Before someone reads your services, looks at your work or decides to contact you, they have to wait for your website. That wait is part of their first impression.
At StandWatch Digital, we treat speed as part of how a website represents its owner. We work to push performance and test scores as high as we can within the site’s needs and technical limits. The point is to help visitors reach your information and use the site without unnecessary delay.
Our free Website Check gives owners a starting point: an automated report on their homepage, covering speed, phones and accessibility, security, and search basics. It can reveal problems worth fixing. It cannot tell the whole story from one score.
Why waiting matters
A slow website asks visitors to spend time before they know whether you can help them. Some will wait. Others will leave.
Google’s performance guidance describes faster loading and timely responses as important to retaining users. It also explains why the experience can be harder on mobile devices with limited processing power and slower connections.
The business evidence is useful, provided it stays in context. In a 2021 Vodafone case study published on web.dev, an A/B test compared two landing pages with the same design and functionality. The optimized version improved Largest Contentful Paint, a measure of when the main content appears, by 31%. It generated 8% more sales.
That was Vodafone’s result, not a forecast for your website or a result we measured for a client. It shows why performance deserves attention without proving that every improvement produces the same return.
For your site, the practical question is simpler: are visitors waiting unnecessarily before they can read, choose or contact you?
What the free Website Check actually checks
The check looks at your homepage from outside your website. It runs Google PageSpeed Insights tests for phones and computers, alongside checks of the page and its technical setup.
The report covers four areas:
- Speed: Google PageSpeed scores for mobile and desktop, load time and page size.
- Phones and accessibility: screen setup, Google’s automated accessibility score and whether content shifts while loading.
- Security: HTTPS and redirects to the secure version of the site.
- Search basics: titles, descriptions, headings, image descriptions, sitemap, structured data and social previews.
Results are explained in plain English and roll up into a report-card score out of 100. The check is free, requires no email to see the results and can take up to 90 seconds.
Its scope matters. It is an automated homepage check, not a full audit of every page, form or visitor task. An HTTPS result does not prove a site is safe from every threat. Automated accessibility checks do not prove that everyone can use it. Search basics do not guarantee rankings.
The report helps identify where a closer look may be needed. It is not a certificate that everything works.
A score out of 100 is not a percentage of perfection
Our overall report-card score and Google’s individual test scores describe different things. The report card brings several areas together. Google’s Lighthouse Performance score summarizes measured performance using a weighted calculation.
A Performance score of 90 does not mean a website is “90% as fast as it could possibly be.” Nor does 100 mean every visitor gets an instant page load.
Google’s Lighthouse scoring documentation classifies Performance scores from 90 to 100 as good. It says a perfect 100 is extremely challenging and not expected. Near the top of the scale, improving the score by one point can require a much larger improvement in the underlying measurements.
We still aim high. But a useful report should explain what the remaining findings mean, which changes would help visitors and what limits remain. A number without that explanation is an incomplete answer.
Why the result can change a minute later
A performance test is a measurement under particular conditions, not a permanent property of your website.
Google’s Lighthouse guidance on score variability explains that results can change even when the website’s code has not changed. Server load, network conditions, changing page content and the testing environment can all affect measurements.
That means consecutive tests, even a minute apart, can return different results. It does not mean the page necessarily became better or worse during that minute. There is no fixed amount by which every site’s score should vary.
Small fluctuations should be read in context. Large or repeated drops deserve investigation rather than being dismissed as “just the test.” Compare mobile with mobile and desktop with desktop, and look at the underlying timings and findings. Lighthouse recommends multiple runs rather than drawing conclusions from a single result.
The useful question is whether performance is consistently good, not whether one carefully chosen run looks impressive.
What we can improve, and what needs a wider decision
The build matters. Page weight, images and scripts all affect what a browser has to download and process. Our How we work page describes our recommendation for lightweight pages while keeping the client’s choice of platform open.
The hosting and delivery setup matters too. A remote server, variable server load or a service the page depends on can introduce delay. Google’s variability guidance specifically notes that web servers do not always respond with the same delay.
These factors are reasons to investigate, not excuses to stop improving. Some can be addressed in the build. Others may require a hosting change or a decision about a third-party feature. Our Cloudflare explainer discusses delivery alongside its limits: a host’s name alone is not a speed guarantee.
We aim for the strongest performance we can achieve for the website’s purpose. We do not promise permanent 100s across every category, identical results on every test or the same loading time on every visitor’s connection.
The visitor’s experience is the goal
Google’s PageSpeed Insights documentation distinguishes lab tests from field data. Lab tests simulate a page load and help diagnose problems. Field data describes real users’ experiences over a collection period, when enough data is available.
A good lab score does not necessarily mean every real visitor has a good experience. The free check’s report should not be read as continuous monitoring or a complete record of those visitors.
A wider review should consider whether important content appears promptly, whether the page responds to interaction and whether elements move unexpectedly. Those are the loading, responsiveness and visual-stability concerns behind Google’s Core Web Vitals.
For us, high scores are useful evidence to work toward. The reason to pursue them is the person on the other side of the screen. Your website should give that person a clear, usable first impression of you and what you offer.