About the service
I speed up sites in the order that actually works: measure first, argue later. Before I touch a single setting I take a baseline on your real pages — the home page, a category or feed, a product card or an article, the search results, the checkout flow — and write down server response time, page weight, the moment content becomes visible and the moment the page becomes usable. I do it twice: on a fast connection and on a deliberately throttled mobile one, because those two sets of numbers differ several times over and they call for different fixes.
Then comes the diagnosis, in plain words. Usually the list is short and unglamorous: one page firing hundreds of near-identical database queries; photographs served at the size they came off the camera instead of the size they are displayed at; a dozen scripts and fonts holding up rendering; caching that is missing altogether, or present and configured so that it never actually hits. From that list I work: page cache and object cache, response compression, modern image formats with a fallback for older browsers, deferred loading for everything below the first screen, correct HTTP cache headers and long-lived names for static files, web server and process pool tuned to your real traffic rather than to a template.
Three things I say up front so they are never a surprise. The work is done with a backup in place, and every change is made so that it can be rolled back on its own — if something behaves oddly we undo one step, not a day of work. Caching is not switched on for the cart, the checkout or the personal account: on those pages a visitor would be shown somebody else's data, and no amount of speed is worth that trade. And I will not promise a particular score in a popular testing tool before the measurement — until I have seen your pages that figure would be a guess wearing the clothes of a commitment. The work ends the way it started: the same measurement, the same pages, the same method, and a report with the before and after side by side.
What the work includes:
— baseline measurement on real pages, fast connection and slow mobile;
— a written diagnosis: database queries, images, blocking scripts, missing cache;
— page and object caching, with cart, checkout and account excluded;
— compression, modern image formats and correct display sizes;
— deferred loading and priorities for the first screen;
— HTTP cache headers and long-lived names for static files;
— web server and process pool configuration;
— a repeat measurement and a report with the numbers.
Then comes the diagnosis, in plain words. Usually the list is short and unglamorous: one page firing hundreds of near-identical database queries; photographs served at the size they came off the camera instead of the size they are displayed at; a dozen scripts and fonts holding up rendering; caching that is missing altogether, or present and configured so that it never actually hits. From that list I work: page cache and object cache, response compression, modern image formats with a fallback for older browsers, deferred loading for everything below the first screen, correct HTTP cache headers and long-lived names for static files, web server and process pool tuned to your real traffic rather than to a template.
Three things I say up front so they are never a surprise. The work is done with a backup in place, and every change is made so that it can be rolled back on its own — if something behaves oddly we undo one step, not a day of work. Caching is not switched on for the cart, the checkout or the personal account: on those pages a visitor would be shown somebody else's data, and no amount of speed is worth that trade. And I will not promise a particular score in a popular testing tool before the measurement — until I have seen your pages that figure would be a guess wearing the clothes of a commitment. The work ends the way it started: the same measurement, the same pages, the same method, and a report with the before and after side by side.
What the work includes:
— baseline measurement on real pages, fast connection and slow mobile;
— a written diagnosis: database queries, images, blocking scripts, missing cache;
— page and object caching, with cart, checkout and account excluded;
— compression, modern image formats and correct display sizes;
— deferred loading and priorities for the first screen;
— HTTP cache headers and long-lived names for static files;
— web server and process pool configuration;
— a repeat measurement and a report with the numbers.
