Web Performance
Project workbook · Path 91

Create a resource budget and an optimization plan, then compare local page behavior without claiming an unmeasured performance score.

The case

A fictional workshop page has a teaching resource budget: HTML 100 kB, images 800 kB and JavaScript 300 kB. A proposed revision keeps HTML at 100, reduces images to 240 and JavaScript to 120. These are scenario values, not measured network transfers from WikiFree.

ResourceBefore kBProposed kB
HTML100100
Images800240
JavaScript300120
STAGE 1

Define the performance question

Specify the user task, device conditions and page state before measuring. A page can transfer few bytes yet block interaction with expensive script work. Separate laboratory tests from real-user field data. The case begins with a hypothetical byte budget so calculations stay distinct from browser observations.

Your task

Write a measurement plan and list the scenario assumptions.

Required output

Performance brief

Record your result and evidence here or in a separate file.

Review gate

Hypothetical byte values are not labeled measured.

If the gate is not met

Add the test conditions beside every future measurement field.

STAGE 2

Inspect the loading path

The browser discovers and processes resources in an order affected by HTML, CSS and scripts. Identify resources needed for the first useful view and those that can wait. Avoid adding preload hints blindly; prioritizing everything defeats prioritization. Keep the core lesson text available without JavaScript.

Your task

Annotate the local fixture’s critical and deferred resources.

Required output

Loading map

Record your result and evidence here or in a separate file.

Review gate

Each proposed deferral preserves the essential content.

If the gate is not met

Test the page with scripting unavailable before calling it resilient.

STAGE 3

Set a resource budget

A budget makes regressions discussable. Calculate the proposed savings and identify which assets change. Optimize images with suitable dimensions and encoding, and remove unnecessary script work. The scenario does not justify a claim about Core Web Vitals or a particular PageSpeed score.

Your task

Calculate category savings and the total reduction.

Required output

Resource budget table

Record your result and evidence here or in a separate file.

Review gate

Total changes from 1,200 to 460 kB, a 61.67% reduction.

If the gate is not met

Compute percentage savings against the original total.

STAGE 4

Run a controlled comparison

Use the same browser conditions and repeat a test rather than selecting the best run. Record what actually happened, including cache state and viewport. The supplied local fixture lets you inspect resource and rendering behavior; it is not a clone of a live production page.

Your task

Record three local runs for the fixture and report the range, if the tools are available.

Required output

Measurement log

Record your result and evidence here or in a separate file.

Review gate

No unperformed metric is filled with an estimate disguised as a result.

If the gate is not met

Leave unavailable measurements blank with a reason.

STAGE 5

Write a prioritized improvement plan

Prioritize changes by expected user benefit, effort and confidence. Separate measured bottlenecks from hypotheses and include a regression check. A performance change should not remove accessibility, content or necessary functionality. State what must be measured after a real deployment before claiming an improvement for users.

Your task

Write three actions and an acceptance check for each.

Required output

Performance handoff

Record your result and evidence here or in a separate file.

Review gate

The plan includes a post-change comparison under the same conditions.

If the gate is not met

Replace guaranteed score claims with measurable acceptance criteria.

Final project

Deliver the measurement plan, loading map, budget calculation, actual local observations and prioritized actions. Explain one optimization that would be inappropriate for above-the-fold content.

Critical condition

Scenario estimates and measured browser results are clearly separated; no performance score is fabricated.

CriterionPointsEvidence
Evidence and correctness30Calculations, source use and results match the supplied case; limitations are honest.
Required artifacts30All five stage outputs and the capstone handoff are present and inspectable.
Reasoning and decisions20Choices, assumptions, alternatives and the case-specific critical condition are explained.
Reproducibility10A reviewer can follow the stated steps and distinguish expected from observed results.
Clear communication10Labels, units, structure and evidence references make the work easy to review.

Score at least 80/100, supply every required artifact, and satisfy the path-specific critical condition. Rework missing evidence before marking complete.

Worked reference

The proposed total is 100 + 240 + 120 = 460 kB, compared with 1,200 kB. Savings are 740 kB or 61.67%. This is a resource-budget calculation, not a measured load-time or PageSpeed result.

Boundaries

This is a local performance exercise, not a live-site optimization or a guarantee of field metrics.