Why has the flow of technology reversed?
Esther and Fred are right that at least in some specific areas like lightweight web services, the flow of technology has reversed. This creates an interesting opportunity for entrepreneurs and investors who,. like us, are on the look out for ways of using services developed for consumers as a framework for business services.
But figuring out which consumer technologies will work in a business context requires taking the analysis a step further. The interesting question is\... why has the flow of technologies reversed? I think there are two key drivers. The first is the maturity of the information technology infrastructure, and the second is the increasing complexity of information services.
The maturity of the IT infrastructure has dramatically reduced the capital required to create and deliver an innovative web service. When we invested in del.icio.us, Joshua Schacter was already supporitng over 50,000 users on borrowed hardware with less than a $200,000 investment. That was possible because Joshua was able to use the rich development environment that surrounds Linux, Apache, MySQL, and Perl (the LAMP Stack), and cheap commodity hardware. He also did not need to build a communications network to enable users to get to his service. He could coount on the Internet. Contrast that development effort with the development of the early internet, where expensive mainframe computers were linked together with dedicated communications circuits. That would not have happened without the support of the govenrment.
The second driver is complexity - A friend of ours mentioned last week that by the late 90s the typical enterprise software suite had over 4000 separate features. The most he had avery seen used in any installation was 750. These systems just got too complex for anyone to be able to use all of the features effectively. Contrast that with the current generation of consumer web services. The most obvious example is Google. When it first came on the scene it was a box on the screen into which you typed what you were looking for. Consumer web services tend to do one thing well.
So it is not surprise that business users are beginning to capitalize on the rapid pace of innovation on top of the LAMP stack and that business users are rebelling against the uneccassary complexity of the current generation of enterprise software and beginning to use tools from their own experince as consumers that are much simpler and easier to use.
All of this suggests that the best enterprise implementations of consumer web services will be lightweight and easy to use.
April 14, 2006 10:00 AM, By Brad Burnham
Tags: complexity consumer webservices
Comments (4)
There is not a single enterprise application that does not require implementation. This is the main barrier to entry. Consumer app typically require only a one person implementation. Enterprise apps even small ones are usually 5 or more. We see the biggest problem is training users how to use apps in their businesses.
We sell a hosted enterprise application and have experimented with pricing, even giving it away for free. In every case, implementation is the big problem even for savvy users. Plus businesses have trust issues and expect a sales process, demos, etc.
We are now experimenting with giving away very small free applications which solve important but simple problems and are trying different ways to make it easy to implement. This seems to us to be where Microsoft got it right all those years ago. Make it easy to migrate data from a competitors product. We expect to make our revenue in up selling to our mature larger application which requires more traditional sales and implementation process. That is where the money is.
Posted by Dan Cornish , April 20, 2006 12:58 AM
I think you're wrong, Brad. 13 years ago during the client server/re-engineering days, developers had to pay for training, pay for books, buy software to get documentation, and use what was offered by vendors. Open source was nowhere, online info was nowhere, and it was really, really hard to learn on your own.
The web changes all that. Cheap hardware, sure. Cheap storage, yeah. But easy access to code snippets, peer input, documentation, user's groups online, cheap components, etc--that's what makes the difference, combined with acceptance of independent software vendors and individual developers. We're innovating and releasing faster than the large software companies.
Does Delicious scale well? Do you add more servers, or just more resources to servers? Because that points to another issue: is it architected well?
Posted by charlie crystle , April 20, 2006 02:04 AM
* consumer to enterprise technology flow is definitely happening in networking hardware
All the "hard problems" are in the digital home (not the enterprise). The home has more demand for higher speed (for photos and video), ease of use, and low-cost. Examples
* wireless for video
* storage for home video
* wiring the home (Moca, HPNA, etc)
* Very High Speed Home Networks (GigE) to remove need for QoS ( for big files and video)
* High-Speed Access (Ethernet and PON)
* these are all tough technical problems that need to be solved at the lowest cost. There is no question that once solved, that these technologies will become adopted in the Enterprise. The consumer is the technology driver today for networking.
* this can also be seen in the big gains consumer networking companies have had in revenues. A lot of the gain comes from low-end enterprise gear (that is surprisely high-featured)
Posted by Iain , April 20, 2006 12:36 PM
Wouldn't complexity within the enterprise actually encourage development flow, especially given biz users' apparent unhappiness with present complexity?
Instead, I would add immediacy to your flow thesis. It is orders of magnitude quicker to get uptake of a consumer web app vs enterprise. And with one-man/small team shops being scooped up for handsome money, as you guys have seen first hand, it has been a rational entrepreneurial decision to focus on this market. Hence, flow toward consumer facing services.
Enterprise web apps will have a huge impact, it's just going to take longer than the consumer space. While people get hung up on "built to flip," there are countless other firms toiling away out of the spotlight prepared to really knock the cover off the ball.
Posted by Andy Nardone , April 18, 2006 03:02 PM