Opening a Second Office in Raffles Place: A Singapore Consultancy's Phone System Decision

In short: A Singapore consultancy opening a second office collected four cloud PBX quotes and found them clustered within a few dollars per user. Price was never the deciding question. The one none of the quotes answered was whether two addresses would behave like one phone system — shared numbering, calls that transfer between sites, and Teams that actually rings.

The lease was signed before the phone system was discussed. That is the usual order, and it is rarely a problem until someone works out that the new office needs to answer calls on day one, and that the phone line is the longest lead item on a fit-out that everybody assumed was mostly furniture.

The company in this scenario is a Singapore management and engineering consultancy — around fifty-five staff, a main office south of the city, and a newly signed satellite floor in Raffles Place chosen because most of their clients are within a ten-minute walk of it. This is a composite, built from patterns that recur across Singapore professional-services firms of this size, not a named client. The operations director did what any sensible operations director does: she asked four providers for a quote.

The quotes came back within a few dollars of each other per user per month. And that, rather than any of the numbers, is where the interesting part of the story starts.

Why Singapore firms treat a phone system as a commodity purchase

Singapore is a genuinely priced, genuinely competitive market for cloud PBX, and that is unusual. In many Asian markets, a business phone system is quoted only after a scoping call. In Singapore, entry-level hosted PBX pricing is published: local providers advertise starting points in the region of S$30 to S$50 per user per month, seat bundles priced by the handful, and managed offerings that start higher for firms wanting more hand-holding. Rates move, so any specific figure should be checked at the time of buying — but the point stands that a buyer here can compare published numbers without picking up the phone.

Published pricing is good for buyers. It also produces a specific failure mode: when four quotes are visible and close together, the decision quietly reframes itself as "which of these is cheapest," because that is the only variable the comparison exposes.

For a single-office company, that reframing is mostly harmless. Any competent hosted PBX will handle one office's extensions, an auto attendant, and a main line. The differences between providers are real but small, and the cheapest option is often a perfectly reasonable answer.

The moment a second address appears, the comparison stops working — because the thing that now matters most is not priced on the quote sheet.

The question none of the quotes answered

What the operations director actually needed to know was whether the two offices would behave like one company or two.

That sounds abstract. In practice it is four very concrete requirements, and none of them appeared as a line item on any of the four quotes.

One numbering plan across both addresses. Can a colleague in Raffles Place be reached by dialling a three- or four-digit extension from the main office, and vice versa? Or does calling the other office mean dialling a full external number, like calling a different company?

Calls that transfer between sites. When a client calls the main line and needs the partner who is at the Raffles Place office today, can the call be transferred — warm, with a word first — or does the receptionist have to take a message and send an email?

One identity to the outside world. Does the firm present one main number and one auto attendant regardless of which office picks up? Or do clients start learning two numbers, and eventually start calling the one where they know someone, which defeats the point of having a main line?

Softphones for people who are in neither office. A consultancy of this kind has people at client sites most of the week. If the phone system's answer to that is "they can use their mobiles," then the phone system is not really the phone system, and the extension list is a fiction.

None of these are exotic requirements. They are what "one company across two addresses" means, mechanically. But they are architectural questions, and the quote request had not asked about architecture — it had asked for a per-user price, so that is what came back.

What goes wrong when this is discovered late

The problems are ordinary and predictable, which is what makes them worth naming.

Two systems that happen to be from the same vendor

The cheapest path for a provider quoting a second office is to stand up a second instance. Same platform, same invoice, entirely separate numbering. It is not misleading — it answers exactly what was asked — but the result is two phone systems that share a logo and nothing else. Extensions do not overlap. Transfers become external calls. The two offices are now two companies as far as the phone is concerned.

A receptionist who becomes the integration layer

When systems do not talk, a person compensates. Calls arrive at the main line for people who are not there, and someone spends part of every day taking messages, sending emails, and calling colleagues on their mobiles to say a client is holding. This cost is real and never appears in any comparison, because it is paid in someone's attention rather than in a monthly invoice.

The Microsoft Teams requirement that arrives afterwards

Most professional-services firms in Singapore run on Microsoft 365, and their internal communication has already moved into Teams. Nobody puts "must work with Teams" in the original quote request, because Teams feels like a separate thing from the phone system. Then the second office opens, half the firm is at client sites, and the obvious question appears: why are we switching applications to make a call to a colleague whose presence indicator is already on screen?

Integrating a PBX into Teams is well-trodden — it is done with Microsoft-certified Session Border Controllers, so calls to extensions, landlines and mobiles can be made directly from Teams — but it is much easier to plan for at the outset than to retrofit onto a system chosen for its per-seat price.

Nobody asked what happens when one office loses internet

A cloud PBX depends on connectivity, and a second office is a second thing that can go down. What happens to calls to the Raffles Place extensions when that floor's circuit fails? Do they fail over to the main office, to mobiles, to voicemail — or do they simply not connect? This is a five-minute conversation during design and a very bad afternoon if it is never had.

The way to actually make this decision

For a firm opening a second location, the buying question is not "cheapest per-user rate." It is "does this behave like one phone system across two addresses, and who is responsible when it does not."

Three things are worth more than the price difference between the quotes.

Architecture, decided before pricing. One tenancy, one numbering plan, one auto attendant, extensions that reach across both offices, and routing rules that reflect how the firm actually works — main line to the front desk during office hours, overflow to a queue, out-of-hours to voicemail-to-email so nothing sits unheard until Monday. This is a design conversation, and it is short. Skipping it is what produces two systems.

An uptime commitment you can point at. Brocent's managed cloud PBX is hosted on Alicloud, AWS and Azure with a 99.99% uptime commitment. That number matters less as a marketing figure than as a statement that somebody has taken responsibility for availability — which is precisely what nobody owns when a firm assembles its own arrangement from the cheapest parts.

Teams integration treated as a requirement, not an upgrade. If the firm already lives in Teams, Microsoft Teams integration via certified SBCs should be in the design from the start. It is the difference between a phone system people use and a phone system people route around.

There is a fourth consideration that only applies to some firms, but applies strongly when it does. Singapore consultancies of this size frequently open in the region next — Hong Kong, Kuala Lumpur, or a China representative office. A cloud PBX that can add local numbers in Hong Kong, mainland China and Japan alongside the Singapore ones, on the same platform, means the third office is a configuration change rather than a repeat of this entire exercise. If regional expansion is plausible within a few years, it is worth weighting now, when it costs nothing to plan for.

What it costs, honestly

Per-extension economics are not where the money is, and it is worth saying so plainly.

Indicative market-reference rates for the components are in the range of US$10–20 per extension per month for a basic extension, US$15–35 for a queue extension with call-queue and IVR features, and US$15–30 per channel per month for SIP trunking, with the hosted PBX platform fee quoted per deployment. These are reference figures ahead of a full pricing update, and final numbers are confirmed during scoping — but the shape is the useful part. Across fifty-five users, the spread between the cheapest and most expensive quote in a Singapore comparison is typically a few hundred dollars a month.

Set against that: one receptionist spending an hour a day bridging two disconnected systems, one partner missing a client call because it could not be transferred, or one office's phones being unreachable for an afternoon. The rate difference is not the variable worth optimising.

The more useful framing is where the phone system sits. For a firm this size, telephony is one service among several that all need to work on day one at the new office — network, wireless, endpoints, security, and the phones. Buying them from five vendors on five contracts, each optimised individually, is how a fit-out ends up with five different people to call when something does not work on move-in morning.

The part of the timeline that surprises people

The design conversation is short. What follows it is not, and this is the most common way a second-office phone system ends up being the thing that is still not working on move-in morning.

Number porting sits on the critical path. If the firm wants to keep its existing main number — and it almost always should, because clients, letterheads, email signatures and a decade of directory listings all point at it — that number has to be ported from the incumbent carrier. Porting is a carrier-to-carrier process with its own paperwork, its own validation steps and its own queue. It is not something a provider can compress by working harder, and it is the single item most likely to be started three weeks later than it should have been. The practical rule is to confirm porting timelines with the current carrier at the same time the lease is signed, not when the fit-out is being scheduled.

Connectivity has to exist before the phones can be tested. A cloud PBX needs the new office's circuit to be live. Business fibre provisioning in a Singapore commercial building depends on what is already in the riser, whether the building has an existing arrangement with the carrier, and how quickly building management grants access for any in-premises work. In a well-served CBD tower this can be quick. It is not reliably quick, and "we assumed the internet would just be there" is a recognisable failure pattern.

The cabling and the desks are the same decision. Where the handsets, access points and floor ports go is decided by the seating plan, and the seating plan is usually still moving while the electrical and data cabling is being installed. Firms that resolve this late end up either running visible cable after fit-out or discovering that the meeting room nobody planned to put a phone in is the one clients get put in.

Someone has to make test calls before anyone moves in. Not just "does the extension ring" but the routing that was designed: a call to the main number during office hours, the same call after hours, a transfer from one office to the other, a softphone call from a mobile network, and voicemail arriving as an email. Half an hour of deliberate testing in an empty office is worth more than any amount of testing done while fifty-five people are trying to work.

None of this is exotic project management. It is, however, the difference between a phone system that is quietly ready on day one and one that becomes the visible symbol of a difficult move. The reason it is worth naming here is that it is entirely invisible on a per-user price comparison — the quotes that were clustered a few dollars apart said nothing at all about who would own this sequence, or when it would start.

Three ways to phone a second office

Comparing local providers on price alone

• What it looks like: four quotes, similar per-user rates, decision made on the lowest number.
• Cost: the lowest available, on paper.
• What it gets you: working phones in both offices.
• Where it fails: the comparison never exposed the cross-office question, so whether the two sites share numbering, transfer calls, or present one identity is decided by whatever the provider happened to build — not by the firm.

A separate PBX for each office

• What it looks like: the new office gets its own system, sometimes from a different provider, often because it was procured by whoever was managing the fit-out.
• Cost: individually cheap; two invoices.
• What it gets you: two functioning offices.
• Where it fails: internal calls become external calls, transfers break, two main numbers circulate, and every future change has to be made twice. Consolidating later means renumbering, which is disruptive precisely because clients have learned the numbers by then.

One hosted cloud PBX across both offices

• What it looks like: a single platform with one numbering plan, one auto attendant, cross-site routing, softphones for people at client sites, and Teams integration where the firm already works.
• Cost: a platform fee plus per-extension pricing; not the cheapest line on a spreadsheet.
• What it gets you: two addresses that behave like one company, one party accountable for uptime, and a third office that is a configuration change rather than a project.
• Where it costs more: it requires a design conversation before purchase, and it is a managed arrangement rather than a self-service subscription.

Frequently asked questions

How does this compare to local Singapore providers on price?

Entry-level hosted PBX in Singapore is genuinely competitive, with published starting rates commonly in the S$30–50 per user per month range and seat bundles priced lower per head. Brocent's indicative component rates sit in a similar band — around US$10–20 per extension per month for a basic extension, with the platform fee quoted per deployment. If the requirement is one office and a main line, price is a reasonable way to choose. If it is two offices that need to behave as one, the comparison is measuring the wrong thing.

Can two offices share one numbering plan?

Yes, and this is the central reason to run both sites on one platform rather than two. Staff dial short internal extensions regardless of which office they or the person they are calling is sitting in, transfers work across sites, and the firm presents a single main number and auto attendant to clients.

Does it integrate with Microsoft Teams?

Yes. Integration uses Microsoft-certified Session Border Controllers, which allow calls to extensions, landlines and mobile numbers to be placed directly from Teams. For a firm already running Microsoft 365 with staff at client sites, this is usually the difference between a phone system that gets used and one that gets bypassed.

What happens if the internet drops at one office?

That is a design question to settle before deployment rather than during an outage. Typical answers include failing extensions over to the other office, to the UC softphone on mobile data, or to voicemail-to-email so nothing is silently lost. The important part is that the behaviour is chosen deliberately and documented, not discovered.

Do staff need physical desk phones?

Not necessarily. The UC softphone gives a consistent corporate call experience on Android, iPhone and desktop, which for a consultancy with people at client sites most of the week is often more useful than a handset on an empty desk. Many firms end up with a small number of desk phones in shared and reception areas and softphones everywhere else.

How long does setup take for a new office?

The technical configuration is not usually the constraint. Number porting, circuit provisioning and the fit-out schedule are. This is why the phone system is worth raising when the lease is signed rather than when the furniture arrives — the design conversation takes an hour, and the lead times behind it do not.

Can we keep our existing main number?

Usually yes, via number porting, and keeping it is normally the right call — a main number that clients already have is an asset. Porting timelines depend on the current carrier and should be confirmed early in the project, because they sit on the critical path in a way that most fit-out schedules do not anticipate.

Where this ends up

The four quotes were not wrong. They answered the question they were asked, which was what a seat costs per month. The question that mattered — whether two addresses would work like one company — was not on the form, so it did not appear in the answers.

That is the general shape of this decision. A phone system for one office is a commodity. A phone system for two offices is an architecture question, and architecture questions are decided before pricing or they are decided by default.

It is also why telephony belongs inside a broader arrangement rather than beside one. On the day the Raffles Place office opens, the firm needs the network working, the wireless covering the floor, the laptops enrolled, the security controls applied, and the phones ringing — and it needs one number to call when one of those five is not true. Brocent's managed IT support covers that ongoing scope as a per-user monthly plan, with cloud PBX designed and run as part of it rather than procured separately and hoped to fit. The pricing page shows Singapore plan rates per user per month, so the whole picture can be costed before anything is committed.

If a second office is on the horizon, the useful hour to spend is the one that happens before the quotes are requested — deciding what "one company across two addresses" needs to mean, and then asking providers to price that. Talk to us if it would help to have that conversation with someone who has designed the arrangement before.