Skip to content
Agencei

Support & troubleshooting

Performance optimization for your websites, applications and APIs

Performance optimization means making a website, application or API faster and more stable under load, based on measurements rather than guesswork. Agencei identifies bottlenecks through profiling, then applies targeted fixes: database queries, caching, frontend rendering, server configuration, delivery of static assets.

This service is for brochure and e-commerce sites penalised by poor Core Web Vitals, business applications whose screens take several seconds to display, and APIs whose response times degrade the mobile experience or break partner integrations.

Slowness always has a measurable cause: an SQL query without an index, an N+1 problem in an ORM, uncompressed images, render-blocking JavaScript, missing cache, undersized server. We measure before, fix in order of impact and measure after, so the gain is demonstrated rather than assumed.

When do we step in?

Core Web Vitals in the red

Google Search Console flags slow pages, LCP exceeds the recommended thresholds and your search ranking suffers. We address rendering, images, fonts and third-party JavaScript.

Store slows down at peak times

Your WooCommerce or PrestaShop site holds up during the day but collapses during a campaign or sales period. We work on caching, catalogue queries and sizing.

Business application screens take too long

A dashboard or list takes several seconds to load because of multiple queries or calculations redone on every display. We rewrite the queries and introduce caching or precomputation.

API too slow for mobile

The mobile app waits a long time for backend responses. We analyse the slowest endpoints, unnecessary serialisation and cascading calls.

Saturated database

The database server CPU is constantly high, locks pile up. We read execution plans, add missing indexes and fix expensive queries.

Cloud costs spiralling

You compensate for slowness by adding servers. Optimization often makes it possible to shrink the infrastructure while improving response times.

How we work

  1. 1

    Baseline measurement

    Recording response times, Core Web Vitals (Lighthouse, PageSpeed Insights, field data), server and database metrics. This is the comparison basis for everything that follows.

  2. 2

    Profiling and bottleneck identification

    APM tools, request traces, SQL execution plans (EXPLAIN ANALYZE), frontend loading analysis (WebPageTest, DevTools). We rank causes by their real impact on the user experience.

  3. 3

    Prioritised optimization plan

    List of actions with expected gain, effort and risk: indexes, query rewrites, Redis or HTTP caching, CDN, image compression, JavaScript splitting, server resource adjustments.

  4. 4

    Implementation in batches

    Fixes are applied in staging then in production, in small batches, with a measurement after each batch. This attributes each gain to its cause and allows a rollback if needed.

  5. 5

    Verification under load

    Load tests with k6 or an equivalent to verify behaviour at the expected peaks, not just in quiet conditions.

  6. 6

    Report and recommendations

    Before-and-after comparison, explanation of the changes, and recommendations to keep performance up: performance budgets, alerts, good practices for the team.

Technologies we use

  • Lighthouse and PageSpeed Insights
  • WebPageTest
  • Chrome DevTools (Performance, Network)
  • APM (Sentry Performance, New Relic, Datadog)
  • EXPLAIN ANALYZE (PostgreSQL, MySQL)
  • Redis
  • Varnish and HTTP caching
  • CDN (Cloudflare, CloudFront)
  • k6 and Apache Bench
  • Brotli, gzip and WebP / AVIF images
  • Nginx and PHP-FPM
  • OPcache

Why choose Agencei?

  • We measure before we act

    No optimization is applied at random. Every action is justified by a measurement and its effect is verified after deployment.

  • From the frontend to the database

    A slow page can come from the browser, the application, the server or the database. We work across the whole chain instead of limiting ourselves to one layer.

  • No unnecessary rewrite

    Most gains come from targeted fixes: an index, a cache, a rewritten query. We only propose a rebuild when it is genuinely necessary.

  • Gains that last

    We leave alerts and performance budgets in place so that slowness does not come back with the next release.

In brief

What is this service?
Agencei performance optimization is a diagnosis and remediation service that makes a website, application or API faster and more stable under load, with measurements before and after.
Who is it for?
It is aimed at online retailers and sites penalised by poor Core Web Vitals, businesses whose business application is slow, and software companies whose API degrades the mobile experience.
What problem does it solve?
Slowness loses sales, search ranking and users' time, and is often compensated with more expensive servers. The service identifies the real causes and fixes them in order of impact.
How long does it usually take?
The diagnosis usually takes a few days. Fixes range from a few days for targeted actions to several weeks for deeper work on the database or architecture.
What factors influence the price?
The price depends on the size of the platform, the number of layers to address (frontend, application, database, infrastructure), the quality of the existing code and the performance targets.
How does the engagement run?
Baseline measurement, profiling and bottleneck identification, prioritised plan, implementation in batches with measurement after each batch, load tests, report with recommendations.
What are the risks?
Optimizing without measuring leads to useless or risky changes. A misconfigured cache can serve stale data; a poorly chosen index can slow down writes.
What alternatives exist?
Increasing server resources (expensive and temporary), enabling a cache or CDN without diagnosis (partial gains), or rebuilding the application if the architecture is the main cause.

Frequently asked questions

What are the most common causes of slowness?

Server side: SQL queries without indexes, N+1 problems in an ORM, missing cache, undersized server. Browser side: oversized images, render-blocking third-party JavaScript, unoptimised fonts, no CDN. On CMS platforms, a single poorly written plugin or module is often enough to slow everything down.

What are Core Web Vitals?

They are Google's indicators for measuring page loading experience: LCP (display of the main content), INP (responsiveness to interactions) and CLS (visual stability). They influence search ranking and are measured with Lighthouse, PageSpeed Insights and Search Console field data.

Do we need to change hosting to go faster?

Not always. A bigger server hides the problem without solving it and costs more every month. We start by fixing software causes; if the infrastructure is still the limiting factor after that, we recommend appropriate resizing or migration.

How long does an optimization engagement take?

A diagnosis usually takes a few days. Fixes range from a few days for targeted actions to several weeks if the database or architecture requires deeper work. We confirm the duration after the baseline measurement.

Do you work on WordPress, WooCommerce and PrestaShop?

Yes. These platforms have their own levers: page and object caching, OPcache, catalogue query optimization, plugin or module cleanup, images, CDN. We handle them with the same measurement methods as for a custom application.

Tell us about your project

Describe your need in a few lines: we come back to you with a first analysis and the next steps.