Tag Archives: economy of scale

Economy-of-Scale for ISPs

I’ve worked with a number of small communities that want to explore the idea of having a community-owned ISP. My advice to small communities is the same as with all clients – economy-of-scale really matters for ISPs.

Economy-of-scale is the economic term for describing how businesses get more efficient as they get larger. It’s fairly easy to understand, and the classic example is to look at the impact of the salary and costs of the general manager of an ISP. Consider the example where the all-in cost of salary and benefits of the general manager is $200,000 per year. If the ISP has 200 customers, that cost works out to $82 per month per customer – obviously impossible to cover. At 2,000 customers, that’s a cost of $8.33 per month per customer – probably doable, but a big strain on being profitable. At 20,000 customers, the cost is reduced to $0.83 per month – a cost that is easy to absorb. With a customer base of 200,000, this cost is only 8 cents per customer per month.

A large percentage of the costs of operating an ISP are fixed or nearly fixed. Any fixed cost acts in the same manner as the general manager’s salary. The larger the size of the ISP, the easier it is to cover fixed costs and the better the chance of being profitable.

It’s possible for a small ISP to break even, but doing so requires the operator to be extremely frugal – any unexpected expense can throw a tiny ISP into a loss. Operating a small ISP of a few hundred customers is best described as a labor of love – because it is not going to be profitable for the owner.

For many years I’ve said that the bare minimum number of customers to enable an ISP to be full-function was around 2,000. That’s enough revenue to cover the labor for a few employees, along with other operating expenses. An ISP with 2,000 customers still needs to keep an eye on expenses because it doesn’t take much to tip the business into losing money. However, with current inflation, I think the minimum size to be effective has probably grown closer to 3,000 customers.

In building business plans, I’ve always seen the real benefits of economy-of-scale kick in around 20,000 customers. That’s enough customers to be able to operate a full-function ISP that can deliver superior customer service. There is enough revenue to hire all of the needed staff and pay them well, including good benefits. ISPs smaller than 20,000 have to forego some of the benefits that come with size.

Interestingly, economy-of-scale doesn’t scale forever. In my experience in the industry, I see that ISPs of a certain size start getting less efficient. It’s going to be unique to the specific company, but ISPs larger than 200,000 or 250,000 start being less efficient.

I’ve always credited this with the phenomenon where the span of management control gets too broad. At some size, the core management team doesn’t know what is going on at the street level. When that happens, it’s inevitable to start seeing bureaucracy creep in, which is the curse of the giant ISPs. Local or regional management starts determining policies, often to the detriment of the overall business. The largest companies manage through a process of pitting regions against each other to earn bonuses. That’s far different than running an ISP with 20,000 customers. Almost all of the horror stories we’ve heard over the years about the poor treatment of Comcast customers can be attributed to regional managers who made cuts or implemented policies that benefitted their bonuses rather than benefitting the customers.

Try as they might, the giant ISPs are never going to have the same level of customer service as the smaller ISPs they compete against. Witness the many decades of duopoly competition between Comcast and Verizon FiOS. They are both big companies, and customers don’t love either of them in the way that I see customers being loyal to smaller ISPs.

Of course, size isn’t everything, and there are small ISPs who are terrible at the day-to-day operations of the business. Economy-of-scale refers to the scaling of costs and has nothing to do with the philosophy of how an ISP treats employees or customers.

My advice to any ISP with under 5,000 customers is to consider how much easier it would be to operate the company if it grew to 10,000 or 20,000 customers.

ISP Economy of Scale

I’ve worked for numerous small communities over the last few years and the first question they always ask me is if their community is large enough to support a standalone fiber ISP business. What they really want to know is if they can somehow operate their own local ISP and still have affordable broadband rates.

The question about how big an ISP must be (in terms of customers) is really a question about economy of scale. The textbook definition of economy of scale is a business where costs decrease per unit through increased output. The ISP business is clearly an economy of scale business since the cost per customer decreases as an ISP adds more customers.

The example I usually use to demonstrate this is to look at the cost of paying for a general manager for an ISP business. The cost of the general manager is what economists would call a fixed cost – it doesn’t vary as you add or lose customers. However, the amount of the general manager salary that must be covered by each customer gets smaller as the customer base grows larger.

A large part of the costs of operating an ISP are fixed like the GM salary. Costs like operating a business office, doing the needed accounting, buying a billing system, etc. are largely fixed. The largest fixed cost is often the debt service required to pay off the cost of building the network.

I have done hundreds of business plans for communities of all sizes and I have developed a few rules of thumb for operating a traditional ISP – one that has the expected number of employees, charges normal industry rates and has to finance the cost of their network.

  • It’s hard to justify a new standalone ISP with fewer than 2,000 customers.
  • Economy of scale kicks in at that point and the business gets more efficient, when measured on a per customer cost up until an ISP reaches between 20,000 and 25,000 customers.
  • After 20,000 customers the cost curve reaches relative stasis – adding customers increases efficiency but also drives additional fixed costs. For example, companies find they need to hire extra accountants; they might need backoffice positions for functions like personnel or benefits management.
  • At some fairly large size, say 100,000 customers, ISPs tend to become less efficient. Large companies tend to become bureaucratic, hire significant middle management and become less functionally efficient as centralize functions and put them into silos.

There are ways defeat the economy of scale curve to some extent and I have clients who used the following ways to be more efficient than other ISPs of the same size:

Reduce Costs. There are various ways to spend less on needed functions than the average ISP.

  • Use existing excess capacity. A City or an electric company can open an ISP and not have to spend money for a business office since they are already operating one. An existing business like an electric company or a telco might be able to use underutilized customer service reps or line technicians without having to hire a whole new staff for a new ISP venture. A utility or city might be able to use the existing billing systems from their water or electric utilities to support the ISP functions.
  • Reduce Expectations. Small ISPs often can save a lot of money by reducing customer expectations. For instance, they might elect to not have repair service on evening and weekends unless there is a real emergency. Small ISPs also don’t need to offer all of the bells and whistles of larger ISPs and can have a simple product offering. A small ISP might elect to not offer cable TV, which for small ISPs usually has a negative margin. This concept can only be taken so far and works best in communities where an ISP is competing against a decent giant ISP.

Avoid Costs. It’s not easy to avoid the cost of being an ISP, but it can be done.

  • Financing from Other Revenues. I have a client with only 600 potential customers that is successful since they decided to fund their network using property taxes rather than having to pay debt from ISP revenues. This allows them to have broadband rates far lower than surrounding communities by avoiding ISP debt payments. I know municipal and cooperatives power companies that have raised electric rates to help cover the cost of an ISP business since that’s what their customers wanted.

Grow Efficiently. The most obvious strategy to beat the economy of scale curve is to get more customers in an efficient manner.

  • Increase Penetration Rates. For a small ISP the difference between a 50% and a 70% penetration rate can be dramatic. Many ISPs become complacent once they generate enough cash to pay the bills while the smart ones continue to sell hard every year to maximize the customers in their footprint.
  • Expand to Another Market. Most small telcos figured this out years ago and operate a CLEC business outside their regulated footprint.
  • Partnering. There are several examples today of cities that have banded together to create a larger ISP. There are numerous cities and electric cooperatives that are partnering with telcos to gain the economy of scale. We are still seeing small telcos getting purchased by larger ones.