IDC tracks Ethernet switches form factors by modular/chassis, managed fixed configuration and stackable, and unmanaged switches for 18 vendors. A little more than a year ago, a vendor came to us and asked if we could break “web managed” switches out of our existing segmentation. We determined at that time that web managed switches consisted of switches we were currently counting in both our managed and unmanaged segments. We then proceeded to speak to all the other vendors of web-managed switches, a number of merchant silicon vendors, and several direct marketing / resellers who sell lots of low-end network gear. At the end of this process we determined we could estimate the size [a] and market share of the web-managed switch market. However, the most significant definitional difference we could identify between managed switches and web-managed switches was the inclusion or lack of, respectively, a console port on the switch.
We noticed one of our competitors is now tracking the web-managed switch market. So what say you makers, sellers, and buyers of web managed switches. Is this a market with defined competitive or logical boundaries or one that will take over the unmanaged switch market for reasons we couldn’t identify? Did we make the right call?
When making changes to either taxonomy or adding quarterly segments we ask at least following questions:
1. Is the new segment a subset of an existing segment or an entirely new market?
2. Will this new segment last for some time or is it just a feature that will be absorbed into existing market segments in less than 5 years?
3. Does this feature or function fundamentally solve a problem existing products can’t solve? Bonus if is a disruptive force in the market.
What kinds of comments are looking for?
1. We’re wrong about the console port thing, there are other aspects of a web-managed switch that set them apart.
2. Examples of technical or political dynamics facing web managed switches that will either cause web management functionality to stay constrained to a limited number of switches or proliferate to a wide(r) number of switches.
What we’ve said in the past and our current position:
The problem lies in how to properly define web managed switches vs. the unmanaged switches on one side and the managed switches on the other in order to enable meaningful vendor to vendor comparisons. It is nearly impossible to put web managed switches in 1 category only. On one side we have HP ProCurve’s web managed switches which have no CLI, no SNMP, no other management protocol, just an HP web manager. HP considers these unmanaged, and always reported these as such. These switches also follow the original idea behind creating web managed switches – having a product with unmanaged-like (read low) ASP with a little bit of management functionality. HP sells 1M+ ports of these switches each quarter, a fairly substantial number. On the other side stands Netgear whose web managed switches include majority of existing management protocols, their ASPs are in some cases even higher than Netgear’s regular managed switches and now Netgear even sells Layer 3 models. The lack of CLI is the only feature that Netgear uses to determine whether or not the model is web managed or not. These switches have all characteristics of managed switches and comparing these to HP’s web managed switches would be misleading. Other vendors fall somewhere between HP and Netgear. Interestingly though, 3Com’s Baseline Plus SFP we managed switch models even support CLI, so that is not a fool-proof distinction either.
Bottom line update:
While we are glad to size and even forecast spending on switches with this feature, we have deemed that publishing market share based on this feature does not represent a fair view of the market.
If you can’t comment, but you made it this far, please at least register your feed back in the poll. Thanks!
Petr Jirovský contributed heavily to this post.
Side note: Consider this an experiment that while isn’t offical IDC business is obviously heavily connected to what we do.