(this is page 76 of 121)





Latest

Opportunities in clouds

Could be because I haven't seen a cloud in the sky all week, and it could be because Google was totally inaccessible all morning (at least from Jamaica), but I've got this cloud computing thing on the brain these days. If you really believe that it is a tectonic shift, and that no company will be able to ignore it, read on for my take on the 3 types of cloud efforts, and a couple of associated opportunties:

Core-to-DNA: This is where Amazon and Google fall today. Building out scalable, flexible infrastructure that is up 99.x% of the time is so core to what they do that they just won't be able not to take advantage of the shift in computing. Whether Amazon's more flexible or Google's more prescriptive approach wins the majority of the mindshare is less interesting than the fact that you will always be able to count on these guys' delivering solid solutions. The business opportunity here will be in providing a layer of portability across them so that no one gets vendor lock-in syndrome.

Must-have-for-strategy: Microsoft, IBM, and Yahoo are good examples of this one. Each has a pole position in some previously dominant platform, each knows they have to move in this direction, but because it is not core to who they are, they're not likely to execute well, at first anyway. No matter though, because the folks with deep pockets (Microsoft, IBM) will spend whatever it takes to bring the B/B+ offerings to market. They may not scale as efficiently, and they may not be as reliable, but these guys will find some market-power way of gaining meaningful traction.

Wannabes: EMC, HP, big banks, and other utilities fall in this last bucket. These are the companies that know that this cloud shift is a big deal, have some sort of vested interest in being able to play in this domain, perhaps even have a starting move, but are quite simply finding that "clouds is hard!" As a result some will stay (if it is critical to their survival), some will outsource, and some will fall out clouds (if they can afford to ignore it). For the time being however, they will all continue to half-execute, confusing endusers and keeping developers on their toes. Naturally, the near term business opportunity here is the best, as all of these guys will need software to try to keep their clouds running (GigaOm has a nice post of what some of this software might want to target as initial opportunities).

I have no doubt that there will be plenty of movement from tier 3 to tier 2 and back— what I am less clear about is who, if anyone, will be able to move to tier 1 and join Google and Amazon in what will be a very lucrative market opportunity.

Comments Permalink

As much as I've come to believe that Paul Graham has now sucked in too much of his own exhaust to write with as much insight as he used to, his latest essay on "being good" as a directing principle for startups has stuck with me like a sharp chicken bone at a barbecue. At first I wanted to dismiss it, thinking of cases like the OLPC debacle, where people's politics and grand visions for making the world a better place get the best of them. But as I let the essay fester, it began to dawn on me that Paul's "doing good" is just another way of saying that you need to have a bigger mission than just flipping to Google, or even just making yet another photo book configurator.

At Tabblo we had this: as Eric (our customer guy) reminded me recently, if you look through our wiki (now buried behind layers of HP VPN goodness), you'll see a number of pages that talk about making people more creative in their self-expression, making them better story-tellers despite the normalcy of their everyday lives. Helping the unwitting blogger, empowering the casual publisher, and all of that other jazz we'd talk about at lunch, late at night, and during those moments when we'd feel like big bad world was crushing us.

But lately I've also been thinking about the fact that all of Web 2.0 has been geared towards enhancing self-expression: from Flickr to Facebook and Blogger to Twitter, the whole ride has been about letting people create content that they can (in the best of cases) use as a vector for connecting to other people doing exactly the same thing. In its best days Tabblo did this well because, after all, part of making people feel like creative story-tellers was finding them an audience who was willing to listen and to engage.

As Adam Green has recently pointed out however, greasing the skids of self-expression is just one of the many ways in which the Internet and what we build with/for it can help people. While it is a very good thing that we've all gorged on this particular goal over the last 5 years, the time may be upon us to start thinking about the stuff that is useful and that can actually make a positive impact in the way that everyday lives are being led throughout the world. After all, lord knows we've got enough crises to which we can apply our entrepreneurial energies.

For my part, I've spent quite a bit of time as of late thinking about online group formation and why it is that even on the most "modern" of sites, it still feels so incredibly encumbered, especially when you compare it to the way groups form in the real world. I'm tired of "friending" people, I don't want to "follow" them, and I really could care less about signing up for another "Evite-killer." But I do know that the seeming ease with we can coalesce around a particular task for fun, profit, or just to do good at school, at work, or within our communities is something that still needs a lot of work in the virtual.

This group-forming detour aside, the guys mentioned above may be right— it's time to do something useful, something good and to leave the self-expression behind for a little while.

Comments Permalink

Meshing with constraints

Not much stuck with me from my days at business school, but one of the few things that did was from this wacky book called "The Goal" about manufacturing optimization. In it, the author uses a boy scout troop, and more specifically a fat kid named Herbie, to make the point that in most cases a system is only as efficient as its least efficient component, and that as such, finding this constraint and being obsessive about how to loosen it tends to yield the best results. Pretty obvious I realize, but over the years I've found that it is a great way to think about emerging technologies and new opportunities.

Today Microsoft is taking the wraps off of Live Mesh, a combination platform/service to sync files and application across all manner of devices. It looks like an interesting project, especially because it seems to leverage some form of feed syndication for its transport layer. But apparently it's more than just a project: it is the grand vision of Ray Ozzie and has sucked up about 2 years and 100 people to get to release (which reminds me of Ozzie's previous Project Groove a.k.a Lotus Notes 2.0 which took serious resources to get to sync files between PCs). Despite it being tiny by Microsoft standards, feeding all of those people for all of that time to get file and application preference sync seems a bit excessive to me.

More importantly, are we really at the point where the world now needs Lotus Notes 3.0? Let's see: 1. devices are proliferating, with smartphones leading the pack... check. 2. More and more personal computing is taking place in "the cloud" on websites like Facebook and Gmail... check. 3. There is still a lot of data being locked into the local hard drives of millions of PC users... check. 4. What we are all dying for is a way to sync all of this stuff so that we can stop thinking about desktop or cloud and seamlessly carry on our personal computing tasks from any device anywhere...really?

Back to Herbie the fat kid and the theory of constraints. Where is it that we are really suffering today? Is it in new protocols to sync data across devices? It seems to me that everyone and their mother is capable of designing sync protocols, and that more importantly, Internet pioneers solved this one even before the days of Lotus 1.0. Instead, there seem to be three clear places where we are currently suffering from constraints that might make for "big play" business opportunities: 1. we never have enough bandwidth in enough places (whether it is cable, DSL, cellular, etc.), 2. we don't yet seem to know how to economically store enough stuff in the cloud, and 3. in the case of portable devices, we never have enough power efficiency.

But more sync protocols? Hybrid platform/services that require buying into one vendor's silo? At best, these things seem to me to be band-aids that are trying to get us around some of these constraints instead of through them.

Interestingly enough, normally M&A-allergic Apple today announced the acquisition of PA Semi, a chip company that focuses exclusively on power management in CPUs, seemingly to address constraint #3 in their iPhone platform.

Comments Permalink

Amazon and innovation

Amazon's Jeff Bezos is the media's favorite love/hate boy, vacillating between genius to dolt as his stock price see-saws with the collective effervescence of the Internet's potential. However, it can not be denied that for the last 10 years he's charted a really exciting course in the development of three major platforms: the web as storefront (Amazon.com), the web as underlying fabric for computing (EC2/S3, etc.), and the web/device ecosystem for consuming content (Kindle). This month he seems to be on top again, with great articles in Businessweek and Wired describing both the company's ability to innovate and its cloud computing initiative.

For those of us now familiar with the way big tech companies think, the best quote is one he gives Businessweek on how to drive innovation:

Q: Every company claims to be customer-focused. Why do you think so few are able to pull it off?
A: Companies get skills-focused, instead of customer-needs focused. When [companies] think about extending their business into some new area, the first question is "why should we do that—we don't have any skills in that area." That approach puts a finite lifetime on a company, because the world changes, and what used to be cutting-edge skills have turned into something your customers may not need anymore.
(via O'Reilly Radar)

If the last year of HP life has taught me anything on this score, it is that this skills-vs-customer-needs issue is the club that anyone who interested in protecting fiefdoms will use to kill innovative but risky ideas. It has many guises, from "we don't know anything about X," to "how could we possibly get X to Y scale," but at the end of the day it ends up feeling like the same thing: "we're already good at X so why don't we do X+incremental instead of something totally new!" Fortunately, there are ways of mitigating this: buy new skills, bring in outsiders, empower the mavericks, realize that X is going away and freak out— but all of these take time and don't come without pain.

That said, all of this customer-focus goodness also strikes me as a bit glib and this week's Amazon media coverage of the cloud computing initiatives points to why: when Amazon started S3/EC2, there was no way that there were customers clamoring for what they've built. No one came to the front door saying "wow, you guys are at 10% utilization, wouldn't it be cool if you could rent me some virtual servers?" and yet, much in the same way Apple has done time and time again, Bezos & crew were able to see evolving trends coalescing into a solution they thought customers would want.

I'm all for listening to the customer, and I'm certainly all for thinking like the customer, but it's important to realize that— at least in technology— there is something at the moment of initial creation that requires creative insight in a way that listening to what your customers want just doesn't allow. Something about skating to where the puck is going to be comes to mind...

Comments Permalink

While the embedded video clip has become as common as the animated GIF, littering webpages with colorful bits of video that help to make a point or add color to a blog post. Unlike text however, most rich media suffers from a the linear consumption problem; that is, if what you are doing is scanning quickly, the various bits of rich media become speedbumps in the path of grazing the underlying content. In some cases, the video is really worth stopping for, but in most you end up feeling like you want those 3 minutes back in your life.

For a while now I've thought that this proliferation of linearly consumed media would mean that rich recommendation systems were going to be making a comeback (hello collaborative filters!), but in order to do so, they'll actually have to work (a dubious proposition), and more importantly, we'll need to find a sub-URL way to indicate ratings in a open way that can be used as input by the various recommendation algorithms.

Another approach which I've recently seen used to great success is the mixing of the rich media with solid scannable navigation aids and pointers into timecodes in the video, so that the interested user can quickly skip around. Much to my surprise, it was the New York Times that first showed me the power of this approach with their "live transcript" of the Obama speech on race. Having a transcript that scrolls along in sync to the video is such a simple thing to do, but makes such a huge difference in lowering the impedance mismatch between the text-heavy web and various rich media formats.

Another great example is being built by a startup called Omnisio which allows for the annotation of presentations (a frequent subject of the embedded video clips I come across), along with the slides the presenter is using. They've currently got a bunch of annoying user-commenting features but once you shut those off (from the user control panel), the resulting experience makes watching presentations 100x more enjoyable and efficient.

Just another lesson in how sometimes the simple solution, well executed, can yield great results.

Comments Permalink