Call Us: 602-426-0542

Chapter 34: Tales from the Towers – The Beginning of the End......

I’ve seen business plans for WISPs from small to large, and from municipal mesh to tower-centric models.  In every one, I try to look at design methodologies that would make the deployment unique technically.  The goal is that th effort pays off in reduced capital expenditures, competitive advantage, and hopefully an overall better financial business model.  One of the problems with being an engineer though is that we rarely assume we are wrong unless someone proves that we are. 


I engineered a system a year ago that I was sure was a solid design.  The financial model fit, the ROI was good and the end game looked great.   Then I was asked to reduce the CAPEX (Capital Expense) as much as possible.  After much teeth gnashing and some soul searching, I realized that my design was not based on the idea of maximizing the profitability of the investment and that my end game wasn’t in sync with that of the investors.  I didn’t think through the fact that the original model hadn’t taken into account multiple levels of future investment funding and how to keep the highest percentage of stock control.

When designing a business plan for a WISP, the first thing to consider is simply this, “what is your exit strategy”.  If you can’t answer that, don’t even start.  If your only answer is, “I’ll sell it to someone”, good luck with that statement to possible investors.  Most WISPs are only worth about 1-2 times annual revenues.  It’s a highly specific business with a limited number of potential buyers which makes it a buyer’s market.  Throw in the fact that many WISPs are small companies with single owners; then the dynamics of the sale are based as much on personal issues as financial.

Most business plans I do are based on the fact they probably won’t get very large and will maintain a single owner.  Many owners look at it as a small business that should eventually reach a comfortable profitability level.   If the business can be sold when they are ready to get out, that’s even better.  This type of design can focus on a solid infrastructure with a fairly linear growth plan.  Corners don’t have to be cut for the sake of minimizing outside investment.  You only need to insure that the investment level is sufficient to carry the company to profitability.  These are the easiest types of business plans to work with.

Developing a WISP business plan that will only have to deal with a single round of funding but has to take into account investors, ROI, and payback schedules is not much different than a sole proprietorship from a technical  point of view.  There is still the assumption of linear growth with more exit strategies but that’s on the financial side.  You don’t cut corners on infrastructure if you know that that it has to be upgraded within the first year or two to support planned capacity needs.  Since expected growth is still based on a more linear model, labor resources that are available in startup may not be available when deployments start cranking up without adding additional staff.  The problem occurs with this model if growth exceeds expectations and the needed capital wasn’t included in the original funding model.  Single owner businesses typically just hunker down, put in some extra hours, maybe apply for some credit with a bank, or just simply dig into their pockets further if possible.  It’s also when many of us quit accusing our mother-in-laws of being a member of the local Coven since they may be the only option.

Single round funded businesses also turn to partners to support the business.  Hopefully the money is made available as a loan, and doesn’t cost partners additional equity.  However, if the business grows more slowly than expected and additional capital is needed, someone may end up owning more of the company.  That is not a position an entrepreneur wants to find himself in.  He may find that his exit strategy may occur much faster than was planned.  If they decide things aren’t going right and they have to partner up more money, the entrepreneur may also find himself looking for work elsewhere if he isn’t critical to the operation. 

The exit strategy of this kind of model is usually based on a sale of the company when it hits certain successful financial and resource milestones.  Investors are in the business to make money and the faster they can do that, the faster they are usually going to get out and move onto something else.  It’s rare for First Round angel investors to try to hold onto all the equity they have unless they see something very special in the company.

Multi-round funding is another animal entirely from a technical standpoint.  Even though first, second, and possibly third-round round investors are obviously out to maximize their investment, each group has different priorities from a capital expenditure level.    For example, a First Round or angel investor wants to control as much stock as possible for as little money as possible.  Even though that sounds pretty obvious, it dictates a different technology design.  The goal of a First Round of investor is to create the image of a company with great growth potential for the Second Round investor.  Second and Third round investor simply want the highest and fastest return on their money.  Think about that for a few minutes.

Let’s say we design an infrastructure capable of supporting 3000 users at an average of 4Mbps.  This design costs $800,000.  If we use a 12-1 user to bandwidth ratio, that means we need about 1 Gbps to support that number of users.  Assume it take about 18 months to build out that many users with a linear growth pattern.

Part of the cash flow analysis of this model includes $150 for installation fees.  An installation probably averages about $250.  For this analysis, let assume $50 per month for service.  If you are installing 200 users per month from startup, then your expenses are about $50,000 per month for installations and your income is about $30,000.  Your cashflow per month for these installations is about $20.000.  Of course you are getting an additional $10K per month in revenue per month starting the next full month after an install.  The point is that it will take at least 10 months to become profitable for most business plans that I’ve worked on.  It then takes about 24 months to completely recover all of the original CAPEX expense.  We will get back to this later but keep it in the back of your mind. 

Since the first round is typically angel investment, it’s the hardest money to raise.  Startup money is the highest risk investment anyone can make in the funding of a business.  It also means that the goal of the first round of funding isn’t to make the best technical design possible or a design built for years of expansion; it’s to make the company as attractive as possible for Second Round investors.  As an engineer, that is a very difficult concept to design for. 

We start by looking at our $800,000 startup costs and then try to figure out two basic things.  The first is to reduce the CAPEX.  Here is where it gets interesting if you are the person designing the network.  You aren’t reducing the CAPEX just to make it easier to find funding, but you are also reducing the CAPEX to get the Second Round investors to jump in and still leave as much equity as possible for the First Round investors.  This means that you may not be designing for a full deployment but may be designing for an interim deployment that is more like a proof of concept.  This is a critical concept to understand if you plan on creating a company that can grow into something that might go public someday as an exit strategy.

For example, $800,000 builds out a system that can support 3,000 people.  Breaking down the $800,000 shows that $200,000 of the original CAPEX is a wireless tower based backhaul system.  This connects your coverage area to a Competitive Local Exchange Carrier (CLEC) that may be located 50 miles and 3 tower hops away.  The cost of bandwidth there is about $5 per MB for a Tier 1 provider.   This bypasses local loop carriers thus saving thousands of dollars every month.  The cost of bringing 1Gbps of bandwidth in over local loop fiber is $13K per month which is about $5K more than using the towers for backhaul if we include tower rental costs.  Basically it will take over 3 years to recoup the capital expenditure but the CAPEX has dropped by 25%.  We just simply move the $5K to the operating expense side. 

But what if we could cut the $13K per month even further to avoid that extra expense?  Additional price quotes from other internet data-circuit providers shows that we can get 100Mbps for about $3,000.  Compared to $13K for 1Gbps, it’s not a great value but that is not what’s important.  What is important is that we are saving an additional $10,500 per month for the first few months or until we swamp the 100Mpbs circuit.   

The problem is that you know that circuit is probably going to be insufficient within 4 months.  You can add an additional 100Mbps for about $1,500 per month so at least your egress bandwidth can grow as needed.  However, expansion of the system was partially based on the fact you would have equipment on 3 different towers between your target area and the data center located in a major metropolitan area.  In addition to losing at least 1 extra tower in the area you were going to cover, the tower locations would also allow you to branch off into the next area.  Instead, you are going to push that $200K expense down the road.  Then instead of a $4-$5 dollar per month per Mb charge in a data center, you will have to get another local loop circuit at a cost of $13-$26 per MB to cover the next area.  Now the real race begins. 

If the next round of funding comes in before you start that expansion, the CAPEX of the towers can now be included into second round of funding and no problem.  If not, you use local loop fiber there also, double the back end infrastructure that is used for system management, and keep moving forward.  Then you hope the next round of funding comes in before you have to expand into your third area. There are many other engineering compromises on the back end that you will also re-evaluate in a multi-round funded model but this one example shows how engineering is affected by the business model.  Since this article is getting really long, I’m going to skip all the other areas, tower by tower, that also have to be evaluated under the same criteria.

We have now saved $200,000 and an extra $10,000 per month in expenses at startup.  That money could now be moved to sign up clients faster.  For example, going back to our installation costs, we remember the difference between $150 for an installation fee and a free installation may be the difference between signing up 200 customers a month and 300 customers a month.  The recurring revenue is what is important to a Second Round investor.  The more clients you have, the more attractive the company looks and the more money can be raised in the next round.  Clearly free installations will attract more clients.  However, you will also lose $30,000 a month in revenue to do that.  That is First Round capital that has to be available that moves from CAPEX to OPEX (operational expense).  This means that the $200K you were going to put into backhaul, now just got moved to signing up customers to get the Second Round of funding faster.  That’s what is attractive to an investor and what you need to be engineering for.

Some of you may say that this really isn’t compromising your design and may just look at it as good business sense.  It is for a multi-round funded model but it puts more pressure on the funding in the first round.  It’s also lower risk than building out a big infrastructure and then not be able to hit target sales numbers.  However, its short term engineering that eventually has to be updated down the road.  In a single round funded model, not putting this infrastructure in may result in not calculating enough investment that they then may have to raise later.  This is also not good as they might not be able to raise it with a tight cash flow position.  Underfunding is as bad as over-engineering, so pick your poison.    

Business managers and investment bankers lean towards the highest profit and greatest return on investment.  In most cases, they may or may not understand the engineering but they do know how to present and position a company to the people who are putting up the money.  Those people aren’t engineers either.  Both sides may be victims of the Dunning-Krueger affect and it takes a lot of work to get those sides together.  If there is respect and an understanding of the roles that each need to play, then it can be a winning combination for an Entrepreneur.  If either side doesn’t grasp the goals of the other side, then it’s going to be a difficult and testy partnership.  It’s important that both sides understand the goals and the funding methodology that will be utilized to achieve those goals.      




Leave a comment

Make sure you enter the (*) required information where indicated. HTML code is not allowed.

We are proud to support the efforts of One Laptop Per Child (OLPC)