WAN Carriers Need Better Bandwidth Pricing Models for Enterprises

This Paying All the Time Nonsense is so 1990s

Two weeks ago I wrote a blog asking "should enterprise IT provide great Internet access?". Essentially, enterprise WANs usually give 3-10 Mbps to a field site and 100-200 users need to fit inside that bandwidth. Meanwhile, these users go home at get 20 Mbps for themselves via cable modem (please no e-mails about cable modem sharing, blah, blah, blah...the fact is I can download at up to 20 Mbps at home). QoS is used to protect VoIP, Video, and critical applications inside this 3-10 Mbps, but this doesn't leave much room for everything else. We have a very robust 5-level QoS model. But the first three queues are VoIP, Critical Applications, and Video. That puts all the rest of the traffic (about 95% of it during a normal day) in the bottom two queues to duke it out. Someone is going to piss someone else off; it's just guaranteed. Then the users start to complain and they are sometimes not complaining about slow Facebook, it's actual enterprise applications that are slowing down. These applications are not in the "Critical Queue", but they are important. Nothing we can do there. Then we get the users e-mailing us and coming up with their own ideas to fix the problem. Oh great. My favorite one was to begin rate-limiting individual TCP sessions using a specific protocol after a certain period of time that the router determines has been too long. Sure, let me look into the Cisco IOS commands for that. So, that brings us to the real heart of the problem. We can't buy more bandwidth from our MPLS carrier because we have to pay for it ALL THE TIME. Weekends when no one's there? Yep. How about overnight? Yep. 12PM when traffic dies down for a while? Send in that check! So the pricing model, and thus the cost model, requires us to keep the bandwidth low. If we could get a huge pipe and pay for what we use, it would be better. Carriers will argue this is because bandwidth is a perishable asset which can't be inventoried (you can't store it up to use later, say, like disk drives), so they have to charge you constantly. To that, I say come into the 21st century! If Apple can give me all the computing power in an iPad, some Finance geek can come up with a better pricing model. The dirty secret is the current pricing model is a cash cow for the carriers. You don't use the bandwidth 75% of the time and they charge you constantly for it. How nice! Thus, I am not satisfied with the bandwidth operating model we live in today with WAN carrier services (MPLS, private line, Internet, take your pick). Users are getting 20 Mbps at home for themselves and then they come to the office and we squeeze 200 users into 12 Mbps. Even with QoS protecting critical apps, it is not a sustainable model to keep telling our field sites “you have enough bandwidth”. They are stepping on each other doing downloads and Internet browsing and it’s angering them. It makes IT appear aloof and unhelpful. In a way, I could order 4 cable modems at the office, get 100 Mbps speed, pay the same price, and just put up some IPSec tunnels. 95% of the time the site would be much happier and more productive. I'll figure out something to do with the VoIP. So, I want some brainstorming from the carriers on next-generation network bandwidth and pricing models. What we have now is last decade stuff. I want innovation. I want thinking outside the box. For example:

  • Burstable circuits with strict reporting and a huge 100Mbps pipe.
  • Time-of-day charging (much less expensive during off hours). Do I dare say "Nights and Weekends free"?
  • A huge 100Mbps Ethernet pipe with charges per megabit with strict reporting.
  • "Family Plans" - bandwidth pools across the entire service contract that can charged against instead of the current per-circuit fee.
  • "Call Friends for Free" - a list of 20-30 destinations that you get free bandwidth to (free bandwidth to Google and internally to your primary SAP server).

Wireless carriers started to realize 10 years ago they could differentiate their product line based on pricing models. WAN carriers that serve enterprise customers need to do the same. I am bored by the carriers that call me and want to talk about their great, new OC-12000000000 undersea cable from China to Antarctica. Yawn. The next one that calls me and says we'll give you 10x the bandwidth for the same total cost I pay today and make it flexible in how the bandwidth is used will immedietly get a return call.

More >From the Field blog entries:

I Got an iPad

Is Enterprise IT Here to Provide Great Internet Access?

Latency in the Data Center Matters

Net Neutrality for Dummies

Cisco Nexus 5000's Poor Results in Data Center Switching Test

Cisco's Overlay Transport Virtualization (OTV) is New, but has Potential

  Go to Cisco Subnet for more Cisco news, blogs, discussion forums, security alerts, book giveaways, and more.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.

Copyright © 2010 IDG Communications, Inc.

SD-WAN buyers guide: Key questions to ask vendors (and yourself)