In summary<\/p>\n
- \n
- There is no price per extension.<\/strong> The cost is divided into central, terminals, installation, traffic and maintenance \u2014 and proposals rarely separate the five components.<\/li>\n
- Year-on-year comparisons are always misleading.<\/strong> On-premise and cloud solutions typically cross paths between the third and fifth year.<\/li>\n
- In per-user models, the bill grows with the hirings<\/strong> without anyone making a new decision.<\/li>\n
- The most frequent hidden cost is oversizing of simultaneous channels<\/strong>.<\/li>\n<\/ul>\n<\/div>\n
\nIn this article<\/p>\n
- \n
- The five cost components<\/a><\/li>\n
- What factors weigh into a custom solution<\/a><\/li>\n
- What weighs in on a cloud solution<\/a><\/li>\n
- How to compare proposals<\/a><\/li>\n
- Hidden costs<\/a><\/li>\n<\/ol>\n<\/div>\n
\u201cHow much does a telephone system cost?\u201d is a question without a useful answer, for the same reason that \u201chow much does a car cost\u201d doesn't have one. But there is a cost structure that always repeats itself, and knowing it allows you to read proposals with different eyes.<\/p>\n
The five cost components<\/h2>\n
Regardless of the solution, the total invoice is always broken down into the same five parts:<\/p>\n
- \n
- A central<\/strong> \u2014 licensing or subscription, normally per extension or per simultaneous call.<\/li>\n
- The terminals<\/strong> \u2014 IP phones, headsets, computer and mobile applications.<\/li>\n
- The installation<\/strong> \u2014 configuration, numbering plan, messages, opening hours, number portability.<\/li>\n
- The traffic<\/strong> - what is paid to the operator for calls and the SIP trunk.<\/li>\n
- Maintenance<\/strong> \u2014 updates, support and changes over time.<\/li>\n<\/ul>\n
Many proposals present only two or three of these items, which makes comparison with another proposal impossible without additional work.<\/p>\n
\nAlways ask for the five separate instalments.<\/strong> A proposal that bundles everything into a single monthly fee could be brilliant or terrible \u2014 there's no way of knowing without breaking it down. And a supplier that resists separating them is telling you something.<\/p>\n<\/div>\n
![\"Relative](\"https:\/\/www.dataroad.pt\/wp-content\/uploads\/2026\/08\/dataroad-custos-central-telefonica-empresas-1024x576.webp\")
The relative weight of each budget line in a telephone system project for an SME.<\/figcaption><\/figure>\n What factors weigh into a custom solution<\/h2>\n
In an on-premise system, the investment is concentrated at the beginning.<\/p>\n
O licensing<\/strong> It typically scales with the number of extensions, but what usually comes as a surprise is the cost of additional modules \u2014 call recording, call queues, CRM integration, reporting. They are almost always extra.<\/p>\n
The telephones<\/strong> They are the most underestimated part. The difference between a basic model and one with a screen and supervisor keys seems small until it is multiplied across thirty stations.<\/p>\n
Our Installation<\/strong> It varies greatly with the complexity of the numbering plan. A company with simple customer service gets set up quickly; one with queues by department, varied schedules and overflow to mobile phones requires real configuration work.<\/p>\n
Our annual maintenance<\/strong> it's the item that most often disappears from proposals \u2014 and the one that ensures the control panel receives security updates.<\/p>\n
What weighs in on a cloud solution<\/h2>\n
Here the logic is reversed: smooth entry, constant cost.<\/p>\n
Our user fee<\/strong> dominates everything else. It is what makes the solution affordable in the first year and what makes it expensive in the fifth, especially if the team grows.<\/p>\n
It is worth understanding what this monthly fee includes: how many simultaneous calls per user, whether national calls are included, if call recording is a paid option, and how many devices each person can use at the same time.<\/p>\n
Our Portability<\/strong> usually has a one-off cost, and the minimum contract period<\/strong> that is where the differences between seemingly identical proposals are hidden.<\/p>\n
How to compare proposals<\/h2>\n
The only honest comparison is made to three to five years<\/strong>, and with an explicit growth scenario.<\/p>\n
Build a table with the cumulative cost of each option at the end of one, three and five years, assuming the number of people you expect to have at each point in time. The result is usually counter-intuitive: solutions that seemed expensive win, and vice versa.<\/p>\n
Two typical scenarios. A stable company, with a single office and twenty to forty workstations, tends to pay less with an on-premises solution from the fourth year onwards. A growing company, with multiple locations or many mobile workers, rarely regrets the cloud \u2014 the extra cost buys flexibility that the other option cannot provide.<\/p>\n
Hidden costs<\/h2>\n
Too many simultaneous channels.<\/strong> It is the most common waste in business telephony. Outside of intensive customer service contexts, a team of thirty people rarely needs more than eight to ten channels. Sizing by the number of employees instead of actual traffic costs money every month.<\/p>\n
The network that is going to have to improve.<\/strong> If the internal network does not prioritise voice, the new switchboard will give poor calls \u2014 and the true bill includes intervention in the structured network<\/a>.<\/p>\n
Back-up power supply.<\/strong> IP phones rely on switches, which rely on power. Without a UPS, a power failure leaves the business unreachable.<\/p>\n
Traffic fraud.<\/strong> A poorly protected PBX exposed to the internet can generate thousands of euros in international calls in a single weekend. It is a cost that does not appear in any proposal and can be avoided with adequate security configuration<\/a> and destination limits.<\/p>\n
- The terminals<\/strong> \u2014 IP phones, headsets, computer and mobile applications.<\/li>\n
- What factors weigh into a custom solution<\/a><\/li>\n
- Year-on-year comparisons are always misleading.<\/strong> On-premise and cloud solutions typically cross paths between the third and fifth year.<\/li>\n