The Short Answer
Acceptable latency for VoIP is generally below 150 milliseconds (ms) one way. If a test reports round-trip time, commonly called RTT, the comparable target is below 300 ms round trip.
For the most natural business conversations, lower is better. A one-way result below 100 ms—or an RTT result below 200 ms—provides a stronger margin for consistently responsive calls.
Once one-way latency reaches 150 ms or higher, pauses become more noticeable. At 300 ms one way, a conversation may feel unnatural, with people unintentionally interrupting or talking over one another.
Use the Tier 1 Telecom Internet Quality Test to check your connection’s latency, jitter, loaded performance and estimated VoIP capacity.
Key Takeaways
- Below 100 ms one way is excellent for VoIP.
- Between 100 and 149 ms one way is generally good.
- Between 150 and 299 ms one way can cause noticeable conversational delay.
- A result of 300 ms or higher one way is poor for normal business calling.
- Many internet tests report round-trip latency rather than one-way latency.
- Fast download speed does not guarantee low latency.
- A connection that performs well while idle may develop serious latency when busy.
- Physical distance, internet routing, congestion, Wi-Fi and network equipment can all affect latency.
The Common Misconception: Latency and Internet Speed Are the Same Thing
Internet speed and latency measure different characteristics of a connection.
Speed measures how much data a connection can transfer over time. Latency measures how long it takes data to travel between two points.
A business can have a 500 Mbps or gigabit internet plan and still experience delayed phone conversations. The connection may have plenty of capacity, but voice packets may take too long to reach their destination.
Think of bandwidth as the number of lanes on a highway. Latency is the time it takes a vehicle to reach the other end. Adding more lanes allows more vehicles to travel simultaneously, but it does not necessarily shorten the distance or improve the route.
That is why upgrading to a faster internet package does not automatically correct every VoIP latency problem.
What Does Latency Mean?
Latency is the delay between sending information and receiving it at another point on the network. It is measured in milliseconds.
During a VoIP call, your voice is converted into small digital packets. Those packets travel from your phone or application through the local network, across the internet, through the VoIP platform and eventually to the other person.
Every part of that path introduces some delay.
A small amount of latency is unavoidable. Problems begin when the combined delay becomes large enough to interrupt the natural rhythm of a conversation.
High latency may cause:
- A noticeable pause before the other person responds
- Two people repeatedly talking at the same time
- Callers accidentally interrupting one another
- Conversations that feel awkward or unnatural
- Delayed keypad tones or menu selections
- Slow response from an auto attendant
- Echo becoming more noticeable
- Calls that resemble a satellite or two-way radio conversation
Latency primarily creates delay. Choppy or robotic audio is more commonly associated with jitter, congestion or other network instability, although these problems often occur together.
One-Way Latency Versus Round-Trip Time
Before interpreting a latency result, determine what the test is measuring.
One-way latency
One-way latency measures how long data takes to travel from the sending endpoint to the receiving endpoint.
Cisco and the International Telecommunication Union use one-way, mouth-to-ear delay when discussing VoIP design. Cisco’s VoIP quality guidance recommends keeping one-way delay below 150 ms for high-quality real-time voice.
Round-trip time
Round-trip time measures how long data takes to travel to a destination and return.
Many browser-based internet tests report RTT because it can be measured from one device without placing monitoring equipment at both ends of the connection. Microsoft, for example, identifies latency below 300 ms RTT as a requirement for optimal real-time voice performance in its Microsoft Teams network guidance.
RTT is often approximately twice the one-way latency, but the relationship is not always exact. Upload and download traffic may follow different routes or encounter different network conditions.
A 200 ms RTT result does not guarantee that each direction takes exactly 100 ms. It does, however, provide a practical indication of the connection’s responsiveness.
What Is a Good Latency Result for VoIP?
The following ranges provide a practical way to interpret VoIP latency:
| One-way latency | Approximate RTT guideline | General rating | Likely VoIP experience |
|---|---|---|---|
| Below 100 ms | Below 200 ms | Natural and responsive business conversations | |
| 100–149 ms | 200–299 ms | Generally suitable for reliable business calling | |
| 150–299 ms | 300–599 ms | Delay may become noticeable during conversation | |
| 300 ms or higher | 600 ms or higher | Conversation may feel awkward and interruptions become likely |
The RTT column is an interpretation guide, not a mathematical guarantee. Network paths are not always symmetrical, and actual mouth-to-ear delay includes processing that an ordinary internet test may not measure.
The objective is not simply to remain below the maximum acceptable number. A consistently low result provides more room for temporary congestion, routing changes and other normal fluctuations.
Why Does High Latency Affect Conversations?
People naturally leave only short pauses between speaking and responding.
When latency is low, the conversation feels immediate. When latency becomes high, the listener hears each statement later than it was spoken.
The original speaker may assume the other person has not responded and begin talking again. At nearly the same moment, the delayed response arrives. Both people then stop, apologize and try again.
This repeated talking over one another is one of the clearest signs of excessive latency.
High latency does not always make the audio itself sound distorted. A call can sound perfectly clear while still being frustratingly delayed.
What Is Latency Under Load?
An idle latency result measures responsiveness when the internet connection is relatively quiet. Loaded latency measures what happens while the connection is also handling substantial upload or download activity.
This distinction matters in a business environment.
A connection might report 20 ms of latency early in the morning but rise sharply when employees begin:
- Uploading large files
- Synchronizing cloud storage
- Running online backups
- Joining video meetings
- Streaming training content
- Accessing cloud-based applications
- Sending traffic from security cameras
- Downloading operating-system updates
If latency rises dramatically during these activities, calls may perform well when the office is quiet and poorly during the busiest parts of the day.
On the Tier 1 Internet Quality Test, an increase above 20 ms under load produces an advisory. An increase from 21 to 40 ms is treated as moderate, while an increase above 40 ms is more significant and may affect calls. An increase above 80 ms receives no loaded-latency points and calls for corrective action.
A good connection should provide both low idle latency and reasonably stable latency while the network is busy.
What Causes High VoIP Latency?
Physical distance
Data cannot travel instantly. Longer distances normally create more delay.
A business in Florida connecting to a nearby regional service location will generally experience less latency than one sending the same traffic across the country or overseas.
Distance is also one reason satellite connections traditionally experience higher latency. The data must travel a much longer physical path before reaching the terrestrial network.
Inefficient internet routing
The geographically closest destination is not always reached through the shortest network route.
Internet providers exchange traffic through specific networks and facilities. An inefficient or congested route can send packets through additional locations before they reach the VoIP platform.
Network congestion
When too much traffic competes for the same connection, packets wait in queues before being transmitted.
This can increase both latency and jitter. The problem may only appear during busy periods, which is why testing under load is important.
Bufferbloat
Routers and internet equipment use buffers to temporarily hold traffic. Excessively large or poorly managed buffers can allow long queues to develop instead of controlling congestion efficiently.
The connection may still report strong download and upload speeds while latency increases dramatically whenever it becomes busy.
Wi-Fi interference
Weak wireless signals, interference and competition between devices can introduce additional delay.
Wi-Fi must coordinate access to shared radio channels. Devices may need to wait or retransmit data when signals collide or arrive incorrectly.
A wired VoIP phone will generally have a more predictable connection than a wireless device operating with weak or congested Wi-Fi.
VPNs and security inspection
A VPN may send traffic through a remote gateway before allowing it to continue to its destination. Firewalls and security services can also inspect or process network traffic.
These functions are often necessary, but inefficient configurations or overloaded equipment can introduce avoidable delay.
Aging or overloaded equipment
Routers, firewalls, switches and wireless access points must process every packet that passes through them.
Equipment that lacks sufficient capacity may perform normally during light use but introduce delay as traffic increases.
Problems within the internet provider’s network
High latency may originate beyond the business’s local network.
Regional congestion, routing problems, damaged infrastructure or overloaded provider equipment can all affect responsiveness. Testing from multiple devices and comparing results at different times can help determine whether the problem is local or provider-related.
How Should You Test Latency for VoIP?
Run the Tier 1 Telecom Internet Quality Test from the same connection your business phones will use.
For a more representative assessment:
- Use a wired Ethernet connection for the initial test.
- Run the test while the office is relatively quiet.
- Repeat it during a busy part of the workday.
- Compare idle performance with performance under load.
- Test from Wi-Fi if employees regularly use wireless softphones.
- Repeat the test when users report delayed calls.
- Compare latency with jitter and the connection’s estimated call capacity.
One test is only a snapshot. Internet routing, congestion and Wi-Fi conditions can change throughout the day.
The testing destination also matters. A browser test measures the path to its own test infrastructure, while an actual VoIP call travels to the phone provider’s infrastructure and then onward to the other caller.
A browser-based result is an effective screening tool, but intermittent or route-specific problems may require monitoring of actual calls and network equipment.
How Can a Business Reduce VoIP Latency?
The correct solution depends on the source of the delay. Common improvements include:
- Connecting desk phones and critical workstations through Ethernet
- Improving wireless coverage and reducing Wi-Fi interference
- Replacing overloaded routers, firewalls or access points
- Configuring Quality of Service to prioritize voice traffic
- Using traffic-management features that control excessive network queues
- Scheduling backups and large transfers outside peak calling hours
- Avoiding unnecessary VPN routing for approved voice traffic
- Selecting appropriate regional VoIP infrastructure
- Asking the internet provider to investigate routing or congestion
- Adding a secondary internet connection where phone availability is critical
A faster internet plan may help when the existing connection is regularly saturated. It will not necessarily correct excessive distance, inefficient routing, Wi-Fi interference or poorly configured equipment.
A Simple Business Example
Consider an office with a gigabit internet connection and 25 cloud phones.
When the network is quiet, the office records 25 ms RTT and calls feel immediate. During the afternoon, cloud backups and large file uploads begin. The RTT rises to 280 ms, and employees start talking over customers because responses arrive later than expected.
The office does not lack raw internet speed. Its problem is that latency rises when the connection is busy.
Appropriate traffic management, scheduling large transfers differently or correcting the router configuration may provide a better result than purchasing an even faster internet package.
Is 20 ms Latency Good for VoIP?
Yes. A result of 20 ms is excellent whether it represents one-way latency or RTT.
It leaves substantial room for the other stages involved in delivering a call and should not create a noticeable conversational delay by itself.
Is 100 ms Latency Bad for VoIP?
It depends on how the result is measured.
A result of 100 ms RTT is very good for VoIP. A result of 100 ms one way is still generally acceptable, but it consumes more of the recommended 150 ms one-way delay budget.
Always determine whether the test reports one-way latency or round-trip time before interpreting the number.
Can High Latency Cause Choppy Calls?
Latency by itself primarily causes delayed conversation rather than broken audio.
However, latency often increases alongside network congestion and jitter. In those situations, users may experience both delayed responses and distorted or interrupted audio.
If the main complaint is choppy or robotic sound, review the connection’s jitter result as well. The Tier 1 guide What Is a Good Jitter Score for VoIP? explains how inconsistent packet timing affects calls.
Will More Bandwidth Reduce Latency?
Not necessarily.
More bandwidth can help when the connection is saturated and packets are waiting behind other traffic. It may have little effect when latency is caused by physical distance, internet routing, Wi-Fi interference, VPN routing or overloaded network hardware.
This is why the complete test result matters more than download speed alone.
Does Fiber Always Have Lower Latency Than Cable?
Fiber commonly provides strong, consistent performance, but connection type is not the only factor.
Network design, routing, congestion, provider infrastructure and the destination all affect latency. A well-managed cable connection may outperform a poorly routed fiber connection to a particular service.
The most useful answer comes from testing the connection under the same conditions in which employees will use the phone system.
Test Your Connection Before Delayed Calls Affect Customers
A fast internet plan does not automatically mean that phone conversations will feel immediate.
Use the free Tier 1 Telecom Internet Quality Test to measure latency, jitter, loaded performance, upload speed and estimated VoIP capacity.
If the connection performs well while idle but develops high latency under load, Tier 1 can help determine whether the issue involves local equipment, traffic management, Wi-Fi, internet service or phone-system configuration.
For help evaluating your network or planning a cloud business phone system, contact Tier 1 Telecom or call (866) 808-4371.
Continue Learning
About Tier 1 Telecom
Tier 1 Telecom helps businesses evaluate, deploy and support cloud-based business phone systems. We focus on practical recommendations, reliable call quality and phone solutions designed around how each organization actually communicates.
