Our company has almost 800 customers at the moment and 4 employees and is profitable!
Charles Wu wrote: >> Once you get to say 1000+ customers, things like having the staff for >> service calls and time to repair for customers are often more important >> than the brand of radio or the original cost of the radio. We do spend >> more on payroll than radios, despite deploying lots of expensive gear. >> Keeping CPE prices down is appreciated and important, but less tangible >> ongoing management, troubleshooting, and repair costs must also be >> considered. The reduction in support costs isn't an expection, it's a >> reality and requirement in many situations. > > When you're working as a startup, labor costs are essentially zero (and if > you're asian like myself, you can call on your kids/relatives/grandparents to > work nights and weekends -- the classic Chinese restaurant business model =) > > However, when working with employees (and I don't care how smart / > hard-working / strong willed you are, there's still only 24 hours in a day) > -- labor costs become a bigger factor as the organization scales > > So this brings up a more interesting debate -- e.g., one-man band / > mom-and-pop vs. organizational strategy > > As an organization, trying to run a WISP with 700 residential customers is a > complete waste of time, however, as a one-man-band -- an 700 customer WISP > can be highly profitable > > The problem here is that there's a nasty chasm between what the one-man band > can handle and what an organization needs to support itself (e.g., it doesn't > scale linearly) > > The picture looks more like this > > 700 customers -- one-man band (or equivalent) -- highly profitable > > Then -- they start hiring employees to grow and scale the business > > Unfortunately, there's a minimum amount of overhead required, and what was > once a profitable business is now bleeding red ink and needs to reach 2,000 > customers before things get good again > > Which creates an interesting question -- if you're such a WISP, do you just > stop and sit tight at 700 customers? Or do you "go-for-broke" by trying to > grow? > > -Charles > > > -----Original Message----- > From: wireless-boun...@wispa.org [mailto:wireless-boun...@wispa.org] On > Behalf Of jp > Sent: Tuesday, January 05, 2010 10:36 AM > To: WISPA General List > Subject: Re: [WISPA] Wimax gear > > On Mon, Jan 04, 2010 at 05:28:49PM -0600, Wallace Walcher wrote: >> Having built my WISP from scratch with my own resources and currently being >> debt free in my operations, I often wonder who the people are who so quickly >> classify Mikrotik and Ubiquity gear as trash. I am making a very good >> living deploying such "trash". > > I'm not ashamed of calling their bluff when they say something is > "carrier class", and it's not even released yet and then has firmware > their either sets the timing wrong to the point of destroying the link > or doesn't do vlans, and the firmware isn't pulled offline because it's > the best stuff available. > > I've got a couple UBNT M links up and like them, and believe it has a > future. I just can't put my whole business on the line while they refine > a product. It is wise and irrestible to try the stuff though. > > I've got a downtown network of UBNT 802.11 gear, and the nanos and > bullets just can't handle the interference as I'd like. It was intended > to be an upgrade from the breezecom FH gear which was slow but reliable. > The UBNT is faster, but less reliable in the presence of local > interference. Now, if someone has an interference problem, we > immediately swap them over to Alvarion 5.4 gear. It is more expensive, > but we know we'll never have a service call after it's put in unless it > gets hit by lightning or the customer wants to upgrade. We would have > been wise to upgrade straight from the old stuff to 5.4. I'd still > recommend the UBNT CPE for truly rural use. > > Then MT is always making something wonky. A couple years ago, you could > crash the MT with a SNMP query. Now, if you put an N card in and upgrade > the firmware in your 433ah to 4.4, you might lose the ethernet ports. I > stay 1-4 months behind on their firmware because it's a mystery what you > might get. Changelogs show less than half of what they change. I do like > them for basic routing and also use their gear for a few links. I think > it's a step up from UBNT for ptp 802.11 based links. I also like MT > because it's pretty low power use, which has practical value for solar > sites or sites needing long battery backup. We don't have the time to > tinker to use it for everything. We tried 900 with SR9 then XR9 and the > reliability wasn't there compared to what we were accustomed to with > Trango and Alvarion. > > Once you get to say 1000+ customers, things like having the staff for > service calls and time to repair for customers are often more important > than the brand of radio or the original cost of the radio. We do spend > more on payroll than radios, despite deploying lots of expensive gear. > Keeping CPE prices down is appreciated and important, but less tangible > ongoing management, troubleshooting, and repair costs must also be > considered. The reduction in support costs isn't an expection, it's a > reality and requirement in many situations. > > A minor glitch that affects a few customers outside of town is not a big > deal, but if the glitch requires half a day on the road or requires > aircraft, boats, snowcats, or sleds, it could cost hundreds of dollars > and mess up a lot of customers. > > I'd fear for my welfare if everything was built on UBNT and MT though. > > We use Alvarion 900, 2.4 (not going forward), 5.4, 5.8, Trango (lots of > 900 installed, but not going forward), MT, UBNT, and now Solectek and > Radwin. > > My WISP is pretty low debt 100% privately owned and financed, and we > often choose higher end equipment. You do get what you pay for, but of > course there are diminshing returns the higher end you go. > > >> My perception is they are either people who are not spending their own money >> - they are working for the investor, or possibly borrowing or leasing the >> equipment, or they are a vendor promoting their own high margin goods. >> Those that are WISPs seem to have the perception that it is better to >> install higher cost equipment, no matter what the cost, if it will provide >> them an expected reduction in support costs. >> >> What I have found in my area is that people who deploy such equipment have a >> very hard go of it, mainly because the replacement costs during the storm >> season eat their lunch. My operational plan is different than some - I >> focus on residential customers on the outskirts of town that do not have >> access to Cable and DSL. Those focusing on business accounts in cities >> would understandably have a different perspective. But I feel very >> fortunate to have a sub $200 total CPE cost (sometimes sub $100) with the >> Mikrotik-type solution. >> >> >> -------------------------------------------------------------------------------- >> WISPA Wants You! Join today! >> http://signup.wispa.org/ >> -------------------------------------------------------------------------------- >> >> WISPA Wireless List: wireless@wispa.org >> >> Subscribe/Unsubscribe: >> http://lists.wispa.org/mailman/listinfo/wireless >> >> Archives: http://lists.wispa.org/pipermail/wireless/ > -------------------------------------------------------------------------------- WISPA Wants You! Join today! http://signup.wispa.org/ -------------------------------------------------------------------------------- WISPA Wireless List: wireless@wispa.org Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/
