MonkeyBrains: Silly Name, Serious Broadband

The lunch crowd at Coffee Bar, a stylish cafe in San Francisco's Mission District, dines with laptops open. It's a reminder that the city is a tech mecca.

To judge by the city's broadband options, San Francisco looks more like a backwater. The national average broadband download speed in the U.S. at the time this article was filed is 10.02 Mbps, according to Ookla's Net Index. The average broadband download speed in the city is 8.46 Mbps, unless you happen to live in certain subsidized housing projects.

At Valencia Gardens, for example, low-income residents typically enjoy connections of around 50 Mbps at no cost, thanks to a city-funded fiber connection.

But Internet connectivity for other San Francisco residents is improving, thanks to changing wireless technology and entrepreneurs who believe they can provide better Internet service than the likes of AT&T and Comcast. (The city is also hoping to accelerate broadband for residents with a federal stimulus funding request. Funding awards are scheduled to be announced on September 30.)

I'd come to Coffee Bar to meet Alex Menendez and Rudy Rucker, co-founders of MonkeyBrains.net, a wireless ISP, Web hosting and server co-location company, and Anders Finn, their sole employee. ("If you set things up properly, you don't need a huge workforce," Rudy explained.)

For years, I'd been paying a little over $40 per month for 6 Mbps down/768 Kbps up DSL from Cyberonic, a small ISP that resold Covad bandwidth. In reality, I seldom saw speeds above 5.0 Mbps.

The average monthly bill in the U.S. for broadband service in April 2009 was $39, according to a June 2009 Pew Internet survey.

When my Internet connection proved too slow for OnLive, a cloud-based gaming service, I began looking in earnest to see whether it would be possible to get something faster without breaking the bank. I also wasn't eager to sign up with AT&T or Comcast, the dominant telecom companies in the area.

Through Yelp.com, I discovered two ISPs in San Francisco that were getting rave reviews, WebPass and MonkeyBrains. I e-mailed WebPass and the founder promptly e-mailed back to say that his company doesn't do residential installations. WebPass focuses on large residential buildings and businesses.

After exchanging e-mail with Alex at MonkeyBrains, I soon had an installation appointment. It would cost $250 to install an antenna on my roof, but Internet service would be free for the remainder of the year, almost six months. The math worked for me.

The monthly price next year -- no contract -- hasn't been nailed down, but Alex says it will probably be in the $30 to $40 range.

Anders, who handles installations for MonkeyBrains, cheerfully risked life and limb on my steep roof to place the broadband antenna. After returning to following morning to make some adjustments, I was getting more than 20 Mbps down and up.

MonkeyBrains guarantees at least 10 Mbps symmetrical and most customers see 15-20 Mbps. Some customers see 25-30 Mbps with a good connection. According to Alex, this depends on the company's backhaul connection, which is in the process of being upgraded. Once that happens, he expects 30 Mbps+ connections will be common.

I could've gotten, at least on paper, 20 Mbps down and 4 Mbps up from Comcast for $55 per month, but I'd prefer to give my business to a small, local company. As someone who already pays too much to AT&T for home voice and mobile service and Comcast for cable TV, I'm looking to give these companies less rather than more money.

This turns out to be a common theme among MonkeyBrains customers.

"We've had various customers tell us during installs that the Comcasts and AT&Ts of the world are ‘big billing systems,'" said Alex. "If you want anything else, like you want to call someone and ask a question, our company lends itself better to that. And unlike Comcast or AT&T, Rudy, Anders, and I are not looking to take over the U.S. We're looking to do really well in our area and if other people want to learn how to do this, we would love to help them do that."

MonkeyBrains has been around for 12 years as a server co-location facility. Thanks to a new generation of affordable wireless technology and the tepid efforts by large ISPs to improve broadband speed while lowering prices, the company is becoming a wireless ISP, or WISP.

Wireless ISPs solve the last-mile problem, the speed bottleneck between fiber optic Internet connections at telecom service providers and the businesses and residences where Internet users work and live.

"About a year and a half ago, we had access to a fiber line, back to our data center, and we started talking about WISP -- doing wireless shots back to this fiber jump point," said Alex. "The first proof of concept was my house...and we started playing with different antennas. The first couple of times it didn't work. And then finally we got it online. At my house I started using it as I usually do, for SSH sessions and Netflix, and it was working great. The throughput was amazing. When there was no one on our network I was getting 40 Mbps symmetrical at my house."

"One of the big breakthroughs," said Rudy, "was these cheaper antennas that are out there, that are outdoor and waterproof and they have little Linux embedded systems so they're capable of doing all kinds of different functions."

Thanks to FCC rules changes in recent years, unlicensed 5.8 GHz outdoor antennas with respectable through-put are now available for a few hundred dollars. At higher price points, equipment costs make it harder for wireless ISPs to compete with DSL or cable.

Alex and Rudy had looked at antennas from a number of manufacturers -- Redline, DragonWave, EnGenius -- but none of them proved suitable. Then they tried an antenna from Ubiquiti Networks.

"Ubiquiti really nailed it and that really allowed us to do this project," said Rudy.

It turns out that MonkeyBrains isn't the only company that sees a future in wireless broadband.

"We certainly have been approached by several companies with that sort of model or business plan recently," said Barry Fraser, telecommunications policy analyst in San Francisco's Department of Technology, in a phone interview. "It's certainly something that has generated a lot of interest among ISPs. It's less expensive than trying to lay cables everywhere."

MonkeyBrains currently has about 100 antennas, each of which may serve one or many customers -- an average of 5-10 per antenna, Rudy estimates. It's adding about 10 customers a week. One of its largest clients is KQED-TV, San Francisco's public television station. The Electronic Frontier Foundation is another.

MonkeyBrains uses the revenue from its large business customers to subsidize the free bandwidth it provides to the community.

"With all ISPs, once you get a customer, the bandwidth doesn't really cost anything," explained Rudy. "A lot of the costs are service, when something breaks, having to do a service call. If the equipment is working, it's almost all profit for the ISP. In that regard, we can sell it for zero dollars, and we're not really shooting ourselves in the foot."

Though other ISPs might argue that bandwidth is a real cost, MonkeyBrains does give Internet connectivity away for nothing. It maintains about 40 free WiFi hotspots in San Francisco's Mission District. Users joining the network see the SSID "MonkeyBrains.net" and some have become customers. It's an unintended form of marketing for the company.

MonkeyBrains is building a network composed of a bunch of overlapping directional antennas. Some of its customers have repeating antennas that provide service to other customers, if they're in optimal locations. The MonkeyBrains network is similar in some respects to what Meraki tried to do in San Francisco several years ago, though Meraki used omnidirectional balloon nodes as opposed to crisscrossing directional transmissions.

Alex says the Meraki comparison only goes so far, noting that Meraki was trying to build a mesh network while MonkeyBrains is building a core directional network with reliable nodes.

"The reason Meraki failed (beside not having a good business model) is that the system relied on having a usable node nearby, which was not always the case," he said in an e-mail. "In the MonkeyBrains WiFi network, each subscriber points at the closest reliable node. Some subscribers who have stable links to a reliable master node often have an open access omni toggled off of it for the surrounding area."

The networking model combines point-to-point connectivity for business customers and point-to-multi-point connectivity for residential customers. Alex concedes that point-to-multi-point isn't popular because of the potential for instability but he insists it's working well so far.

Rudy believes that organizations like KQED could easily become their own local wireless ISP, to provide free Internet connectivity and access to the content they produce.

"We're trying to evangelize the idea that anyone can be an ISP," Alex chimed in.

This is more of a philosophy than a plan at the moment, but the brains behind MonkeyBrains see a time when Internet broadband is free, or at least much less expensive, than it has been.

"People are so used to paying for the Internet, they're so used to paying for bottled water, for all this other stuff, when really it's something that -- maybe it shouldn't be free, but it should be closer to free," said Rudy.

Alex and Rudy go back and forth on this point, with Alex saying that Internet connectivity "should be free on a certain level."

And so it is: MonkeyBrains makes shared connections available to the community and users get a few Mbps at no cost. For faster connections, you pay.

Rudy says that if you wade through the contracts offered by large ISPs you'll find that they limit your ability to share your connection.

Alex added, "We're definitely more state-of-nature. If you can figure out how to resell or retransmit our signal, that's on you."

Hedging that statement, Rudy suggests that he'd probably approach anyone making serious money that way to discuss a revenue sharing arrangement.

MonkeyBrains is still a business after all, albeit one that's defiantly local and community-oriented.