the presence of one or more firms whose dynamic capabilities create dominant, defendable market positions can deter entrants and force weak competitors to leave the industry.” (page 137)

Shake it on, Baby

Now, I can tell you, that to be sure, there is no shortage of entrants into the cloud application platform industry. Force.com may have started things off a few years ago, but today there are at least 40 companies, and probably more than that, who play in the space. An upcoming Gartner report profiles many of these companies. I don’t think we’re in the shakeout period yet; however, I put better-than-even money on the idea that we’re in the last stages of expansion before a shakeout gets started. I submit as evidence the following:

1. CogHead, a leading (but small) technology company in the Cloud application platforms space, went bust last December. Its assets were ingested by the large, enterprise-flavored SAP. I am not sure what SAP plans to do with those assets, but I don’t believe they bought CogHead’s stuff to put it in a museum.

2. Microsoft Azure (and .NET Services), which is kind of a hybrid approach somewhere between Amazon EC2 (cloud system infrastructure) and Google App Engine (APaaS), will hit general availability sometime soon. This will mark the first big-time enterprise software player to have their own Cloud application platform offering.

3. VMWare has jumped into the fray with its planned acquisition of SpringSource – whose CloudFoundry offering constitutes a Cloud application platform has to be the main reason the company shelled out over 400 million, for what is a tiny open-source company (admittedly, Spring, around which the core business of SpringSource is built, is a great and widely-adopted framework for Java programmers). It will be interesting to see how VMWare plans to incorporate the SpringSource and Hyperic assets into its portfolio, and what (if any) other complementary acquisitions they will make. I’m not sure what to expect, but if I were a small Cloud application platform startup counting on a neutral VMWare container within which to run my “shared-hardware” multi-tenant platform, I’d be a little bit more worried than I was a few weeks ago.

Twist and Shout

The future for APaaS and other types of Cloud application platforms looks very bright to me. I’m particularly bullish on APaaS because of the productivity benefits available to developers using a combination of development of new code and composition of business-oriented services and components offered by the provider. I don’t think there’s a big enough market for all of the companies producing Cloud application platform software to wind up as “winners”, though. I also think that the time is drawing close where the other “mega-vendors” – IBM, Oracle, and SAP – must either launch their own Cloud application platforms or acquire one, or risk being left behind. Keep your eyes open for acceleration in the demise of Cloud application platform companies, and deceleration among new entrants.

Whoops!

I’ve been so busy with writing deadlines that for the last month I’ve had no time to update this blog. I hate to write things that aren’t “meaningful” to me, and I’m not able to generate meaningful blog posts with the same frequency of some of my peers. So, I ask that you bear with me as I try to get back into the swing of things.

Upcoming Speaking Engagements

I’ll be delivering a pair of presentations at Symposium this year – a great honor for a new, wet-behind-the-ears Gartner analyst. The first, “Citizen Developers: When End-Users Surpass the IT Department” is a Maverick (read: controversial within Gartner) pitch about the waves of end-user development we’ve seen over the years and the big one that is on the horizon at the nexus of Cloud Computing, Enterprise Mash-ups, Digital Natives, and Modern Development Tools. The second, “Buckle Your Seatbelts: Web and Cloud AD Convergence Ahead” speaks to the coming enterprise shift toward Web AD in the Cloud, and makes a few predictions. I’ll also be participating in a debate with Tom Bittman, called “Private vs. Public Cloud Computing: Does Control Trump Scale?”, in which we’ll explore the implications of private, public, and hybrid approaches to enterprise use of cloud computing. I’d be lying if I said I wasn’t a little terrified about speaking at Symposium, but I’m also excited to have this great opportunity.

Speaking of … speaking, tomorrow is a local analyst briefing in Dallas, where I’ll be giving a couple of 20-minute “rapid fire” sessions. One is on Enterprise RIA, and the other is a cut-down version of my upcoming Symposium pitch on Web & Cloud AD convergence. Should be interesting and it gives me an opportunity to work in some audience feedback before my final writing deadline for the Symposium presentation.

I recently finished reading a couple of books I’ve been chewing on for a while. As a full-time Gartner analyst and part-time MBA student, I don’t get a lot of time to read apart from my discipline, but I managed to squeeze in The Unfinished Game: Pascal, Fermat, and the Seventeeth-Century Letter that Made the World Modern as well as an interesting book that hits closer to my day job, Dot.Cloud: The 21st Century Business Platform. As an added bonus, I’m throwing in a short discussion of Enterprise Web 2.0 with EGL.

The Unfinished Game

I’ve got a dirty secret – I’m mathematically challenged, and it has always been so. I’ve had to work extra hard to be extra average at math. Secretly, I’d like to be good at math, and understand some of the advanced compression, encryption, and other algorithms that kick around the interwebs. Every now and then, I’ll pick up a math-oriented book, and I usually put it down in despair. Not so The Unfinished Game, which is as much about history as it is about probability.

In a nutshell, the book describes a snail mail conversation – an ancient form of collaboartion – between Pascal and Fermat regarding the correct way to devise the payment of a wager on an unfinished game of dice. The narrative weaves mathematical explanations between discussion of the time, place, and personalities of the story. A quick and compelling read, you’ll come out of it feeling that you’ve learned something truly important, borne of a discovery that took ages by modern standards. It is striking to consider how quickly we are able to use the social Web to iterate over interestin ideas,  in comparison to the pen and paper ways of old. Strongly recommended (4/5).

Dot Cloud

Peter Fingar’s book describes his visionary view of the forces leading to a Cloudy future. Fingar is a noted author of business strategy books, and Dot.Cloud is no exception. Rather than describe in exhaustive detail the technological underpinnings of the Cloud, Fingar provides actionable advice, compelling examples, and inspiring stories as he pontificates on the “21st century business platform”. The book is really eight discrete essays, and each explores another aspect of business in the Cloud. There is a heavy emphasis on BPM and end-user interaction and collaboration using technology tools. I highly recommend this important book, especially if your knowledge about the Cloud’s business implications is a bit cloudy (5/5).

Enterprise Web 2.0 with EGL

IBM’s EGL is a business-oriented language and deployment tool for the development of Web 2.0 Ajax applications that span several platforms including mainframes/iSeries (COBOL), Java EE, and the browser (JavaScript). It is especially well-suited for enterprises with heavy legacy investments that they want to resurface with Web 2.0 instead of replacing those assets or spending a lot of time building a SOA strategy. The book is an easy read – especially if you skim the language syntax examples – and illustrates that IBM’s engineers put a lot of thought into the design of EGL. It’s worth reading even if you don’t plan to use EGL in your enterprise, because it has content geared both at programmers and managers who need to understand the platform, and also because EGL provides an interesting thought experiment. There aren’t many examples of languages/compilation tools that take one high-level syntax and compile down to several other high-level languages. While you probably don’t need to read this if you are outside the IT modernization space, it is interesting nonetheless, and the conversational tone makes it easy to get through. The first couple chapters in particular are well worth a read for any IT professional as they provide a great introduction to the role of IT in competitive strategy (for non-technology firms) and to the subject of Web 2.0. Recommended (3/5).

In the Batter’s Circle

I’ve just started reading Michael Ogrinz’s Mashup Patterns, and Mark Levin’s Liberty & Tyranny is in the warmup area.