VoIP for Small Business: A Practical Checklist Before You Sign

Most small businesses sign a VoIP contract because a demo looked good on a laptop. The demo showed a phone lighting up and a menu that responded, and both things were genuine — they just did not test the parts that determine whether the system works in your building on a Tuesday afternoon. A small business does not need a procurement department. It does need a checklist, a trial, and a dozen questions asked before the signature.

Write down what actually happens on your phones

Before looking at any product, spend half an hour listing what your phone system does today. This sounds bureaucratic and it is the single most useful preparation you can do, because it converts a vague sense of "the current system is fine" into a list you can test against.

Go through each of your lines and note what happens on a call: who answers, what the caller hears, how they are transferred, whether there is a queue, what the voicemail does, whether there is an announcement or music on hold, whether anyone dials out often, and whether any line is used by equipment rather than a person.

That last category matters more than most. A fax machine, a credit card terminal, a door entry panel and a paging system all consume lines and all behave differently on a new system. Write them down separately and treat them as a distinct work stream, because they are the part most likely to break on cut-over day.

Test the trial with your own use case, not the demo script

If a supplier offers a trial, the temptation is to run through the features they showed you. That validates their setup rather than your requirement. Instead, run a test that mirrors a real working day:

1. Place three calls at the same time from different people and see what happens to each
2. Transfer a call mid-conversation, both to a person and to a queue
3. Let a call go to voicemail, then retrieve it from a different device
4. Dial out and leave a message, then check how it appears to the recipient
5. Put a call on hold, transfer it to another extension, and pick it up
6. Check what an external caller hears during each of the above

Two of those tests find most problems. The three-concurrent-calls test exposes whether your site is provisioned for the number of simultaneous conversations you actually have, which is different from the number of people you employ. The external-caller test exposes the things that are invisible from inside — a missing caller ID, a gap in the music on hold, a greeting that cuts off.

Ask about the things that only show up under load

These are the questions that separate a system that will work for you from one that will work in the vendor's demo environment. Ask them of any supplier, and get the answers in writing:

• What happens to my calls if the internet connection drops — do they go to voicemail, to a mobile, or nowhere
• Is there a local survivability option, and what does it require
• Can I keep my existing phone numbers, and who handles the porting
• What is the process and lead time for adding a new extension
• What happens to my voicemail and recordings if I move to a different system later
• Is there a charge for support, and what are the response times
• What is the contract length, and what does early termination cost

The first question is the most important and the most often answered vaguely. A business with one internet connection has a single point of failure, and the answer to "what happens if it drops" determines whether a bad afternoon is a nuisance or an outage. Some systems can fail over to a mobile number or to a local survivable branch; others simply drop the calls. Knowing which you are buying matters, and if the answer is unsatisfying, that is information worth having before go-live rather than after.

Understand the numbers before you compare them

Suppliers will talk about users, extensions and concurrent calls as if they were interchangeable. They are not, and the difference is where surprises come from.

Newrock publishes these limits separately for each model — for instance the OM20G lists maximum users, IP extensions, trunks, built-in FXS/FXO ports and concurrent calls as distinct figures. When you compare a small-business product against another, line those figures up against each other rather than comparing headline user counts, and be aware that a product with more users can have fewer concurrent calls.

For a small business the practical question is simple: does the model fit the current requirement with room for growth, and does it fit without forcing you to buy a larger licence than you need in year one?

The question that most often goes unanswered is the one about what happens when the internet connection drops. Every system has an answer, but only some suppliers will say it out loud before you sign.

Plan the numbers, because that is where the work is

For most small businesses, keeping existing numbers is the single largest source of anxiety in a voice migration, and it is also usually the most manageable. It is a carrier-side process with a lead time, which means it needs to be started early — before the hardware is even chosen — because the porting schedule, not the hardware, is what determines your cut-over date.

Ask the supplier who handles the port, what documents they need from you, and what the lead time is. Then start that clock. If there is any number you cannot port, find out now rather than on the day.

Decide what happens to the old system

Before go-live, decide what happens to the existing phone system: when it is disconnected, when the service is cancelled, and who removes it. Cable left in the wall and a service still running are the two most common leftover costs.

Also decide what happens to the numbers themselves. If the old system holds a number you want to keep in reserve, note it now — porting away releases it, and getting it back is not always straightforward.

Have a plan for the first week

Whatever the supplier promises, budget for support in the first week. Not because problems are inevitable, but because the most common early issues are the boring ones: a handset that needs a firmware update, a hunt group order that is one step off, a voicemail PIN that nobody has changed.

Write down the supplier's escalation contact, the hours they cover, and the reference numbers for your account before go-live, while everything is calm. On day three of a cut-over, having that in a document is worth an hour of your time.

Frequently asked questions

How long does a small business VoIP migration take?

It depends mostly on number porting, which is a carrier process with its own lead time, and on how much analogue equipment you have to replace. The technical cut-over itself is often shorter than people expect; the preparation is what takes the time.

Can I keep my existing business phone numbers?

Usually yes. Porting existing numbers is a standard part of a migration and most suppliers include it. Confirm the lead time early, because it gates the cut-over date, and confirm whether any number cannot be ported before you rely on it.

What if the internet goes down?

It depends on the design. Some systems fail over to a mobile number or another network, some can keep a branch working locally, and some simply drop the calls. This is worth settling in advance — it is the difference between a recoverable annoyance and a lost day of calls.

Do I need a new router for VoIP?

Not necessarily, but voice traffic is more sensitive to delay and jitter than general browsing, so the existing internet connection and router need to handle it. Ask about voice traffic handling, and about QoS if you share a link with other traffic.

Related reading on the Newrock site

• OM20G all-in-one IP-PBX
• How to Choose an IP-PBX: 8 Questions to Ask Before You Buy
• What an IP-PBX Actually Costs: Line Items People Forget
• VoIP Security: Where Attacks Actually Happen in a Business Network

Conclusion

A small business does not need a long procurement process to get VoIP right. It needs a written list of what your phones do today, a trial that runs your own use case, a dozen questions about behaviour under failure and under load, and a number-porting conversation started early. Do those things and the contract is a formality rather than a leap of faith.