Answer library
What should you check before an audit, after a release or before a client conversation? Start with a question. Each answer includes its scope, limits and sources.
The useful output is a dated priority list with evidence, not a single score.
Read the answerHow this library works
An answer is a starting point for a decision. Check what it covers and which evidence it uses.
Get to the point without digging through a long article.
See what can be assessed and what the answer cannot settle.
Each question has references and a last-reviewed date.
Find a question
Go to your team’s group or search for a topic. Open any question without leaving the page.
3 answers
An agency audit should start with the agreed URL scope and then cover the technical signals that can change the next delivery decision: crawl and indexability, performance, accessibility, public security signals and page weight. The useful output is a dated priority list with evidence, not a single score.
Choose a tool that preserves the URL, source module, severity or priority, evidence and retest context for every finding. An agency needs a client-readable decision trail, so a raw scanner export is only useful when it can be turned into a scoped backlog without hiding what was not checked.
Start by grouping findings by affected URL or template, then compare scope, confidence, user-facing risk and implementation cost. A high-impact finding is one that is supported by evidence and affects an important part of the agreed site scope; it is not simply the item with the most alarming colour.
3 answers
Before taking responsibility for a client site, record the domain, important templates, current redirects and canonical signals, robots and sitemap access, baseline performance, visible accessibility risks and public security headers. This creates a shared starting point and prevents an inherited issue from being mistaken for a new regression.
A freelancer can use a public audit to collect repeatable evidence for a handoff: what was checked, which URL is affected, how urgent the issue is and what should be retested. The handoff becomes safer when open questions and out-of-scope areas are written beside the result.
Use three layers: what was checked and when, the top findings grouped by user-facing risk (not by tool name), and what you recommend fixing first with a realistic effort note. Keep the client summary to one page and link to evidence only when they ask for detail — small-business clients decide on clarity and trust, not export volume.
5 answers
A SaaS team should audit the marketing site as a connected delivery surface: crawlability and technical SEO, performance across important templates, accessibility of the product story and forms, public security signals, and the clarity of the resulting backlog. The audit should make the next engineering or content decision easier.
When Core Web Vitals regress, compare the same URL group and device segment, then separate field status from a controlled lab diagnosis. Use the lab to investigate payload, rendering and interaction causes; use CrUX when available to understand the real-user trend. Do not turn missing field data into zero.
An automated audit is a useful pre-launch gate for observable website signals, but it is not the whole release review. Use it to catch repeatable technical, performance, accessibility and public security findings, then add manual checks for the critical journey and verify the deployed version in the same scope.
State the audited URL set, modules, measurement date and data sources; explain what each module can and cannot prove; separate lab from field data; and list authenticated, backend and legal areas that stayed out of scope. A SaaS report should make it obvious which conclusions are evidence-backed and which need a follow-up manual check.
A website carbon footprint module estimates emissions per pageview from data transfer, an energy model (CometWeb uses SWDM v4), grid carbon intensity and a chosen network/device scenario — it does not directly measure server or user-device energy. That estimate is useful for tracking transfer-reduction progress release over release and for supporting sustainability conversations with prospects who ask about it, but it is one input, not a certification.
3 answers
Before a seasonal campaign, check the public campaign landing page, key category and product templates, crawl/indexing signals, page weight and Core Web Vitals, accessibility of the purchase path, and visible security configuration. Freeze the scope and date so campaign changes can be retested against the same baseline.
Walk through the checkout in a test environment using only a keyboard, then repeat it with a screen reader. Check that focus is visible and follows the steps, fields have clear labels, and errors explain which field needs fixing. Run automated checks on accessible pages, then test validation and the payment handoff manually.
Prioritise campaign fixes by asking which issue affects an important template or journey, how confident the evidence is, how quickly it can be shipped and whether the same component appears elsewhere. Validate the fix after release with the same URLs and test conditions, then wait for field data to reflect the new window.
3 answers
Connect SEO, performance and accessibility work around the same URL or template scope and a shared dated backlog. This keeps a technical SEO issue from being separated from the performance or interface change that caused it, while still preserving the specialist checks and their limits.
A web team can use one prioritized quality backlog when each item keeps its source module, affected URL, evidence, status, owner and retest condition. The value is traceability across releases, not the claim that one tool replaces every specialist system.
After a migration or template change, re-check the same URL inventory: status codes, redirect chains, canonical targets, hreflang pairs, internal links to key templates and indexation signals in Search Console. Compare against a dated pre-migration export where possible and record what was intentionally removed versus accidentally dropped.
Sources help you assess an answer. They do not guarantee rankings, AI citations, legal compliance, security or conversions.
An answer helps set direction. An Insight report shows the data and priorities for a specific site.
