Performance Optimisation

Faster for the people actually using your site

We optimise against field data from your real visitors — not a lab score from a server in Oregon. Measured before, measured after, held there every month.

A caching plugin is not a performance strategy

It hides the symptom for logged-out visitors on cached pages. Your checkout, your search, your logged-in users and your mobile traffic on a bad connection all still hit the slow path underneath.

Lab scores mislead

A 98 in a testing tool means nothing if your customers are on 4G with an older phone. We optimise against the 75th percentile of your real traffic.

Speed decays

A one-off optimisation project is undone within two quarters by new images, scripts and tags. Performance has to be maintained, not delivered.

The bottleneck is rarely the front end

Most WordPress slowness is server-side: unindexed queries, bloated autoloaded options, a plugin doing an API call on every page load.

Measure, fix the biggest thing, measure again

Baseline from real users

Field data for LCP, INP and CLS across your top templates, split by device and connection. Lab traces only to explain what the field data shows.

Server and database first

Slow query profiling, autoloaded option cleanup, object caching, PHP version and worker tuning. This is where the seconds usually are, and where most vendors never look.

Cut the payload

Plugin-by-plugin audit of what each one loads and where. Unused CSS and JS removed, scripts deferred, third-party tags challenged — every tag has to justify its cost.

Images and fonts

AVIF and WebP with correct sizes per breakpoint, explicit dimensions to stop layout shift, lazy loading below the fold, self-hosted subset fonts with a sane fallback.

Cache and edge, last

Page cache, object cache and CDN configured once the underlying site is genuinely fast — so caching is an accelerator, not a disguise.

Hold the line

Budgets set per template and monitored continuously. If a deploy pushes LCP past its budget, we’re alerted and it’s fixed — before you notice it in your analytics.

We commit to the numbers, in writing

On Operate Pro, your site’s Core Web Vitals targets are written into the agreement. If field data falls outside budget for a full month and it isn’t caused by a change you asked for, we work it back at no extra cost.

What changed, and what it was worth

Largest contentful paint on a catalogue template, field data, 90 days before and after.

Mobile add-to-cart rate over the same window. Attribution is correlational — we report it as such, not as a guarantee.

Two ways to buy speed

A one-time overhaul fixes today. A plan keeps it fixed — most sites need both, in that order.

One-time overhaul

One-time project · all six steps · before/after report

Operate Pro

Budgets monitored continuously · 4 hours dev time

Operate

Monitoring and monthly reporting · no committed budget

Performance questions

Will you break our site making it faster?

Every change goes through staging with visual regression checks first, exactly like an update. Aggressive optimisation is where most speed work breaks sites, so we ship it the same careful way.

Do we need to change hosting?

Usually not. We tune what you’re on and only recommend a move when the platform is genuinely the ceiling — and we don’t resell hosting, so there’s no commission behind that advice.

What about our page builder? We hear they’re slow.

Builders add weight, but they’re rarely the whole story. We measure what your specific build actually costs and cut it down; rebuilding templates is a last resort, quoted separately.

How long until we see a difference?

The overhaul runs two to three weeks. Server and database work usually shows up in field data within days; front-end changes take about 28 days to fully appear in Core Web Vitals reporting.

Can you guarantee a PageSpeed score of 100?

No, and anyone who does is selling you a lab number. We commit to field-data budgets for real users — that’s what affects rankings and revenue.

See where your seconds are going

A $499 audit against your real field data: which templates are slow, which plugins are responsible, and what the three biggest wins would be. Findings are yours either way.

Fixed price. Delivered within two business days.