Skip to main content

Optimization Log

An optimization log turns "tuning" into a professional, auditable process. You can prove what worked, revert what broke, and avoid guesswork.

Why This Matters

Without a log, optimization becomes guesswork. You make 10 changes, performance improves, but you cannot identify which change helped (or which one introduced a regression). A structured log:

  • Proves impact with before/after metrics for each change.
  • Enables rollback by documenting how to reverse every change.
  • Builds client trust with professional documentation.
  • Prevents conflicts by recording tool assignments (e.g., Perfmatters handles CSS, LSCache handles page caching).

Optimization Log Template

Track every change in a structured table:

StepChange MadeTool UsedResultMetrics ImpactStatusNotes
1Disabled Cloudflare auto-minifyCloudflareOKAvoid double-minify conflictsKept
2Remove unused CSSPerfmattersOKLCP: ___ -> ___KeptAdded exclusions for nav/menu
3Delay JSPerfmattersOKINP: ___ -> ___KeptExcluded critical scripts
4Preload hero imagePerfmattersOKLCP: ___ -> ___Kept
5Enable guest modeLSCacheOKTTFB/cache hit improvedKept
6Enable Brotli + HTTP/3CloudflareOKNetwork handshake improvedKept
7Lazy load offscreen imagesPerfmattersOKBytes reducedKeptLCP excluded
8Menu regression fixPerfmattersFixedUI restoredKeptAdded exclusion for menu script

Rule: log every change immediately.

Rollback Plan

If an optimization breaks something, follow a feature-by-feature rollback approach. Do not disable an entire plugin at once:

IssueRollback Action
CSS broken layoutDisable "Remove Unused CSS" temporarily
JS buttons not workingDisable "Delay JS" temporarily
Page blank or 500 errorDisable all Perfmatters optimizations
Cache loop issueDisable LSCache Guest Optimization
Site shows old versionPurge Cloudflare cache + LSCache
Broken scriptsCheck browser console + add JS exclusions
caution

Always rollback feature-by-feature, not plugin-by-plugin. Disabling an entire plugin removes 10 optimizations when only 1 caused the issue.

Before/After Metrics Table

Capture these metrics before starting and after completing optimization:

MetricBeforeAfterGain
LCP (Mobile)
INP
CLS
FCP
TTFB (Global)
Page Size
HTTP Requests
PSI Score (Mobile)
PSI Score (Desktop)

Save screenshots of PageSpeed Insights results as proof for clients and for your own reference.

Final Settings Record

After completing optimization, document the final configuration snapshot:

CategorySetting
PHP Versione.g., 8.3
Web ServerOpenLiteSpeed / Nginx
Cloudflare StatusActive, SSL Full (Strict), APO enabled
LSCache SettingsPage TTL, Guest Mode, ESI status
Perfmatters SettingsCSS optimization, JS delay, preload list
Lazy Load ConfigExcluded images, threshold
CDN Cache StatusHIT ratio, Edge TTL
TTFB ValuesHomepage, product page, search

Client Delivery Template

For client-facing optimization projects, prepare a delivery report:

SectionContent
Speed SummaryOne-paragraph overview of what was achieved
Before/After ComparisonSide-by-side metrics with screenshots
Technical Stack SummaryServer, caching, CDN, and optimization tool configuration
Core Web Vitals ProofPageSpeed Insights screenshots showing "Good" status
Maintenance RecommendationsMonthly checks, update cadence, monitoring alerts
Support ContactHow to reach you for issues or questions

Checklist

  • Optimization log table created and maintained for every change.
  • Rollback plan documented for each optimization.
  • Before/after metrics captured with screenshots.
  • Final settings snapshot recorded.
  • Client delivery report prepared (if client project).
  • Log stored in accessible location (Notion, Git, shared docs).

Summary

An optimization log transforms ad-hoc performance work into a professional, auditable process. Log every change with its tool, result, and metrics impact. Document rollback steps for each optimization. Capture before/after metrics to prove value. This discipline ensures that every optimization project is repeatable, reversible, and demonstrably effective.

What's Next