Why owning physical infrastructure can be financially detrimental

We’ve received a question from someone who understands the telecommunication industry well. We felt that the question reflects some of the concerns raised by part of the industry, so it would make a great subject for an article.

The question went as follows:

If you invested 100-150K in a system, paid it off over time and now run at nearly zero cost with up to date software, why would you change?”

And it is a very good question because a lot of established businesses have invested heavily into their telecommunications infrastructure and feel they are now only starting to see a return on that investment.

To answer the question I’m going to approach it from an economic standpoint, as this is one of the most important decision-making factors for when purchasing/upgrading equipment for CXOs and directors.

Let’s talk Economics

There is a well-known economic term that we feel fits the situation mentioned in the problem. The issue is related to the idea that the money has already been invested in the existing system and we feel it’s a good example of Sunk Cost Fallacy.

As many readers will know, a sunk cost is a payment made that is not able to be recovered. Investment is telecommunication hardware is a great example. Once the money has been invested, it’s gone and shouldn’t be a consideration in future decisions.

But that’s not the reality. The reality is we are emotionally invested into the investment we’ve made in our current system and the feeling of waste drives our future decisions when in fact, it shouldn’t be a factor.

We go on vacation, and it’s terrible. The weather is lousy, the service poor. There are cockroaches in the room. But we suffer on through the rest of the vacation, making the most out of it, because we’ve already paid for it.

If we had not fallen foul of the Sunk Cost Fallacy, we might have returned home, done some of the things we always want to do but never find the time. Visit some places, see some friends we never seem to find the time to see. We probably would have had a much better time.

So, in answer to the question posed to us. The 100-150k investment is gone. It doesn’t necessarily mean you need to convert immediately to a cloud-based system, but certainly, the existing investment in infrastructure should not be an influence on the decision.

The endowment effect

People have a tendency to prefer to hold on to something they already have ownership of, rather than acquire something they obviously don’t own, that is more effective and more affordable.

A hardware system will become increasing deficient in performance and will have increasing costs in repair, replacement and personnel compared to a cloud-based system, which doesn’t require the same investment in infrastructure and maintenance.

You are prone to keep with what you have, even though it might not be the most cost-effective solution, especially when you can’t see past Sunk Cost Fallacy. It’s financially detrimental.

Finally, it’s possible the person asking the question is affected by Loss Aversion. They have made an investment in their infrastructure and is pulled more towards their tendency to avoid loss than the potential to acquire gains. There’s nothing we can do but try to identify when we are allowing Loss Aversion to dictate our decisions.

Economies of scale

If you go out to sea in a small boat with a fishing rod and catch a few fish, the cost of the boat hire, your time and everything else is very high in proportion to the fish you catch, this is your cost per unit. If you own or buy from a big fishing company, the cost efficiency is increased and thus, brings the cost per unit down.

For a small or medium sized telco (50 servers max), the cost you are likely to be paying for servers is high, even if you get a good deal. However, if you buy from a cloud hosting company, you are leveraging the economies of scale in your favor. Even the smaller cloud hosting companies have 10,000+ servers, so they will be buying those servers much cheaper than you can. The same principle applies to ongoing costs as well, such as power and transit.

In short, anything that is deployed in scale is more cost effective. Furthermore, high quantity, doesn’t mean poor quality. Look at the way cars and phones are made.

Cloud solutions are flexible, cost effective and high in quality (when deployed correctly), they offer telecommunications companies so many benefits, that unless there are immutable facts as to why you can’t utilise these technologies (such as a USP), then you are one of those irrational individuals swimming against the tides of economics. In other words, throwing money away.

The Cost of High Availability

If you are a half-serious telco, you know you can’t operate with just 1 switch. If you are a serious telco, you know you can’t operate in just 1 datacenter. If you are a truly global telco, you have POP’s in each continent. But what about the costs? If you are renting space and have your own servers, there are huge amounts of expenditure going to waste due to the cost of running it yourself (in comparison to large cloud companies), unless there is a specific reason why you need physical equipment.

More importantly, what are the costs of running your own hardware AND achieving five nines (99.999% uptime) availability?

99% is just not good enough. That would be just over 7 hours of downtime (per year).This can be delivered by a single server but it is really not telco worthy.

99.9% gets it within the hour (per year), which is more reasonable. It’s possible to deliver this availability by a single server, but you will run into upgrade and maintenance problems eventually.

99.99% now we’re down to about 5 minutes. Here you really need a multi-server solution, even if it’s a cold standby.

99.999% is just under 30 seconds. If you are lucky, you will deliver this from a single datacenter, but I wouldn’t put any money on it.

The problem here though is cost. I can literally throw a server on budget hosting (even at 5 nines availability), for about $50 / month. I could technically throw up another 2 or 3 in different places and create pseudo availability. From an administrative level, however, information would not be synchronized correctly, they would be 3 separate systems and would require additional human hours to compensate for a lack of a well-designed solution.

You could do the work yourself, reinvent the wheel with lots of useful tools, learn the skills and support the infrastructure, which is great if you are a DBA & Network architect and security professional and you can justify your time spent on this!

But ultimately there is a non-linear cost increase when you increase the amount of 9’s that you require in uptime. Therefore, it’s a great benefit to use systems which can inherently scale horizontally to avoid unnecessary costs.

Maintenance Costs

Another key point which cannot be ignored is maintenance costs. This task should not be taken lightly. We are not just talking about creating backups, archiving and purging of the database, but replacement hardware as it runs out, security updates, regular penetration testing, software upgrades/patches. If you are not regularly doing all of these tasks, then you are doing it wrong and leaving yourself at unnecessary risk.

Not strictly maintenance, but what happens if you wish to expand a component of your system? You may need a skilled engineer in that field. There is a possibility that the language used or the product range of your switch will grow, but there is a greater risk that the number of engineers specifically supporting your make and model of hardware will decrease.


With few exceptions, all the assets that you own will suffer from financial depreciation and it is not different with hardware. Furthermore, who will want to buy physical equipment that is out of date and costly to run?

Enter The Cloud Solution

As humans, we often make irrational justifications for things that seem perfectly logical, but when we become aware of the better options, we can see our irrational decision, reflect and make better choices.

Look at the way we now acquire entertainment. We no longer feel the need to own our favorite movies when we know that they, plus more, can be delivered as a service to our TV screens in seconds. We can benefit from this service architecture and all of its advantages too.

The cloud is not something to be afraid of. When implemented correctly, it has intrinsic characteristics which naturally solve core concerns, such as effortless in country deployment, management & resiliency. Overcoming these problems with a cloud solution gives your telco a competitive edge.

ConnexCS is a cloud softswitch, with zero cap-ex, relieving you from your sunk cost burden moving forwards. We can’t promise that the endowment effect won’t come into play because we know that you’ll love our product!

We sit on the shoulders of giants and benefit from the economies of scale delivered through cloud computing & resource centralization (deployed in geo-redundancy of course). Our system is built to natively sit on the new era of technologies, making redundancy just an emergent property that is effortlessly deployed.

Like to know more about ConnexCS and how we can improve your telecommunication operations? Speak with me via private message, leave a comment below or visit our website: 


Leave a Comment

Your email address will not be published. Required fields are marked *