Research & Analysis Checklist for SaaS & Startups

State the specific decision, such as whether to prioritize a feature, pursue a segment, or adjust pricing, and lock a decision date. This keeps analysis focused on actionable outcomes rather than interesting but low-impact data.

Write H1/H0 statements, for example, 'H1: PLG free plan increases activation rate in SMB by 15% within 60 days.' Hypotheses let you choose the right methods and sample sizes, and prevent post-hoc rationalization.

Choose primary KPIs like signup-to-activation, PQL rate, or demo-to-close, and guardrails such as churn, support load, and infrastructure costs. Precommitting metrics avoids cherry-picking later.

Specify firmographics, technographics, and roles, for example, mid-market, 100-1000 FTE, using AWS, data team of 3+, and buyer VP Data. Pull lists from your CRM and enrich with Clearbit or ZoomInfo to match criteria.

Decide the method mix, for example, 12 user interviews, review mining, and Similarweb + Ahrefs analysis, and set confidence thresholds like 90% for directional GTM and 95% for pricing changes. This balances speed and rigor.

List known blind spots, such as limited data for enterprise buyers or lack of visibility into private pricing, and capture risks to mitigate. This prevents overconfidence and informs where to invest more data collection.

Pull from G2, Capterra, Product Hunt, and analyst notes to enumerate direct and indirect competitors, plus substitutes. Tag each with category, ICP focus, and GTM motion to drive relevant comparisons.

Snapshot pricing pages with Visualping or Wayback Machine and store structured fields in Airtable or a repo, including SKUs, limits, overages, and add-ons. Versioned history enables temporal analysis of monetization strategy shifts.

Use Playwright or Puppeteer with proxy rotation to fetch feature lists, API endpoints, and rate limits; parse with Python and Beautiful Soup. Respect robots.txt and throttle requests to avoid blocking.

Ingest reviews from G2, Capterra, and app stores via AppFollow or Appbot, then tag themes, roles, and sentiment. This reveals churn drivers, unmet needs, and competitive differentiators in buyers’ language.

Pull Similarweb channel mix, Ahrefs keyword rankings, and Google Search Console queries to measure demand capture. Track branded versus non-branded terms to spot competitor positioning and content gaps.

Use Clearbit, People Data Labs, or Apollo to add firmographics, tech stack, and role titles to CRM records. Enrichment ensures your sample reflects the ICP, not a convenience cohort.

Create a canonical ID keyed by domain, Crunchbase permalink, or LinkedIn URL to merge variants like Acme Inc., Acme.io, and Acme AI. Accurate joins are the backbone of cross-source analysis.

Define column names, units, and definitions for metrics like Activation, PQL, and NRR; store in Notion or a repo. Consistent schemas reduce rework and misinterpretation across analysts and stakeholders.

Use dbt tests or Great Expectations to detect duplicates and outliers, and normalize fields like country codes, currencies, and date formats. Automated checks save hours and make results reliable.

Schedule ingestion with Airflow, Dagster, or Prefect and monitor freshness with alerts when datasets exceed 24 or 72-hour thresholds. Stale data leads to wrong calls in fast-moving markets.

Mask emails and names by default, encrypt at rest, and gate access through roles in BigQuery, Snowflake, or Redshift. Privacy compliance avoids risk and keeps analysis shareable across teams.

Document how you sample by segment and adjust weights if your dataset overrepresents one ICP. Weighting reduces bias when combining survey, review, and internal data sources.

Catalog core, differentiator, and parity features across competitors, linking each claim to URLs, changelogs, or docs. Evidence-backed matrices prevent debate over hearsay in roadmap planning.

Compare price points, metering units, usage limits, and enterprise add-ons; note coupon usage and procurement discounts from public quotes or community chatter. Pricing archetypes inform your monetization experiments.

Use Similarweb for traffic sources, Facebook Ad Library for paid creatives, and Ahrefs for non-branded rankings. This shows where competitors actually win attention and where you can counter-program.

Count marketplace apps, Zapier triggers/actions, and native integrations; score API coverage and rate limits from docs. Ecosystem breadth is a proxy for stickiness and enterprise readiness.

Monitor changelogs, release RSS feeds, GitHub repos, and engineering job postings on Greenhouse or Lever. Frequent releases and SRE job spikes can hint at scale issues or aggressive expansion.

Analyze homepage hero copy, proof points, and social proof across competitors; code themes like 'security', 'speed', or 'cost' using a simple tagging rubric. Positioning gaps reveal differentiable narratives.

Set up DAU, WAU, MAU, activation, and PQL definitions in Mixpanel or Amplitude, and confirm event naming standards. A solid baseline allows comparing market insights to actual in-app behavior.

Plot retention by segment, for example, SMB vs mid-market, admin vs IC, and tech stack cohorts. This illuminates where your product is already resonating and should guide qualitative deep dives.

Export Zendesk or Intercom tickets and Gong or Chorus call transcripts, then categorize blockers like SSO gaps, limits, or migration pain. VoC from real accounts grounds competitive claims in reality.

Recruit 8-12 participants per ICP and use a semi-structured guide focused on switching triggers, alternatives tried, and must-have criteria. Record with Zoom and transcribe via Otter or Whisper for coding.

Use Van Westendorp to bracket willingness to pay and Gabor-Granger or conjoint via tools like Sawtooth for packaging sensitivity. Tie outputs to your metering unit and guardrails like gross margin.

Cluster needs by Job-to-be-Done, anxieties, and desired outcomes using a coding tool like Dovetail or Aurelius. JTBD framing helps translate themes into features that win deals and reduce churn.

Create a 1-2 page readout that lists decisions, evidence strength, and recommended actions, with links to backing analysis. Busy leaders need the punchline first to unblock roadmaps and GTM moves.

Publish Looker Studio, Tableau, or Metabase dashboards with source queries in Git, and use Jupyter or Hex for deeper analysis. Reproducibility ensures updates do not break the narrative.

For each conclusion, state confidence level, main assumptions, and what data would change your mind. This makes trade-offs explicit and guides follow-on research.

Use install base estimates from BuiltWith or Wappalyzer, usage proxies, and pricing to build a bottom-up market model. Cross-check with top-down analyst estimates for sanity.

Schedule weekly scrapes and diff checks for pricing, feature pages, and docs; send Slack alerts on significant changes. Staying current turns research into a system rather than a snapshot.

Store data, code, and reports in versioned repos; add README, data provenance, and owners. Proper lineage and permissions reduce rework and prevent accidental leaks.