What is low latency? Definition, FAQs, and 10 tips to minimize latency
What is low latency?
In networking, low latency is the ability of a computing system to rapidly transmit data packets over a network from source to destination with minimal delay (latency). When latency is low, data is transmitted fast enough to deliver real-time, responsive, and interactive digital experiences. Zero latency is optimal but isn't possible in real-world apps and networks.
Low latency is paramount in environments like real-time trading or in-app communication, where real-time processing and instant feedback are critical to the user experience.
How to measure latency
Latency can be measured in seconds, milliseconds (ms), and even microseconds. The bar for low latency varies by app and industry, and is set to meet user expectations of a fast, seamless, premium experience on a certain platform.
For example, 53% of mobile users will leave a webpage if it takes more than 3 seconds to load. Web browsing may be able to tolerate seconds of latency, but for real-time communications, latencies higher than 100 milliseconds can cause delays that derail the user experience.
Defining low latency for various applications: How low is low?
Here’s how low latency is defined for a variety of common applications.
In-app notifications: Less than 1 second. Apps rely on notifications to relay updates to users in near real-time. You can build in-app notifications with low latency using Sendbird.
Streaming video and audio: Low latency of 1-5 seconds maintains real-time engagement between sender and receiver. High latency results in delivery delays and laggy viewing.
Real-time communication: A low latency of 150 ms or less is needed for live calling or chat to facilitate smooth, natural conversations in various industries, from digital health to fintech and beyond. High latency leads to crosstalk.
Online gaming: A low latency of 100 milliseconds or less is essential for maintaining competitive integrity and responsiveness in online gaming and esports. Players rely on real-time feedback for split-second decisions, and high latency can degrade gameplay.
Web browsing: Latency under 100 ms ensures prompt page loads and a responsive UX while surfing the web. High latency delays input responses from users, slows rendering, and disrupts navigation.
Augmented reality (AR) and virtual reality (VR): 20 milliseconds or less. AR/VR experiences require near-zero latency to feel immersive and real. High latency leads to a discrepancy between the user’s movement and visual feedback, causing discomfort and disorientation.
Fintech use cases: 1-10 microseconds. Low latency is essential in high-frequency financial trading and financial services apps, where market prices change in milliseconds, and it’s a competitive advantage to trade faster. High latency can result in missed opportunities.
In general, the more time-sensitive and interactive your use case, the lower the latency will need to be.
What is ultra low latency?
Ultra-low latency refers to exceptionally minimal delays, significantly lower than what is typically categorized as "low latency," in the transmission, processing, and reception of data in networking, computing, and telecommunications contexts. This term implies a focus on achieving the fastest possible response times, often measured in microseconds (µs) or even nanoseconds (ns), depending on the application's requirements. It’s vital to industries that rely on real-time data exchange, such as online gaming, high-frequency financial trading, and real-time communications.
Jitter vs. high latency
Jitter refers to variations in latency over time. Fluctuations in latency (brought on by changes in network traffic or routing times) can create an inconsistent user experience, especially for real-time communications. This unpredictability can be more disruptive to the user experience than consistently high latency. Ideally, both jitter and latency (or ping) would be low. Tracking these metrics allows you to gauge how quickly and consistently data packets are transmitted from one point to another.
What is latency in networking? The relationship between latency, bandwidth, and throughput
In simple terms, latency in networking is the delay in data transmission over a network. Propagation delay in networking refers to the time it takes for a signal to travel from the sender to the receiver over a network. This delay is a fundamental aspect of network performance. It is influenced by the physical properties of the transmission medium (such as copper wire, fiber optic cable, or air in the case of wireless communications) and the distance between the communicating devices. Propagation delay is measured in units of time, typically milliseconds (ms) or microseconds (µs), and is a critical factor in a low latency network.
What is bandwidth in networking?
The maximum amount of data that can be transmitted over a network connection in a given period, also known as the maximum capacity, is known as bandwidth in networking. By allowing more data to travel across the network, higher bandwidth leads to lower latency in networking.
What is throughput in networking?
The amount of data that is transmitted over a network in a given time period is called throughput in networking. While bandwidth is the upper limit for data transfers, throughput is the actual data transfer rate experienced by users, reflecting how much data can effectively be sent.
What is the relationship between bandwidth and throughput in networking?
Increasing network bandwidth and network throughput can lower latency, enabling more data to move across a network. That said, ample bandwidth and throughput don’t guarantee low latency, as data travel time also depends on factors like network congestion, processing delays, and more aspects we’ll cover in the next section.
What causes high latency in networking?
High latency in networking can be attributed to several factors, each affecting the time it takes for data to travel from its source to its destination. Understanding these factors is crucial for diagnosing and mitigating latency issues while building a low latency network.
It starts with an app’s request to a server. As the requested data beams across the network toward its destination, it encounters a variety of network components — switches, routers, protocol changes, translators — spread over miles of network cable, fiber, and connections.
Millisecond delays are introduced at each juncture: processing delays, transmission delays, queueing delays, propagation delays, as well as router and switch processing times. These delays compound as data travels between network connection points. Moreover, each hop in the journey and device between the data source and its destination adds to the overall network latency.
To lower latency, you’ll start by identifying and addressing the specific technical factors behind high latency in your specific app use case. These are the 10 most common causes of high latency in in-app communications:
Distance in a low latency network
The fundamental cause of latency is physical distance. The greater the distance, the longer it takes for data to travel from the source to the destination. This is known as propagation delay, as mentioned above.
According to a recent survey in Forbes, reducing latency was the second most-cited reason (44%) for surveyed businesses to add new locations, or to use content delivery networks (CDNs). What is a CDN? CDNs are a network of geographically distributed edge servers located at strategic points around the world, reducing the time it takes for data to reach global user bases.
Network congestion
If a network is carrying too much traffic, multiple data packets must compete for limited transmission resources, which causes latency. This typically happens when too many clients join a network without the bandwidth to support them all, causing delays or even data loss.
Transmission medium
Different transmission mediums — fiber optic cables, copper wires, wireless connections — all propagate signals at different speeds. Fiber optic cables transmit data faster than copper wires, for example. Each discrepancy adds to latency, which compounds over connection points.
Hardware quality
The network devices that data passes through — routers, switches, hubs — can contribute to high latency. Each device needs time to process, inspect, and forward data packets.
Upgrading to high-performance hardware, networking equipment, and low latency data storage solutions can help to reduce latency. Greater processing power allows for faster packet processing, routing, and forwarding.
Routing
What is routing? Routing in networking is the process by which data packets are forwarded from one network segment to another, typically moving through networks from the source to the destination. This process involves determining the optimal path for data to travel across multiple networks, which is crucial for efficient data transmission. Routers, the devices responsible for routing, use routing tables, algorithms, and protocols to decide the best route for the packet to take to reach its endpoint.
The process of routing data packets from source to destination involves making decisions about the most efficient path through the network. Inefficient routing decisions or changes in network routes due to link failures or congestion can add latency.
Communication protocol overhead
Each layer of the network protocol stack adds overhead to a data transmission. For example, encapsulating data packets in headers for routing and error-checking purposes adds bytes of data to be transmitted, causing latency.
Data packet inspecting and filtering in network security
While essential to network security, packet inspection and filtering can increase network latency. Each inspection requires the device to analyze the packet headers and payload, increasing the processing overhead and time per packet.
Network configuration
Your network settings contain several factors that can reduce the efficiency and speed of data transmission. Factors to pay attention to include: suboptimal routing paths, bandwidth allocation, and inadequate buffer sizes.
Server performance
Server performance directly impacts the time it takes to process incoming server requests. A server under a heavy load or with limited capabilities accrues processing delays and latency, especially in apps that require complex computation or data processing.
How to lower latency: 10 tips to minimize network latency
To achieve a low latency network, developers can optimize for the efficiency, capacity, and responsiveness of apps and network systems, specifically how they perform actions and handle data. For example, optimized code and efficient processors can handle inputs and outputs quicker, reducing time delays for users. Here are 10 tips on how to lower latency in your next app or use case.
1. Optimize network requests
You can minimize the number of network requests, and therefore the number of round-trips required to complete a task, by bundling them when possible. Fewer requests lead to a lower overall round-trip time (RTT), and, therefore, a low latency network.
2. Minimize overhead
API network requests incur overhead due to data encapsulation, transmission, and processing. You can optimize requests to send only essential data, minimizing unnecessary headers or metadata to reduce data transmission load and thereby reduce latency.
3. Prioritize critical data loading
Load critical resources first — essential scripts, messaging data, user information — to accelerate the initial rendering of a messaging notifications interface (or notification center) and deliver a responsive, low-latency experience.
4. Implement data caching
You can accelerate response times to common user actions by locally caching data frequently accessed on devices. This reduces the need to fetch the same data repeatedly from the server, helping reduce latency.
5. Compress and optimize media
Optimizing images and media before loading them into the app notifications interface reduces the data transmitted over the network to lower latency. Lazy loading for media files also helps to prioritize visible content loading and improve responsiveness for in-app chat experiences.
6. Invest in connection resilience
Without a backup plan, network interruptions can slow or halt user experiences. Implementing offline modes for networks or background sync allows you to send and receive messages seamlessly, even in challenging network conditions.
7. Simplify your UI
Reducing the complexity of your user interface (UI) cuts down the rendering time for chat view and enhances your app’s responsiveness. For example: using pagination or limiting the number of visible notifications in the UI can help lower latency.
8. Use WebSockets for real-time communication
Efficient communication protocols and transport mechanisms, like WebSockets or HTTP/2, reduce latency. WebSockets maintain a persistent connection for real-time communication and allow instant message delivery without constant polling, which increases network latency.
9. Minimize API call time
You can optimize server response times by minimizing processing overhead for API calls. Using efficient algorithms and data structures to handle tasks helps to reduce the round-trip time for API calls, and in turn, reduce latency.
10. Monitor and analyze performance to minimize latency
Identify and address latency bottlenecks in your app with performance monitoring tools and debuggers. Regularly monitor and analyze metrics for API calls, data load times, and user interactions to pinpoint areas of improvement.
Paying attention to these techniques will allow you to ensure a fast, seamless, and responsive app experience.
5 low latency use cases: How low latency is the lynchpin of digital UX
For businesses that compete based on digital user experience, lowering latency can be the difference between failure and success. In ecommerce, for example, even a one-second delay in page load time can result in a 7% decrease in conversion rate.
Here are 5 important business use cases that rely on low latency:
Online marketplaces: The performance of online marketplaces hinges on low latency networks. Top shopping apps rely on low latency to enable interactive interfaces, smooth navigation, and quick page loads that make browsing products fast and easy.
Online gaming: Low latency is essential to responsive, fair, and competitive gameplay. Any lag or discrepancies between play actions and game response can significantly degrade the online gaming experience.
Telehealth: Low latency allows health providers to offer real-time communications, remote consultations, and timely interventions. For example, telestroke consultations require neurologists to evaluate patients and recommend a prompt intervention remotely. When network latency is high, visual cues and patient responses can get lost.
Live streaming: Low latency ensures a smooth, engaging, interactive streaming experience for both viewers and creators. Users can interact without noticeable and distracting delays, fostering a sense of community that drives participation.
Financial services: In financial markets, milliseconds can make a difference in trade execution and profitability. Low network latency ensures that financial insights and transactions are sent and executed in real time so users can capitalize on fleeting opportunities.
In industries where even milliseconds can count, maintaining the lowest network latency is imperative.
An important application of low latency: In-app notifications
In-app notifications are personalized, relevant, and engaging messages, supported by rich media, sent within a company’s app. They are frequently used to communicate critical information, offer updates, or engage users with information relevant to their current app usage.
Low latency is crucial for in-app notifications because it ensures that messages are delivered promptly, thereby enhancing the user experience and engagement. For example, consider a live sports scoring app that sends in-app notifications for game updates, scores, or important events like goals or penalties. Low latency is critical in this context because users expect to receive these updates in real time. If there's a significant delay in receiving the notification, the information may no longer be relevant or engaging, as users could already have learned about the event from another source or from the live broadcast itself. This diminishes the app's value and can lead to user dissatisfaction, reducing engagement and potentially decreasing app usage or retention.
More importantly, for time-sensitive or critical alerts concerning public health and safety, low latency is not just important for user satisfaction but can also have significant real-world implications. For example, consider a tsunami alert. For residents in coastal areas, receiving evacuation alerts with the lowest possible network latency can mean the difference between life and death. An alert system with low latency ensures that the warning reaches the population as quickly as possible, giving them the maximum amount of time to evacuate to higher ground or seek shelter. Delayed notifications, on the other hand, can significantly reduce the time available for evacuation, increasing the risk of casualties.
Here are the main reasons why low latency is critical for in-app notifications:
Real-time communication: In-app notifications provide users with breaking information, updates, news, or alerts. Low latency ensures notifications reach users promptly, allowing them to stay informed and engage with time-sensitive content.
Increased user engagement: Low latency contributes to a more responsive, interactive experience with in-app notifications, increasing the chances that users will engage with notifications because they arrive at just the right moment.
Effective collaboration: Platforms that enable collaboration, such as Slack, rely on timely notifications to deliver updates, messages, and alerts to users, facilitating smooth and efficient communication.
Enhanced user satisfaction: Low latency notifications are essential to a quality user experience, ensuring that the flow of communications feels seamless and timely within a responsive UI.
Sendbird offers a robust in-app notifications solution built with latency-reducing features and more, to help you build latency-free app communications without the extra effort.
Offline messaging: When low latency doesn’t matter
A survey of American adults who live in rural areas finds that 1 in 5 say that access to reliable high-speed internet is a problem. Meanwhile, internet and communications remain unreliable or unavailable in areas of Africa, Asia, and Oceania.
If you're trying to reach users who lack stable internet, lowering network latency won’t be much help. Implementing modern messaging features, however, gives users new ways to connect when coverage is lacking.
Offline messaging: This feature allows users to access existing messages and compose new messages without internet connectivity. Once users regain connection, all unsent messages are delivered automatically. Offline messaging allows users to communicate asynchronously when out of coverage, which is handy in many real-world scenarios.
Local caching: This feature lets your app store frequently accessed data locally, allowing users to access messaging and chat history while offline. This lets users access core functionality and chat history, possibly even with AI chatbots and customer support, while offline.
Ensure users stay engaged with low latency in in-app communication
Is low latency good? Yes, and it’s becoming more important every day! As time spent online increases, low latency applications will become even more vital, allowing businesses to engage and retain users with premium digital experiences.
Sendbird makes it easy for developers and businesses to build low-latency in-app communications - especially in-app notifications. Our proven APIs for in-app notifications, video and voice calling, and chat all come with built-in features that reduce network latency to ensure smooth experiences without delays.
You can start a free trial of Sendbird today, or contact us to learn more!