Research & Analysis Checklist for Web Development

Write a one-page brief that states user problem, constraints, and 2 to 3 measurable KPIs like conversion, time to interactive, and error rate. This keeps research focused and connects findings to delivery outcomes.

Catalog core features, pricing, authentication models, and SDK coverage across 5 to 8 competitors. Highlight gaps that materially impact adoption like SSO support, rate limits, and per-seat pricing.

Run Lighthouse and WebPageTest on the same pages and profiles, recording LCP, INP, TTFB, and payload sizes. Document network conditions and device profiles so results are reproducible.

Scrape GitHub releases or RSS feeds to gauge frequency of security patches, breaking changes, and feature velocity. Stability signals reduce maintenance risk and inform upgrade plans.

Tag themes from GitHub issues, Stack Overflow, and G2 like flaky auth flows or poor SDK ergonomics. Prioritize pain points that directly map to your target user segments.

Compare REST and GraphQL schemas, pagination patterns, auth methods like OAuth 2.0 with PKCE, rate limits, and sandbox provisioning time. Faster time to first successful call correlates with adoption.

Use GA4 or Amplitude to map step-level drop off, latency at each step, and error rates. Snapshot current numbers before changes so improvements are attributable.

Break down funnels and Core Web Vitals by mobile vs desktop, 3G vs Wi-Fi, and region. This reveals whether performance work needs regional CDNs, image compression, or adaptive loading.

Use Hotjar, FullStory, or Heap to identify rage clicks, dead zones, and scroll depth gaps. Triangulate with analytics to validate whether design changes or copy fixes are needed.

Synthesize 8 to 12 interviews using affinity mapping with tags like trust, speed, and control. Tie each theme to a measurable behavior in analytics to avoid anecdotal bias.

Create tags like auth errors, API pagination, and timeouts in your help desk. Monthly trend reports highlight classes of bugs that need automated tests or design changes.

Use Datadog, ELK, or BigQuery to group by status codes, p95 latency, and endpoint frequency. Tie hotspots to code owners and estimate users impacted per day.

Run aXe and Lighthouse a11y audits, then verify keyboard navigation, focus order, and contrast manually. Accessibility defects block adoption and often correlate with general UX issues.

Map route types to rendering strategies in Next.js, Remix, or Nuxt based on personalization needs, cacheability, and TTFB budgets. Document how each choice affects edge caching and hydration cost.

Use Vite or Webpack bundle analyzer to identify large transitive dependencies, polyfills, and duplicated packages. Propose esm-first replacements and code splitting points.

Contrast serverless limits like cold starts and execution time with containerized workloads on Kubernetes. Note impact on long-lived connections like WebSockets and streaming.

Use a decision matrix across PostgreSQL, MySQL, and document stores with read to write ratio, indexing needs, and multi region replication. Justify choices with EXPLAIN ANALYZE samples and storage costs.

Define what to cache at CDN with TTL and ETag, what to memoize in the app, and when to use Redis for materialized results. Document invalidation rules to keep data consistent.

Compare REST with OpenAPI, GraphQL, and tRPC for your domain, considering caching, N+1 risk, and federation. Define versioning policy to avoid breaking integrators.

Score libraries using npm download trends, release recency, open issue velocity, and bus factor. Prefer projects with clear roadmaps and type definitions.

Combine Lighthouse lab runs with real-user monitoring to collect LCP, CLS, and INP. Prioritize fixes that improve p75 for your top traffic pages and devices.

Use k6 or Locust to simulate peak concurrency, think-time, and common workflows. Capture p95 latency, error rate, and saturation metrics to identify bottlenecks early.

Instrument traces, metrics, and logs consistently, then pipe to Datadog, New Relic, or Grafana. Define alert policies that page on user impact thresholds, not noise.

Set availability and latency SLOs per critical route, then track monthly error budgets. Gate risky deploys when budgets are exhausted to protect reliability.

Run a lightweight STRIDE or LINDDUN session on new flows, then automate Snyk or Dependabot and npm audit. Triage findings by reachable paths and production exploitability.

Enforce a strict CSP with nonces or hashes, set Secure and HttpOnly cookies with SameSite, and rotate JWT secrets. Verify CSRF protections on all state changing requests.

Document what personal data is collected, where it is stored, retention periods, and processors. Confirm GDPR and CCPA requirements and update your DPA and consent flows.

Track time to green, queue time, and test retry rates across GitHub Actions or GitLab CI. Fix top flaky tests and parallelize slow jobs to reduce lead time.

Report unit, integration, and E2E coverage with Jest, Vitest, and Playwright or Cypress. Add tests for high risk flows like payments, auth, and data export.

Measure time to first review, time to merge, and review depth by team. Pilot smaller PR guidelines and code owners to improve velocity without sacrificing quality.

Check OpenAPI accuracy, example requests in Postman or Insomnia, and Storybook component coverage. Prioritize docs tied to high support volume and new integrators.

Score reach, impact, confidence, and effort for each recommended change. Create a top 10 list that aligns with business goals and current capacity.

Use cycle time and throughput data for Monte Carlo simulations to produce probabilistic dates. Share ranges, not single dates, to set realistic expectations.

Write a one-page summary with key findings, chosen option, tradeoffs, and owners for follow up tasks. Link to dashboards and raw data to maintain transparency.