< Technical support />

Training your teams in web performance is an excellent choice to ensure you master the fundamentals. With support, the focus is more on the concrete and operational aspects: we guide you and explain the reasons why a particular choice is justified or not. You thus learn "on the job" with the best school: the field.

Coaching is offered in the form of day packages, one-off or recurring, depending on the desired frequency of intervention and the size of your organization.

We can provide you with a quote upon simple request by email or phone.

We intervene remotely, via video conference, at a pace defined together: participation in your rituals (sprint planning, reviews…), dedicated work sessions, or on-demand requests.

The duration varies depending on your challenges: from a few weeks to secure a redesign to several months for a complete skills upgrade for your teams.

A web performance expert is a specialist in website loading time and responsiveness. They measure, diagnose, and fix what slows down a page, from the server to the browser. Their scope covers both front-end and infrastructure, whereas a front-end developer or a system administrator only masters one half.

It is this intersection that defines the profession. A degraded LCP can be caused by an oversized image, a slow server, missing cache, an unsuitable tag order, or a blocking third-party script : five causes spread across three different technical domains. Deciding between them requires knowing how to measure before fixing.

Expertise is recognized by three concrete signs: it relies on field data rather than a tool score, it quantifies the expected gain before intervening, and it names the target metric for each fix. A recommendation without a target metric is not an expert recommendation; it is a generic best practice.

Both meet different needs. Outsourcing addresses an existing problem in the short term; upskilling prevents it from recurring. Performance always degrades through accumulation, over the course of publications and deployments: without a trained team, the site relapses a few months after the intervention.

The decision criterion is the frequency of your production releases. If your site evolves every week, your teams will decide on its daily performance, not a one-off provider. Training then becomes more profitable than intervening, and support takes the form of skills transfer rather than a turnkey service.

In practice, combining the two yields the best results: an initial intervention to bring the site up to standard, followed by support that equips your teams and arbitrates major technical choices. The budget shifts from correction to prevention, which is cheaper in the long run.

Four situations justify it: Core Web Vitals consistently in the red on field data, a redesign project at the time of technical choices, an unexplained drop in conversion, or an internal disagreement on the causes of a slowdown. The sooner is always the cheaper, particularly during a redesign.

The most profitable time remains the writing of the specifications. Choices of stack, hosting, theme, and templates commit performance for years: correcting them afterward costs several times the price of having made them correctly from the start. A few days of upstream support avoids a full optimization service afterward.

The most reliable warning sign is not a tool score, but the gap between what you measure and what your visitors experience. A good lab score with bad field data indicates a problem that standard tools do not show, and this is typically where an outside perspective becomes useful.