Understanding Internet Peering and Why It Matters

Every time you send an email, load a web application, or join a video call, your data takes a journey across the internet. That journey’s speed, reliability, and efficiency depend on something most business owners have never heard of: internet peering. It’s the invisible plumbing that determines whether your cloud applications feel instant or sluggish, whether your VoIP calls are crystal clear or plagued by jitter, and whether your business internet connection delivers the performance your ISP promised.

For UK businesses relying increasingly on cloud-first infrastructure — Microsoft 365, Azure, AWS, Salesforce, VoIP — the quality of your ISP’s peering relationships is arguably more important than raw bandwidth. You can have a 1 Gbps leased line and still experience poor performance if your provider’s peering is substandard. This guide explains what peering is, how it works, why it matters for your business, and how to make smarter decisions about your internet connectivity.

What Is Internet Peering?

Internet peering is an arrangement between two networks to exchange traffic directly, without paying a third party to carry it between them. Think of it like two neighbouring businesses agreeing to share a private road instead of both sending their deliveries out onto the public motorway. The traffic still gets where it needs to go, but it takes a shorter, faster, more predictable route.

The internet is not a single network — it’s a vast collection of independently operated networks called Autonomous Systems (ASes). Every ISP, cloud provider, content delivery network, and large enterprise operates its own AS. When your data needs to travel from your office to a server on a different network, it must cross the boundaries between these autonomous systems. Peering is the mechanism that makes those crossings happen.

Without peering, your data would need to travel through intermediary networks (called transit providers) to reach its destination. Each additional network hop adds latency, introduces potential points of failure, and increases the chance of congestion. Peering eliminates those unnecessary detours by creating direct connections between networks that frequently exchange traffic with each other.

How Peering Works in Practice

When two networks establish a peering relationship, they agree to exchange traffic destined for each other’s customers at no cost to either party. This “settlement-free” arrangement works because both networks benefit: each saves money on transit costs and delivers better performance to its users. The physical connection is typically a fibre-optic cable linking the two networks’ routers, either at an internet exchange point or through a dedicated cross-connect in a data centre.

The exchange is governed by the Border Gateway Protocol (BGP), the routing protocol that underpins the entire internet. Through BGP, each network advertises which IP address ranges (prefixes) it can reach, and its peering partner updates its routing tables accordingly. When your ISP peers with Microsoft, for example, your ISP’s routers learn the most direct path to Microsoft’s servers — and your Office 365 traffic takes that shortcut instead of bouncing through multiple third-party networks.

Public Peering vs Private Peering

Peering comes in two distinct forms, each with different characteristics, costs, and use cases. Understanding the difference is essential for evaluating how well your ISP is connected to the networks your business depends on.

Public Peering

Public peering occurs at Internet Exchange Points (IXPs) — shared facilities where multiple networks connect to a common switching fabric. Rather than establishing individual connections with every network, a participant connects once to the exchange and can then peer with any other member. It’s the most cost-effective way for networks to establish peering relationships at scale.

At a public exchange, all participants share the same physical switching infrastructure. Traffic is exchanged over a shared VLAN, and each network’s router sits on the exchange’s peering LAN alongside hundreds of other routers. A single port on the exchange fabric gives a network access to peer with every other connected member, provided both parties agree to the peering arrangement.

The advantages of public peering are clear: low cost per peer, access to a large number of networks, and the ability to establish new peering sessions quickly. The trade-off is that the shared infrastructure means less control over capacity and potentially higher latency during peak periods, though modern exchanges are engineered to handle enormous throughput with minimal congestion.

Private Peering

Private peering involves a dedicated, point-to-point connection between two specific networks, typically via a cross-connect in a colocation data centre. There’s no shared switching fabric — just a direct fibre link between the two networks’ routers. This provides guaranteed bandwidth, lower latency, and complete isolation from other traffic.

Private peering is typically established between networks that exchange very large volumes of traffic — often tens or hundreds of gigabits per second. The cost is higher than public peering because each relationship requires its own physical infrastructure, but the performance and reliability benefits justify it for high-volume, business-critical interconnections.

In practice, most well-connected UK ISPs use a combination of both. They maintain public peering at major exchanges like LINX for broad connectivity, supplemented by private peering with the networks that carry the majority of their traffic — typically the large cloud providers (Microsoft, Google, Amazon), content delivery networks (Cloudflare, Akamai), and other Tier 1 carriers.

Internet Exchange Points and the UK Peering Landscape

The United Kingdom has one of the most mature and well-developed peering ecosystems in the world, anchored by the London Internet Exchange (LINX) — one of the largest IXPs on the planet. The UK’s peering infrastructure is a significant competitive advantage for British businesses, providing access to low-latency, high-capacity interconnections with global networks.

LINX: The Heart of UK Internet Peering

Founded in 1994, LINX operates the UK’s primary internet exchange with points of presence across multiple London data centres, including facilities in Docklands, Slough, and the City. With over 950 member networks and peak traffic exceeding 4.3 Tbps, LINX is one of the most important internet infrastructure assets in Europe.

LINX provides the shared switching fabric where ISPs, cloud providers, content networks, and enterprises meet to exchange traffic. When your ISP is a LINX member, your traffic to other LINX-connected networks can be exchanged locally in London rather than being routed through international transit networks. This keeps UK-to-UK traffic within the country, reducing latency and improving reliability.

LINX also operates LINX NoVA (in Northern Virginia, USA) and has interconnections with other major exchanges, providing UK networks with efficient paths to international destinations. For UK businesses with operations or customers in North America, this transatlantic peering infrastructure is particularly valuable.

Other UK Internet Exchanges

While LINX dominates the UK peering landscape, several other exchanges serve important roles:

• LONAP (London Network Access Point) — an alternative London exchange with over 100 members, often favoured by smaller ISPs and enterprises seeking a less congested peering environment
• IXManchester — serving the North of England, critical for reducing latency for businesses in Manchester, Leeds, Liverpool, and surrounding areas
• IXScotland — based in Edinburgh, keeping Scottish internet traffic local rather than routing it through London
• IXLeeds — a growing exchange serving Yorkshire and the wider North, reducing dependence on London-centric routing
• LINX Cardiff — extending LINX’s reach into Wales

The presence of regional exchanges is increasingly important for UK businesses outside London. Without a local exchange, traffic between two businesses in Manchester that use different ISPs might need to travel to London and back — adding 10–20 ms of unnecessary latency. Regional exchanges keep local traffic local, which is critical for latency-sensitive applications like VoIP, video conferencing, and real-time collaboration tools.

Transit vs Peering: Understanding the Difference

To fully appreciate why peering matters, you need to understand the alternative: IP transit. These are the two fundamental mechanisms by which networks connect to the rest of the internet, and every ISP uses a combination of both. The balance between them directly affects the quality of service you receive.

IP Transit

IP transit is a commercial arrangement where one network pays another (typically a larger, Tier 1 carrier) for access to the entire internet routing table. The transit provider agrees to carry the customer’s traffic to any destination on the internet, regardless of where it needs to go. It’s the “default route” — the fallback path for any traffic that doesn’t have a more specific peering route available.

Transit is essential: no network can peer with every other network on the internet, so transit fills the gaps. However, transit traffic typically traverses more network hops, passes through more congested links, and is subject to the transit provider’s own capacity constraints and routing decisions. It’s also more expensive for your ISP, which is why aggressive peering strategies are commercially attractive.

Peering

Peering, as we’ve discussed, is a direct exchange between two networks. It’s faster, more predictable, and typically settlement-free. The more peering relationships your ISP maintains, the less traffic needs to traverse transit networks, and the better your overall internet experience.

A well-run ISP aims to peer directly with the networks that carry the majority of its customers’ traffic — the so-called “heavy hitters.” In the UK, this typically means direct peering or private network interconnections with Microsoft (for Office 365 and Azure), Google (for Workspace and GCP), Amazon (for AWS), Cloudflare, Akamai, and the major UK content and media networks. Traffic to these destinations might account for 60–80% of a typical business ISP’s total volume. If these flows are peered rather than transited, the performance improvement for end users is substantial.

How Peering Affects Business Internet Quality

For UK businesses, the practical impact of peering quality shows up in three critical areas: latency, reliability, and consistency. These aren’t abstract technical metrics — they directly affect your employees’ productivity, your customers’ experience, and your operational efficiency.

Latency: The Speed You Actually Feel

Bandwidth tells you how much data your connection can carry. Latency tells you how quickly that data arrives. For interactive applications — VoIP, video conferencing, remote desktops, real-time collaboration, and cloud-hosted business software — latency is far more important than bandwidth. And peering is the single biggest factor in determining latency for UK businesses accessing cloud services.

Consider a practical example. Your office in Birmingham connects to Microsoft 365 via an ISP that peers directly with Microsoft at LINX in London. The traffic path is: your office → ISP’s Birmingham PoP → ISP’s London backbone → LINX peering point → Microsoft’s network. Total latency: approximately 8–12 ms. Now consider the same office using an ISP without Microsoft peering. The traffic might go: your office → ISP → transit provider in London → transit provider’s European backbone → another transit provider → Microsoft’s network entry point (possibly in Amsterdam or Frankfurt) → back to Microsoft’s London servers. Total latency: 25–45 ms, with significantly higher variability.

That 15–35 ms difference might sound trivial, but it compounds with every interaction. Every click in a web application, every keystroke in a cloud document, every packet in a voice call is affected. Across a full working day for a team of 30 people, those milliseconds translate into tangible productivity losses and noticeably degraded user experience.

Reliability: Fewer Points of Failure

Every network hop your traffic crosses is a potential point of failure. Routers can crash, fibre links can be cut, and transit networks can experience outages. When your ISP peers directly with a destination network, your traffic crosses fewer boundaries and depends on fewer third-party networks to function correctly. This inherently improves reliability.

Well-peered ISPs also typically maintain multiple peering sessions with critical networks — at different exchanges, in different data centres, via both public and private peering. This redundancy means that if one peering link fails, traffic automatically reroutes via an alternative path, often without any noticeable disruption. ISPs that rely heavily on transit have less control over these failover paths and may experience longer, more unpredictable outages when a transit provider has problems.

Consistency: Predictable Performance Throughout the Day

Transit links are shared resources. During peak hours — typically 09:00–17:00 for business traffic and 19:00–23:00 for consumer traffic — these shared links can become congested, causing performance degradation exactly when you need your connection most. Peering links, especially private peering, are sized for the specific traffic volumes between two networks and are far less susceptible to this kind of congestion.

The result is a connection that performs consistently whether it’s 03:00 on a Sunday or 11:00 on a Monday morning. For businesses running time-sensitive operations — trading platforms, healthcare systems, customer-facing e-commerce, or 24/7 support centres — this consistency is non-negotiable.

Choosing an ISP with Good Peering

Most businesses choose their ISP based on price, bandwidth, and perhaps SLA terms. Very few consider peering quality — yet it’s often the single most important differentiator between providers offering otherwise identical service tiers. Here’s how to evaluate an ISP’s peering credentials and make a more informed choice.

Questions to Ask Your ISP

• “Are you a member of LINX and/or LONAP?” — Membership of a major UK exchange is a baseline requirement. If they’re not present at any UK exchange, their peering capability is severely limited.
• “Do you maintain direct peering with Microsoft, Google, Amazon, and Cloudflare?” — These four networks carry the vast majority of UK business internet traffic. Direct peering with all of them should be standard for any serious business ISP.
• “How many networks do you peer with?” — More peers generally means better routing options. A well-connected UK ISP might peer with 200–500+ networks.
• “What is your peering policy?” — ISPs with an “open” peering policy peer with anyone who asks. Those with “selective” or “restrictive” policies may have fewer peering relationships, though the ones they maintain might be higher capacity.
• “Can you provide latency measurements to specific cloud platforms?” — Any ISP confident in their peering will happily share this data. Reluctance to provide it is a red flag.
• “What is your AS number?” — With their AS number, you can look up their peering relationships on PeeringDB (peeringdb.com), a public database where networks voluntarily register their peering information.

Red Flags to Watch For

• No presence at any UK internet exchange — This means all their traffic is carried via transit, with no local peering benefit.
• Reliance on a single transit provider — If that provider has issues, the ISP’s entire customer base is affected.
• Unable or unwilling to discuss peering — If their sales team doesn’t know what peering is, their network engineering is unlikely to be sophisticated.
• Reselling another provider’s network — Many budget ISPs are resellers with no control over routing or peering decisions. Your traffic quality is entirely at the mercy of the underlying carrier.
• No UK-based network operations centre (NOC) — Peering issues require rapid response from network engineers who understand the UK interconnection landscape.

What Good Peering Looks Like for a UK Business ISP

A well-peered UK business ISP typically has the following characteristics:

• Membership of LINX (and ideally LONAP or a regional exchange for redundancy)
• Port capacity of 10 Gbps or more at major exchanges
• Direct private peering with Microsoft, Google, Amazon, and Cloudflare
• Multiple upstream transit providers (at least two, ideally three) for redundancy
• Presence in Tier 3+ data centres like Equinix, Telehouse, or Global Switch in London
• An open or selective peering policy with 200+ established peers
• Published AS number and PeeringDB entry with up-to-date information

Latency Implications: What the Milliseconds Mean

Understanding latency in the context of peering requires moving beyond abstract numbers and connecting them to real-world business impact. Different applications have different latency sensitivities, and the difference between a well-peered and poorly-peered ISP can push your applications from “excellent” into “frustrating” territory.

Application	Acceptable Latency	Noticeable Degradation	Unusable Beyond
VoIP phone calls	< 80 ms	80–150 ms	> 150 ms
Video conferencing (Teams, Zoom)	< 100 ms	100–200 ms	> 200 ms
Remote desktop (RDP, Citrix)	< 50 ms	50–100 ms	> 150 ms
Cloud CRM & ERP (Salesforce, Xero)	< 100 ms	100–250 ms	> 300 ms
Real-time collaboration (Google Docs)	< 75 ms	75–150 ms	> 200 ms
Email & file sync	< 200 ms	200–500 ms	> 500 ms

A UK ISP with strong peering to Microsoft’s network will typically deliver 5–15 ms latency to Microsoft 365 services from most UK locations. A poorly-peered provider might deliver 30–60 ms to the same services. For VoIP and video, this difference is the gap between “seamless” and “noticeably laggy.” For remote desktop users, it’s the difference between a responsive session and one where every mouse click has a visible delay.

Crucially, latency also affects jitter (the variation in latency between packets) and packet loss. Traffic that takes a longer, more congested path through multiple transit networks is more likely to experience variable delays and occasional dropped packets. VoIP and video conferencing are exceptionally sensitive to both: jitter above 30 ms causes audio distortion, and packet loss above 1% causes visible video artefacts and audio dropouts.

The Broader UK Peering Landscape

The UK’s peering ecosystem is evolving rapidly, driven by several trends that directly affect business internet quality.

The Decentralisation of Peering

Historically, almost all UK peering happened in London. Traffic from Edinburgh to Edinburgh would travel 600 miles south to London, cross a peering point, and travel 600 miles back north. This is changing. The growth of regional exchanges in Manchester, Leeds, Edinburgh, and Cardiff is keeping more local traffic local. For businesses outside London, choosing an ISP that peers at regional exchanges as well as in London can deliver measurably lower latency.

Cloud Provider On-Ramps

Major cloud providers are increasingly deploying their own edge infrastructure within UK data centres, creating direct on-ramps that bypass the public internet entirely. Microsoft’s Azure ExpressRoute, AWS’s Direct Connect, and Google’s Cloud Interconnect all offer dedicated, private connections from your ISP’s network directly into the cloud provider’s infrastructure. These services build on peering concepts but offer even lower latency, higher bandwidth, and guaranteed service-level agreements.

If your ISP has a presence in the same data centres as these cloud on-ramps, they can offer you dedicated cloud connectivity that avoids the public internet entirely. This is particularly valuable for businesses with heavy cloud workloads, data sovereignty requirements, or applications that demand consistent sub-10ms latency to cloud platforms.

Content Delivery Networks & Edge Caching

CDNs like Cloudflare, Akamai, and Fastly deploy cache servers inside ISP networks and at peering points, bringing popular content closer to end users. When your ISP peers effectively with these CDNs — or better yet, hosts CDN edge nodes within their own network — common content and web applications are served from nearby caches rather than distant origin servers. This dramatically reduces latency for web browsing and SaaS application delivery.

The Growing Importance of IPv6 Peering

As IPv6 adoption accelerates globally, the quality of your ISP’s IPv6 peering matters increasingly. Some ISPs maintain excellent IPv4 peering but neglect their IPv6 interconnections, resulting in degraded performance for IPv6-only services and dual-stack environments. As more content and cloud services move to IPv6, this gap will become more significant. Ask your ISP about their IPv6 peering coverage alongside IPv4.

Practical Steps for UK Businesses

Understanding peering is valuable, but translating that knowledge into better business outcomes requires concrete action. Here are the practical steps any UK business can take to benefit from better peering.

Audit Your Current ISP’s Peering

• Ask your ISP for their AS number and look them up on PeeringDB
• Run traceroutes from your office to your key cloud services and count the hops
• Measure latency to Microsoft 365, Google Workspace, and any other critical cloud platforms at different times of day
• Compare these measurements against what other ISPs advertise for the same destinations

Factor Peering into Your ISP Selection

• Include peering questions in your RFP or tender process
• Request trial connections or proof-of-concept periods to validate real-world performance
• Weight latency and peering quality alongside price and bandwidth in your evaluation criteria
• Consider ISPs that offer dedicated cloud on-ramps if you’re a heavy Azure, AWS, or GCP user

Consider Multi-ISP or SD-WAN Strategies

• If no single ISP provides optimal peering to all your critical services, a dual-ISP setup with policy-based routing can direct specific traffic over the best-peered connection
• SD-WAN solutions can dynamically select the optimal path per application, leveraging the peering strengths of each provider
• This approach adds cost and complexity but delivers the best possible performance for businesses with diverse cloud dependencies

Why This Matters More Than Ever

The shift to cloud-first, hybrid-working business models has fundamentally changed what “good internet” means for UK businesses. A decade ago, the internet was primarily used for email and web browsing — applications that are tolerant of moderate latency and occasional packet loss. Today, the internet is the network. Your phone system runs over it. Your file storage lives on it. Your line-of-business applications are hosted in the cloud. Your remote workers depend on it entirely.

In this environment, the quality of the path your traffic takes — determined largely by your ISP’s peering relationships — is as important as the bandwidth of your physical connection. A 100 Mbps leased line from a well-peered ISP will outperform a 1 Gbps connection from a poorly-peered one for the applications that matter most to your business. Understanding peering gives you the knowledge to make that distinction and choose accordingly.

The UK’s peering infrastructure — anchored by LINX and increasingly supported by regional exchanges — provides an excellent foundation. But that foundation only benefits your business if your ISP actively participates in it. Not all providers are equal, and the difference in end-user experience between a well-peered and poorly-peered ISP is dramatic.