How to Set Up Quality of Service for Business Applications

Every modern business depends on its network. From VoIP calls that connect your sales team to clients, to cloud-based ERP systems that keep operations running, to video conferences that replace the daily commute — your applications are only as reliable as the network that carries them. Yet most UK businesses treat all network traffic equally, letting a junior’s YouTube stream compete with the finance director’s SAP transaction for the same bandwidth. The result? Choppy calls, frozen screens, and a creeping sense that your expensive internet connection simply isn’t up to the job.

The truth is, the connection is usually fine. The problem is prioritisation — or rather, the total lack of it. That’s where Quality of Service (QoS) comes in. QoS is a set of network management techniques that allow you to prioritise certain types of traffic over others, ensuring your mission-critical business applications always get the bandwidth, low latency, and minimal packet loss they need to perform reliably.

This guide walks you through everything you need to know about setting up QoS for your business applications — from understanding the core concepts and identifying which applications matter most, to configuring your equipment, testing your setup, and avoiding the common pitfalls that trip up even experienced IT teams. Whether you’re an IT manager at a 50-person firm in Manchester or a business owner running a growing agency in Bristol, this is your practical, no-nonsense roadmap to a better-performing network.

What Is Quality of Service and Why Does It Matter?

Quality of Service (QoS) is a broad term for any technology or configuration that manages network traffic to reduce packet loss, latency, and jitter. In practical terms, it means telling your network equipment: “When things get busy, make sure these applications get served first.”

Think of it like a motorway with a dedicated lane for emergency vehicles. Without QoS, every packet of data — whether it’s a critical VoIP call, a massive backup upload, or someone streaming a podcast — fights for space in the same lanes. When congestion hits, everything slows down equally, and the applications that are most sensitive to delay (voice, video, real-time collaboration) suffer disproportionately.

With QoS enabled and properly configured, your router or switch examines each packet, classifies it according to rules you’ve defined, and places it into a priority queue. High-priority traffic — such as VoIP or your cloud CRM — gets processed first. Lower-priority traffic — like software updates or social media — waits until there’s capacity available.

The Business Case for QoS in the UK

The shift to cloud-first working has made QoS more important than ever for UK businesses. According to Ofcom’s 2025 Connected Nations report, 87% of UK businesses now rely on at least one cloud-hosted application for daily operations. Microsoft 365 alone is used by over 4.2 million UK businesses. When your entire operation runs over an internet connection, the quality of that connection isn’t just an IT concern — it’s a business-critical dependency.

The problem is particularly acute for SMEs. Large enterprises typically have dedicated MPLS circuits, SD-WAN overlays, and in-house network engineering teams to manage traffic. A 30-person accountancy firm in Leeds or a growing recruitment agency in Birmingham usually has a single broadband or leased line connection, a business-grade router, and no QoS configuration whatsoever. This is the gap that proper QoS setup fills — giving smaller businesses enterprise-grade traffic management without enterprise-grade costs.

Understanding the Core QoS Concepts

Before you start configuring anything, you need to understand the key metrics that QoS manages. These are the same metrics that determine whether your applications feel “fast and responsive” or “slow and unreliable.”

Bandwidth

Bandwidth is the total capacity of your connection, measured in megabits per second (Mbps). A typical UK business leased line provides between 100 Mbps and 1 Gbps symmetrical. A standard FTTP broadband connection might offer 300–900 Mbps download but only 50–100 Mbps upload. QoS doesn’t create more bandwidth — it allocates what you have more intelligently.

Latency

Latency is the time it takes for a packet to travel from source to destination, measured in milliseconds (ms). For voice calls, anything above 150ms one-way becomes noticeable. For video, 200ms starts to feel awkward. For a standard web application, up to 400ms is generally acceptable. QoS reduces latency for priority traffic by moving it to the front of the processing queue.

Jitter

Jitter is the variation in latency from one packet to the next. If one packet takes 20ms and the next takes 120ms, you have 100ms of jitter. Voice and video applications are extremely sensitive to jitter because they need a steady, predictable stream of data to produce clear audio and smooth video. QoS mitigates jitter by ensuring priority traffic is processed consistently.

Packet Loss

Packet loss occurs when data packets fail to reach their destination, typically because a congested router or switch has dropped them. Even 1% packet loss can make a VoIP call sound robotic or cause a video feed to freeze. QoS reduces packet loss for critical applications by ensuring their packets are the last to be dropped when buffers fill up.

Application Type	Maximum Latency	Maximum Jitter	Maximum Packet Loss	Bandwidth Priority
VoIP / Telephony	150ms	30ms	1%	Highest (EF)
Video Conferencing	200ms	50ms	1%	High (AF41)
Cloud ERP / CRM	300ms	100ms	0.5%	High (AF31)
Email / Collaboration	500ms	200ms	2%	Medium (AF21)
Web Browsing	1,000ms	N/A	3%	Low (AF11)
Backups / Updates	N/A	N/A	5%	Lowest (BE)

Step 1: Audit Your Network Traffic

You cannot prioritise what you cannot see. The first step in any QoS deployment is understanding exactly what traffic is flowing across your network, how much bandwidth each application consumes, and when peak usage occurs.

How to Conduct a Traffic Audit

Use your router’s built-in monitoring. Most business-grade routers from manufacturers like Cisco, Meraki, Ubiquiti, FortiGate, and DrayTek include traffic analysis dashboards. These will show you top applications by bandwidth consumption, peak usage times, and per-device statistics. If your router doesn’t offer this, it may be time for an upgrade.

Deploy a network monitoring tool. For deeper visibility, tools like PRTG Network Monitor (which has a free tier for up to 100 sensors), ManageEngine NetFlow Analyzer, or the open-source ntopng can analyse NetFlow or sFlow data from your switches and routers. These tools give you granular, per-application, per-user visibility over days or weeks.

Run the audit for at least one full working week. Network patterns vary significantly between Monday mornings (when everyone logs in and syncs) and Friday afternoons (when usage typically drops). A single snapshot will mislead you. Capture a full week, including any periods of known poor performance, to get an accurate baseline.

Document your findings. Create a simple spreadsheet listing each application, its average bandwidth consumption, peak bandwidth consumption, sensitivity to latency and jitter, and its business criticality. This document becomes the foundation of your QoS policy.

Typical bandwidth distribution for a 40-person UK professional services firm, based on CloudSwitched client data.

Step 2: Classify Your Applications into Priority Tiers

Once you know what’s on your network, the next step is to sort every application into a priority tier. Most QoS implementations use between three and five tiers. We recommend four for most UK SMEs — it provides enough granularity without becoming unmanageable.

Tier 1: Real-Time / Mission-Critical

These are applications where even small amounts of delay or packet loss cause immediate, noticeable degradation. VoIP telephony, video conferencing (Microsoft Teams, Zoom, Google Meet), and real-time trading platforms fall into this category. These applications typically use well-known ports or protocols (SIP, RTP) that are easy to identify and prioritise.

Tier 2: Business-Critical Interactive

These are the applications your staff use throughout the day to do their jobs. Cloud CRM systems (Salesforce, HubSpot), ERP platforms (Sage, NetSuite, SAP Business One), accounting software (Xero, QuickBooks Online), and practice management systems. These applications are sensitive to latency but more tolerant than real-time traffic. Users notice when they’re slow, but they don’t break completely.

Tier 3: Standard Business Traffic

Email (Exchange Online, Gmail), file sharing (SharePoint, OneDrive, Google Drive), internal wikis, project management tools (Monday.com, Asana), and general web browsing related to work. This traffic is important but can tolerate moderate delays without significantly impacting productivity.

Tier 4: Best-Effort / Low Priority

Software updates (Windows Update, application patches), cloud backups, personal web browsing, streaming media, social media, and any non-business traffic. This traffic gets whatever bandwidth is left over after the higher tiers have been served. Importantly, it still works — it just yields capacity when the network is congested.

Step 3: Choose Your QoS Method

There are several approaches to implementing QoS, and the right choice depends on your network equipment, your connection type, and the complexity of your environment. Here are the three most common methods used in UK business networks.

DSCP Marking Explained

Differentiated Services Code Point (DSCP) is the most widely used QoS classification method in modern networks. It works by writing a small value into the header of every IP packet, telling every device that handles that packet what priority level it should receive.

The most common DSCP values for business networks are:

DSCP Value	Name	Per-Hop Behaviour	Typical Use
46	EF (Expedited Forwarding)	Low loss, low latency, guaranteed bandwidth	VoIP / real-time voice
34	AF41	High priority, low drop probability	Video conferencing
26	AF31	Medium-high priority	Cloud ERP, CRM, LOB apps
18	AF21	Medium priority	Email, file sharing
10	AF11	Low priority	General web browsing
0	BE (Best Effort)	No guarantees	Backups, updates, personal use

Which Queuing Mechanism to Use

Once traffic is classified, your router or switch needs a queuing mechanism to decide what gets sent first. The three most common options are:

Strict Priority Queuing (PQ): The highest-priority queue is always served first. Simple and effective for VoIP, but risks starving lower queues if high-priority traffic spikes. Use this only for a small, well-controlled amount of real-time traffic (typically no more than 10–15% of total bandwidth).

Weighted Fair Queuing (WFQ): Bandwidth is divided among queues according to configurable weights. A queue with a weight of 40% will receive roughly 40% of available bandwidth during congestion. This is the most common approach for business networks and the default on many Cisco routers.

Class-Based Weighted Fair Queuing (CBWFQ): Combines strict priority for real-time traffic with weighted fair queuing for everything else. This is the gold standard for business QoS — VoIP gets absolute priority up to a defined limit, and all other traffic is handled fairly according to weights. Most modern business routers support this out of the box.

Step 4: Configure QoS on Your Network Equipment

The specific configuration steps vary depending on your hardware, but the general process is consistent across all platforms. Here’s a practical walkthrough of the key stages.

Configuration on a Typical Business Router

Access your router’s management interface. For most business routers, this is a web-based dashboard accessible via the device’s IP address (commonly 192.168.1.1 or 10.0.0.1). Some platforms, particularly Cisco IOS, require command-line configuration via SSH or console cable.

Define your traffic classes. Create classification rules that match traffic to your priority tiers. Common classification methods include:

• Source/destination IP address: Route all traffic to your VoIP provider’s IP range into the real-time queue
• Port numbers: SIP traffic on UDP 5060, RTP on UDP 16384–32767, HTTPS on TCP 443
• DSCP markings: Trust markings from VoIP phones or other devices that self-mark
• Application signatures: Deep packet inspection to identify applications regardless of port (available on UTM firewalls and SD-WAN platforms)

Create your queuing policy. Assign each traffic class to a queue and define bandwidth allocations. A solid starting point for a typical UK SME:

Apply the policy to your WAN interface. QoS policies should be applied to the interface facing your internet connection (the WAN interface), specifically in the outbound direction. Inbound QoS is less effective because by the time traffic arrives at your router, it’s already traversed the congested link. For inbound prioritisation, you need your ISP to apply QoS on their end (available with managed leased lines) or use an SD-WAN solution.

Set your shaping rate slightly below your actual connection speed. This is a critical and often-missed step. If your leased line provides 100 Mbps, set your QoS shaping rate to 95 Mbps. This ensures that congestion and queuing happen on your router (where your QoS policies apply) rather than at your ISP’s equipment (where they don’t). Without this, your router may send traffic faster than the ISP link can handle, causing the ISP’s equipment to drop packets indiscriminately.

Platform-Specific Guidance

Cisco Meraki: QoS is configured under Security & SD-WAN > SD-WAN & Traffic Shaping in the Meraki dashboard. You can create traffic shaping rules based on application categories and assign bandwidth limits and priority levels per application. Meraki’s built-in application identification makes this straightforward for non-specialists.

Ubiquiti UniFi: Navigate to Settings > Internet > WAN and enable Smart Queues (which implements the fq_codel algorithm). For more granular control, use Traffic Management to create custom rules. UniFi’s approach is simpler but less flexible than enterprise platforms.

FortiGate: Create traffic shaping policies under Policy & Objects > Traffic Shaping. FortiGate combines application control with traffic shaping, allowing you to identify applications by signature (even encrypted ones) and apply per-application bandwidth guarantees and limits. This is particularly powerful for businesses with complex application landscapes.

DrayTek Vigor: Popular in UK SME environments, DrayTek routers offer QoS under Bandwidth Management. You can define bandwidth classes based on IP range, port, or DSCP and set guaranteed/maximum bandwidth per class. The interface is less polished than Meraki or FortiGate but perfectly functional.

Step 5: Handle VoIP and Microsoft Teams Specifically

VoIP and Microsoft Teams deserve special attention because they are the applications most visibly affected by poor network performance. When a webpage loads slowly, users wait an extra second. When a voice call breaks up, the conversation stops entirely.

VoIP QoS Requirements

Each concurrent VoIP call requires approximately 85–100 Kbps of bandwidth using the G.711 codec, or 30–40 Kbps using G.729. For a business with 10 concurrent calls, that’s roughly 1 Mbps of strictly prioritised bandwidth — a trivial amount on a modern connection, but one that must be guaranteed without exception.

Configure your QoS policy to identify VoIP traffic by:

• DSCP marking EF (46) from IP handsets
• RTP traffic on your VoIP provider’s designated port range
• SIP signalling on UDP/TCP 5060 or 5061 (TLS)
• Destination IP ranges published by your VoIP provider

Microsoft Teams Optimisation

Microsoft publishes a regularly updated list of IP ranges and ports that Teams uses. For optimal QoS, you should prioritise:

• Audio: UDP 50000–50019 — mark as DSCP EF (46)
• Video: UDP 50020–50039 — mark as DSCP AF41 (34)
• Screen sharing: UDP 50040–50059 — mark as DSCP AF21 (18)

You can configure Windows PCs to self-mark Teams traffic using Group Policy. Navigate to Computer Configuration > Policies > Windows Settings > Policy-based QoS and create rules for each traffic type, specifying the Teams executable (Teams.exe) and the corresponding port ranges and DSCP values. This marking is honoured by your network equipment if you’ve configured trust boundaries correctly.

Step 6: Consider Your Internet Connection Type

The effectiveness of your QoS configuration depends significantly on the type of internet connection you have. Not all connections are equal when it comes to traffic management.

Leased Lines

A dedicated, symmetrical connection (typically 100 Mbps to 10 Gbps) is the gold standard for businesses that depend on real-time applications. Leased lines from providers such as BT, Virgin Media Business, and Colt offer SLAs that guarantee latency, jitter, and uptime. Many leased line providers will honour DSCP markings end-to-end, meaning your QoS policies extend beyond your router to the ISP’s network. A 100 Mbps leased line typically costs between £200 and £400 per month depending on location and contract length, rising to £500–£1,200 for gigabit connections. For businesses running VoIP, video conferencing, and cloud ERP simultaneously, this investment pays for itself rapidly.

FTTP Broadband

Full Fibre to the Premises offers excellent download speeds (up to 1.8 Gbps with some providers) but asymmetrical upload speeds (typically 50–115 Mbps on business packages). QoS is effective on FTTP, particularly for managing upload bandwidth, which is where congestion most commonly occurs. However, ISPs will not honour DSCP markings on broadband connections — your QoS is limited to your own router. Business FTTP packages cost between £30 and £70 per month, making them an attractive option for cost-conscious businesses willing to accept the asymmetry.

SD-WAN Overlays

For businesses with multiple internet connections or multiple offices, SD-WAN provides an intelligent overlay that can apply QoS policies across all connections, automatically routing critical traffic over the best-performing link. SD-WAN platforms from Cisco Meraki, Fortinet, and VMware (VeloCloud) include sophisticated QoS capabilities built in. Monthly costs range from £25 to £80 per site for the SD-WAN licence, plus the cost of internet connections at each location.

Step 7: Test, Monitor, and Refine

Configuring QoS is not a set-and-forget exercise. Network traffic patterns change as your business grows, as you adopt new applications, and as your team’s work habits evolve. A QoS policy that was perfect six months ago may need adjustment today.

Testing Your Configuration

Generate controlled test traffic. Use tools like iPerf3 to generate traffic at known rates across your network while monitoring how your QoS policy handles it. Simultaneously place a VoIP test call and check for quality degradation. If your VoIP remains crystal clear even when iPerf3 is saturating the connection, your QoS is working.

Use VoIP quality monitoring tools. Services such as Spearline, Voipmonitor, or the built-in call quality dashboards in Microsoft Teams Admin Centre and 3CX provide objective measurements of call quality (MOS scores, jitter, packet loss) over time. A Mean Opinion Score (MOS) of 4.0 or above is considered good; below 3.5 indicates problems.

Test during peak hours. Your QoS policy only activates when there’s congestion. Testing at 2 AM on a Sunday tells you nothing. Test at 9:15 AM on a Monday when everyone is logging in, syncing files, and joining their first Teams meeting of the week.

Ongoing Monitoring

Set up SNMP monitoring on your router to track queue utilisation, drop rates, and per-class bandwidth consumption. PRTG, Zabbix, and LibreNMS can all pull this data and alert you when thresholds are breached. Review your QoS statistics monthly and adjust allocations as needed.

Key metrics to watch:

• Queue drop rate: If your priority queue is dropping packets, you’ve allocated too little bandwidth to it
• Queue utilisation: If a queue is consistently below 50% utilisation, you’ve over-allocated and are wasting capacity
• Application response times: If users still complain about slowness in a prioritised application, the issue may be upstream (ISP, cloud provider) rather than local
• Best-effort starvation: If Windows Updates and cloud backups are failing consistently, your lower-tier allocations are too aggressive

Common QoS Mistakes and How to Avoid Them

Having configured QoS for hundreds of UK businesses, we’ve seen the same mistakes repeatedly. Here are the most common pitfalls and how to avoid them.

Mistake 1: Prioritising everything. If every application is high priority, nothing is. We’ve seen businesses mark 80% of their traffic as “critical,” which completely defeats the purpose of QoS. Your strict priority queue should never exceed 15–20% of total bandwidth. Be ruthless about what truly qualifies as mission-critical.

Mistake 2: Forgetting the upload direction. Most congestion in business networks occurs on the upload side, especially on asymmetrical broadband connections. If you only configure QoS for download traffic, you’re missing the primary problem. Always apply shaping rules to outbound traffic on your WAN interface.

Mistake 3: Not setting the shaping rate below line speed. As mentioned earlier, if your QoS shaping rate matches or exceeds your actual connection speed, congestion happens at the ISP’s equipment where your QoS policies don’t apply. Set your shaping rate to 90–95% of your measured connection speed.

Mistake 4: Ignoring DSCP re-marking. Many consumer devices and applications mark their traffic with high-priority DSCP values by default. If you blindly trust all DSCP markings, a single rogue application can claim priority for all its traffic. Always re-mark untrusted traffic at your network edge.

Mistake 5: Setting and forgetting. Your network changes constantly. New applications, new team members, new working patterns, and new cloud services all affect traffic flows. A QoS policy that worked perfectly in 2024 may be completely wrong for your network in 2026. Review and adjust quarterly at minimum.

Mistake 6: Not testing during congestion. QoS is only active when the network is congested. If you test during off-peak hours and see good performance, that tells you nothing about whether your QoS configuration is actually working. Always test under realistic load conditions.

QoS for Remote and Hybrid Workers

The rise of hybrid working has introduced a new challenge for QoS. When your team works from home, they’re using residential broadband connections that you don’t control. You can’t configure QoS on their home router (well, you could, but it’s impractical at scale).

There are, however, several strategies that help:

Prioritise at the application level. Microsoft Teams, Zoom, and other conferencing platforms have their own internal QoS settings. Ensure these are enabled via your organisation’s administration policies. Teams, for example, can be configured to prefer lower-bandwidth codecs when it detects a constrained connection.

Deploy a cloud-managed SD-WAN agent. Some SD-WAN providers offer lightweight agents that can be installed on remote workers’ laptops, providing application-aware traffic management even on home broadband. This is increasingly popular among UK businesses with large remote workforces.

Recommend minimum home broadband standards. Publish a clear guide for remote workers: a minimum of 30 Mbps download and 5 Mbps upload, a wired Ethernet connection for video calls (not Wi-Fi), and no other heavy users on the connection during working hours. Some UK businesses offer a £25–£50 monthly allowance toward a better broadband package as part of their remote working policy.

Use DSCP marking on the endpoint. Configure Group Policy to mark traffic from business applications with appropriate DSCP values. Even though the home router may not honour these markings, they ensure that when the traffic reaches your company VPN concentrator or cloud gateway, it can be prioritised correctly on the corporate side.

How Much Does QoS Implementation Cost?

One of the most common questions UK business owners ask is what QoS actually costs to implement. The answer varies significantly depending on your current equipment, network complexity, and whether you do it in-house or engage a managed IT provider.

Component	DIY Cost	Managed IT Provider Cost	Notes
Network audit	£0 (your time)	£300–£800	1–2 week monitoring period with detailed report
Router upgrade (if needed)	£200–£2,000	£200–£2,000 + £150 install	Only if current router lacks QoS capabilities
QoS configuration	£0 (your time)	£400–£1,200	Depends on number of sites and complexity
Testing and optimisation	£0 (your time)	£200–£500	Load testing, VoIP quality verification
Ongoing monitoring	£0–£50/month (tools)	£50–£150/month	SNMP monitoring, quarterly reviews
Total first-year cost	£200–£2,600	£1,500–£6,500	Including hardware if upgrade is needed

For most UK SMEs with existing business-grade equipment, QoS configuration is a one-time cost of £400–£1,200 when done by a managed IT provider, with minimal ongoing costs. Given that poor network performance costs the average 25-person SME approximately £8,600 per year in lost productivity, the return on investment is typically measured in weeks, not months.

When QoS Alone Isn’t Enough

QoS is a powerful tool, but it has limits. It cannot create bandwidth that doesn’t exist. If your 50-person office is running on a 20 Mbps broadband connection, no amount of QoS tuning will make Microsoft Teams, Salesforce, and cloud backups all work smoothly simultaneously. QoS optimises the allocation of existing capacity — it does not increase it.

Signs that you’ve outgrown QoS alone and need a bandwidth upgrade:

• Your priority queues are consistently at maximum capacity during business hours
• Even with QoS, VoIP quality drops below a MOS of 4.0 during peak times
• Your best-effort queue is starved for hours at a time, causing backup and update failures
• You’re adding staff or applications faster than your connection can accommodate
• Your measured connection speed has degraded below your contracted rate (contact your ISP)

In these cases, the answer is a combination of QoS and a bandwidth upgrade — typically moving from broadband to a leased line, or upgrading from a 100 Mbps leased line to a 500 Mbps or 1 Gbps connection. QoS then ensures that the larger pipe is used efficiently.

Getting Started: A Practical Checklist

Here’s a concise checklist to take your QoS implementation from concept to production:

1. Audit your traffic for a minimum of one working week using your router’s built-in tools or a dedicated monitoring solution
2. Classify applications into four priority tiers based on business criticality and sensitivity to delay
3. Verify your equipment supports QoS — check for DSCP marking, traffic shaping, and multiple queues
4. Define your bandwidth allocations — 15% strict priority for real-time, 30% business-critical, 35% standard, 10% network control, 10% best-effort
5. Set your shaping rate to 90–95% of your measured connection speed
6. Configure trust boundaries — trust DSCP from managed devices, re-mark from everything else
7. Apply the policy to your WAN interface in the outbound direction
8. Test under load during peak business hours, not at 2 AM
9. Monitor continuously with SNMP and application-level quality tools
10. Review quarterly and adjust as your applications and workforce evolve

Final Thoughts

Quality of Service isn’t glamorous, and it rarely makes the list of exciting IT projects. But for any UK business that relies on cloud applications, VoIP, and video conferencing — which is to say, virtually every UK business — it’s one of the highest-impact, lowest-cost improvements you can make to your network.

The difference between a network with properly configured QoS and one without is the difference between a Teams call that’s crystal clear and one where you spend half the meeting saying “sorry, you broke up there.” It’s the difference between Salesforce loading in under a second and your sales team staring at a spinning wheel while a colleague’s cloud backup chews through the upload bandwidth. It’s the difference between a network that supports your business and one that actively hinders it.

The good news is that QoS is well within reach of every UK business. The concepts are straightforward, the equipment is affordable, and the configuration — while it requires some technical knowledge — is a one-time effort that pays dividends every single working day. Whether you tackle it in-house or bring in a managed IT provider like CloudSwitched, the investment in getting your network priorities right is one you will never regret.