Website Speed Benchmarks: What Belfast Businesses Should Aim For

"Is Our Website Fast Enough?"

It is one of the most common questions we get asked, and the honest answer is that "fast" on its own is not a target. Fast compared to what? Measured how? For which page, on which device, on which connection?

A site that scores well on a developer's laptop in the office can be genuinely painful on a phone on mobile data on the way into town. And a site that is objectively slow can still outrank a faster competitor if everything else about it is stronger.

What you need is not a vague ambition to be fast. You need thresholds that are defined, a way of measuring against your actual competitors, and an order of operations for when you fall short.

The Thresholds That Are Actually Defined

Most speed advice is vibes. Core Web Vitals are not — they are published, specific, and used by Google as a ranking signal. Start here, because these are the only numbers in this article that come with official pass marks attached.

Largest Contentful Paint (LCP)

How long until the main content of the page appears. Usually your hero image, heading or main banner.

• Good: 2.5 seconds or under
• Needs improvement: 2.5 to 4 seconds
• Poor: over 4 seconds

LCP is the one most local business sites fail, and it is almost always an image problem.

Interaction to Next Paint (INP)

How quickly the page responds when someone taps or clicks. It replaced First Input Delay and it is considerably harder to pass, because it measures every interaction rather than just the first.

• Good: 200 milliseconds or under
• Needs improvement: 200 to 500 milliseconds
• Poor: over 500 milliseconds

INP problems are JavaScript problems. Heavy page builders, chat widgets and tracking scripts are the usual culprits.

Cumulative Layout Shift (CLS)

How much the page jumps about while loading. The reason you tap the wrong thing because an ad or banner loaded late and shoved the content down.

• Good: 0.1 or under
• Needs improvement: 0.1 to 0.25
• Poor: over 0.25

CLS is usually the cheapest of the three to fix — images without dimensions, late-loading fonts, injected banners.

The important detail: these are assessed at the 75th percentile of real visits. Passing means three quarters of your actual visitors get a good experience, not that it looked fine once when you tested it. Our Core Web Vitals guide covers the optimisation work in detail.

Lab Data Versus Field Data

This distinction explains most of the confusion about website speed, and it is worth getting straight before you benchmark anything.

Lab data is a simulated test — PageSpeed Insights' Lighthouse score, GTmetrix, WebPageTest. Consistent, repeatable, immediately available, and useful for diagnosis. It is a controlled experiment, not a measurement of your customers.

Field data is what real visitors actually experienced, gathered from Chrome users over the previous 28 days. This is what counts for ranking, and it is what appears at the top of a PageSpeed Insights report when enough traffic exists.

The Lighthouse performance score out of 100 is lab data. It is a rough diagnostic, it fluctuates between runs, and it is not a ranking factor. Chasing a green 100 while your field data sits in the red is a common and expensive mistake.

Use lab data to find problems. Use field data to know whether you have one.

Low-traffic sites — which describes a lot of Belfast SMEs — may not have enough visitors to generate field data at all. In that case, lab data plus mobile testing on a real phone is what you have, and you should weight the mobile results heavily.

Benchmarking Against Your Actual Competitors

Generic industry averages are close to useless for a local business. What matters is how you compare to the five or six businesses appearing alongside you in Belfast search results. Here is how to find out, and it costs nothing but an hour.

Build the comparison

1. Search your main terms — "web design Belfast", "emergency plumber Belfast", whatever applies — from a Belfast location, in an incognito window.
2. List the top competitors appearing organically and in the local pack. Six is plenty.
3. Run each through PageSpeed Insights, on mobile. Record LCP, INP and CLS from the field data section where it exists, and the Lighthouse score where it does not.
4. Test the pages that matter, not just homepages. Your key service page against their key service page.
5. Put it in a spreadsheet with your own site in the same list.

Read the results honestly

• If you are the slowest in the list, speed is a competitive liability and worth prioritising now.
• If you are middling but everyone fails Core Web Vitals, there is a real opportunity — passing when your competitors do not is a genuine differentiator.
• If you pass and they pass, speed is table stakes in your market. Further optimisation has diminishing returns and your effort belongs elsewhere, probably in content and local SEO.

Re-run this quarterly. Competitor sites get rebuilt, and a market where everyone was slow can tighten up in a year.

Test like a customer, not like a developer

Your office connection flatters your website. Before you conclude anything:

• Test on an actual mid-range Android phone, not just an iPhone and not just Chrome DevTools throttling.
• Test on mobile data rather than wifi.
• Test with a cleared cache — your repeat visits are much faster than a first-time visitor's.
• Test the pages people actually land on from search, which are often not your homepage.

What Usually Goes Wrong, in Order of Frequency

After enough audits, the same issues recur. This is roughly the order in which we find them on Northern Ireland small business sites.

1. Images

Comfortably the biggest single cause of slow local business websites. A photographer supplies 4000-pixel JPEGs, they go straight into the CMS, and every visitor downloads several megabytes to see a 600-pixel-wide image.

Fix: serve modern formats, size images to their display dimensions, compress properly, and lazy-load anything below the fold — while explicitly not lazy-loading your LCP image, which is a mistake that makes things worse. Our guide to image optimisation covers the specifics.

Typical impact: large. This alone often moves a failing LCP into passing.

2. Too many third-party scripts

Analytics, tag manager, chat widget, review widget, booking system, two remarketing pixels, a cookie banner and a font service. Each one loads code from someone else's server, and you are at the mercy of all of them.

Fix: audit what is actually there, remove what nobody uses, defer what is not needed immediately. The chat widget nobody has answered in eight months is costing you real INP.

3. Bloated themes and page builders

An off-the-shelf theme built to do everything ships the code for everything, whether your site uses it or not. Heavy page builders compound it.

Fix: this is usually not fixable by tweaking. It is a rebuild decision, and it is one of the clearest arguments for a purpose-built site over a template.

4. No caching

Every visitor causes the server to rebuild pages that have not changed. Straightforward to fix and frequently overlooked. See our caching and speed guide.

5. Cheap or distant hosting

Budget shared hosting, or a server physically far from your customers. For a business serving Northern Ireland, hosting located in the UK or Ireland with a CDN in front of it is the sensible default.

Fix: better hosting is usually the single cheapest meaningful performance improvement available, and it is often a smaller monthly figure than people expect.

6. Fonts

Multiple custom font families, several weights each, loading before any text renders. Fix: fewer families, fewer weights, font-display: swap, self-host where practical.

What Speed Is Actually Worth

Be realistic about this, because speed gets oversold.

Speed is a ranking signal, but a modest one. Google has been consistent that relevance and quality dominate. A fast page about the wrong thing does not outrank a slower page that answers the question. If you are slow and your competitors are slow, fixing speed alone will not vault you to the top.

Where speed pays more reliably is conversion and abandonment. Every second of delay loses people who never see your site at all, and they leave without registering anywhere in your analytics except as a bounce. That effect is strongest on mobile, on data, which is exactly where local searches happen.

The practical position: get into the "good" band on Core Web Vitals, because it is achievable and it removes a liability. Then stop optimising and spend the effort on conversion and content, where the returns are larger.

A Sensible Order of Work

If you have just benchmarked and found you are short:

1. Fix images. Biggest impact, lowest risk, cheapest.
2. Audit third-party scripts. Remove the dead ones.
3. Enable and configure caching.
4. Sort fonts and layout shift. Quick CLS wins.
5. Review hosting if the first four have not been enough.
6. Consider a rebuild if the platform itself is the constraint.

Measure before and after each step, using field data where you have it. Doing all six at once means never knowing which one mattered.

Where to Start

If you want to know where your site actually stands — measured properly, compared against the Belfast businesses you are competing with, with a prioritised list of what to fix — we offer a free website audit that covers exactly that.

For ongoing work, our website development and performance practice builds sites that pass Core Web Vitals from the start, rather than treating speed as something to retrofit. Get in touch if that is what you need.